Complete Guide to Midjourney v5. 2: Learn Various Midjourney Techniques & Styles and Proper Promptin

1. Introduction: You can now be a Picasso, Van Gogh or Banksy. You can be an interior designer, a Logo Maker, a food photographer, a UI designer, all in a few seconds, this amazing image of a fashion shoot. I created this in 10 s using the following prompt. How about this snowy landscape? Again, did it with just a few words. There's literally millions of different Art that you can create with Midjourney. Just like how the calculator allowed users to quickly solve calculations that are record speeds. Midjourney enables you to create amazing Art even faster. But that doesn't mean it's easy. It does take time to master it out. Room Prompts. Prompts are the inputs to your Midjourney engine. Just the slightest difference in Prompts can completely change your image. Even adding a comma can make your Creation go from this to this. Adding one word can take your masterpiece from this to this. And that's what this course is all about. This is the only course that serves as a complete Guide to Midjourney Prompting, starting with the basics and turning you into a master by showing you the Techniques. Most other courses do this by drowning you, by showing you countless different examples of how you can use Midjourney. But they don't teach you the underlying techniques, skills, building blocks, and parameters that you need to know. The only way you can completely understand Midjourney and how it works is by knowing those building blocks. In this course, we learned all these specific Midjourney Prompt building blocks and techniques that you would ever need, like Emphasis, Lighting, Camera, and Styles. In fact, we go into Over 30 plus Techniques. For each one, we go through both the theoretical and practical. At the end of this course, you will have everything you need to create your own prompts that generate whatever you can imagine. There's something truly amazing about instantly taking something in your mind and making it real. You will also get access to several resources, including 40 pages of prompts that we use inside this course. You'll be able to create photorealistic images, landscapes, Mock-ups, cartoons, Icons, Interior Design. Anything that you can imagine, get started today and I promise you will not regret it, will see you inside 2. Course Overview and What AI Is: Welcome to the course. This course has been crafted in a way that even individuals with minimal computer and knowledge can begin generating astonishing images and designs using Midjourney. Now, if you consider yourself in advanced user, that's fine. Feel free to skip certain videos that cover basic steps like installing and using CT of navigation and through Discord and documentation pages. However, I do strongly encourage everyone to try and follow each step of the course. By doing so, you will develop a systematic knowledge and gain the necessary knowledge of using Midjourney as a graphical Generative tool. You're private and commercial purposes. Now, let's go through an overview of our course and our objectives. So this isn't the first chapter where we will go through the basics and learn about what is AI, will also check different use cases of Midjourney and the resulting images that I can employ. We will also explore what is Midjourney and its mission. And we'll also briefly check out their public Gallery page. And then at the end of the chapter, we will observe a few alternatives to Midjourney. In this second chapter, we will install and set up Discord, will register an account, and we'll also create a Midjourney account and learn how to connect it to our Discord. Unfortunately, at this point, Midjourney can only be accessed through discord, so we'll need to use it, will go through different subscription plans that are available. We'll check out the Discord up Community of Midjourney. We'll go through the basic functionalities of Discord and Midjourney. And then we'll also set up our private Discord server that would help us to organize things and improve our productivity. What effect? Now, from the third chapter, we'll begin our practice. We will learn what prompts are, will learn how to structure them. And we will start to learn how to compose Visually Appealing scenes with different subjects, surroundings, and different stylization. We will learn how to combine multiple prompts in one and then adjust a weight or impact of each prompt. Which is really how you can customize an image together. And it's really cool how this all works. We'll also go through all of the different parameters and Midjourney supports. Then we'll begin with more advanced Prompts. Also check other Midjourney Versions and their use cases. And we'll also learn a full spectrum of Midjourney functionalities and parameters and explore how those extra commands that will help us to extract the maximum use out of this tool. Perfect. Let's move on now to the fourth chapter on the top-right. Now this will be all about photorealistic Content and specific topics. We'll actually learn and apply some techniques from the world of real professional photographer. You will learn how different camera models, films, lenses, and lighting setups can affect our scene. And we'll then master our Composition and prompt to achieve maximum photo realism. Well, Use than Image References to try to blend multiple images into one. Well then utilize prompt crime writers that we learned in chapter three and also go through specific topics like people photography, fashion, Interior Design, and many other special topics that I'm very excited about. Starting from chapter five, we'll learn other use cases of Midjourney, will learn how to generate Logos, Mock-ups, UI designs, illustrations, and so on. We'll also explored the most popular use cases of using Midjourney for commercial purposes. Like using it to generate UI elements for applications and games. And along the way, we will learn how to use best practices for each individual type of illustration. In chapter six, we learn how to enlarge our images using Ai up scalars. To improve the quality and dimensions. We'll go through basic imaging editing and photoshop to polish the results and make them perfect. And then in chapter seven, we also discuss things like Midjourney Limitations, Ethical Considerations, and copyright laws. So please do not skip this chapter because it's very important. Well then try to use chat should be team to help us generate prompts and ideas. And it will explore a couple of prompt banks and the Midjourney roadmap. And then we'll finalize the course and discuss key takeaways. So that's the overview of the whole course and objectives So this is our first introductory chapter. And what we'll do is we'll embark on an exciting journey into the captivating world of AI. Okay? Now, it's a remarkable ability now to create studying images. So get ready to discover how AI can unlock your creative potential and bring your imagination to life. Now let's start by understanding what Ai is and how it can be harnessed to manifest your ideas. So let's start off with a baseline. What exactly is artificial intelligence? At its core, AI refers to the development of computer systems that can perform tasks and that typically require human intelligence. It evolves, creating algorithms and models that can process information. It can learn from data and make decisions or so-called predictions. Now, in this realm of Image creation, AI algorithms can actually analyze patterns, colors, shapes, styles to generate images that match underlying data. Now, Ai has opened up a whole new world of possibilities for artists, designers, and creatives. It provides a much powerful to set, to bring their ideas to life in ways that were previously unimaginable. With AI, you can explore new dimensions or creativity. And you can express your revision and very unique and compelling ways. Machine-learning and deep learning are fundamental techniques that go underneath Ai and that drive image Creation. This is what allows us to generate realistic and high-quality images. Machine learning, or ML for short, is a powerful approach that allows AI models to learn from large data sets by identifying patterns and relationships within that data. It enables the model to understand the underlying structures and futures of images. Deep learning, which is actually a subset of ML, takes us process further by utilizing neural networks. Those networks are designed to simulate the human brain structure and function, allowing them to process complex information and extract meaningful representations. Now, through training, those deep neural networks can develop an understanding of various image characteristics such as shape, color, textures, and objects. This training enables the models to generate images that resemble relevant scenes. And the combination of machine learning and deep learning techniques provides a powerful framework for image generation in Ai. Those techniques have already revolutionized the field, allowing this to creatively and Visually Appealing Images and highly realistic images to the power of the computational intelligence. Artificial neural networks, or CNNs for short, are a computational models that basically try to represent what's going on in your brain. Initially inspired by the biological account or Part a, it ends have evolved to prioritize practical results and solve complex problems efficiently. Now what are the key strengths of ANN is their ability to comprehend and represent intricate relationships that may not follow a linear path. Now, unlike traditional algorithms, ANS can capture non-linear patterns. I can make sense of very complex data. So ended ENS consists of interconnected artificial neurons organized in various different layers. Now those neurons are designed to process and transmit information. So the output of some neurons become the input for others. Allowing for the creation of complex networks, enable learning, and allow decision-making. Now by adjusting the strength of connections between neurons, an ENS can learn from training data and then adapt their behavior accordingly. This process is known as training or learning, and it involves iteratively adjusting the networks perimeter to optimize its performance in solving a specific task. This interconnected structure and learning process. A and ends can recognize patterns, classify objects, make predictions, generate outputs based on their input data Now they have already become very powerful tools in various fields, including image recognition, natural language processing, and autonomous systems. While edit a and ends were initially inspired by the biological brain, there are true strength really arise in their ability to efficiently process complex information and extract meaningful insights from data. So here we see a very, very basic example of how AI can learn to associate certain characteristics of an object and then try to predict the actual input data. So along the process, one eventually Ai recognizes or example this shape here It's associates the input with the data that is already known, which is illustrated here. And then it basically predicts that this should be an elephant. And also, we see that this part is unsupervised learning, meaning that the AI learns to associate the characteristics of this object with what it's learned already. But at this moment, it doesn't know it's an elephant. Okay? And then comes the supervised learning parts. So here is a person that would need to attach a label saying that this is an elephant. So just note that this is a very basic explanation of the whole process. And with recent developments and Ai, it's already becoming a semi-autonomous process. But exactly this process allows tools like Midjourney to reverse the process. So when we actually type in the text, for example, elephant, the AI starts searching the underlying dataset or objects and characteristics that are associated with this word. And then using it's artificial neural network or ANN, it tries to guess and produces an image that it thinks we've prompted for. Okay, so let's summarize it. Ai learning from input images involves a two-step initial guessing and guided refinement. When presented with an image and put, the Ai attempts to guess the object or elements contained within it. This initial guess is based on the pre-existing knowledge and the patterns as learned from various amounts of training data. However, to improve accuracy, the AI relies on supervised help from developers and analysis. They provide annotations or labels. For example, that this is an elephant that guide the Ai in identifying the correct object and the image. Now these annotations or labels act as valuable feedback, allowing the AI to learn and refine its understanding of various visual concepts. So texts to Image Prompting basically reverses this process, right? It's, instead of starting with an image and guessing the objects, the AI generates an image based on our textual Prompt or labels. Now this approach combines the fusion of language and vision. The AI models are trained on, on, on an extensive dataset that pairs textual descriptions with corresponding images. Now, during the training process, the AI learns to understand the complex connections between words and visual concepts by analyzing them provided that datasets, the models, acquire the ability to translate words into visual representations. Now this understanding enables them to generate images that aligned with given textual descriptions. For example, this amazing corky here playing a trumpet. By combining the language intervention models and they mutually informative manner. Ai models can generate images that are closely aligned with the provided textual prompts. And this and return opens up so many exciting possibilities for creative expression. Visual storytelling, and generating visuals not based on an artist or based on Art, but based on textual descriptions instead. Now, imagine a world where imagination knows no bounds, where your wildest ideas and creative visions can be transformed in a tangible visual representation. Now with cutting edge tools like Midjourney, you have the power to breathe life into your ideas. Whether you dream of all inspiring landscapes that transport your viewers to distant realms. You have gained character designs and concepts that leap off the page and capture player's attention. Fashion concept designs that push the boundaries of Style and expression. Or interior decor ideas that create unique and immersive spaces. Here, Ai is your trusted ally on this creative journey. Okay, So it's time to now unleash your creative vision and let Midjourney be the catalyst that propels your creativity to new heights. Guys, the future of Art and expression is now literally in your hands on a keyboard. And with Midjourney, your possibilities are boundless. 3. Midjourney Use Cases and Examples: In this video, we'll explore the versatility of a AI-generated Content, will check different applications and Use Cases of that Content. And then finally, we'll also explore the impact of a AI-generated images in the realms of digital design, entertainment, and Fashion. Now, in Fashion, Ai has already started to affect the Creative Landscape. Offering innovative solutions or Fashion Designers. They can now exploring new creative horizons and push the boundaries of the fashion industry. Ai algorithms can generate absolutely unique and never before seen designs that capture the essence, Style, and innovation of whatever the author is saying. Now this fusion of a human expertise and a AI-generated concepts gives a possibility to explore the Uncharted fashion territories. Now, let's go through a few examples that we will be studying in this course. Let's see here. So this is just one of many where this prompt, we've got a Fashion model on a Fashion walk in quite an unusual and surrealistic dress. So let's take a closer look at this one that we generated that completely from Ai. Here's another similar one. And you can really see how AI interprets our Prompt and tries to create something unique and original. Now, let's trick out one more. Here we go. This is a really photorealistic one. Here we have another very accurate and high-quality photo, which was actually just generated by Midjourney by one of the prompts that will learn about in a later section. And the prompt actually was a, let's see here, streets Style magazine photo with a full body shot of a Mexican man with black hair. Confident look, walking through the crowded street holding his phone and wearing a navy blue modern suit. And we can see that this is the result than Midjourney version 5.1. It absolutely nailed it, nailed it. And we can see that we have specified in her prompt, everything is in this photo. Now, let's see some other example, Use Cases. So Midjourney has already started to transform the field of design and arts by automating the creation of captivating artworks, illustrations, and graphics. With Midjourney designers can now unleash their creativity and witnessed the magic of AI in action without spending days for manual drawings. By leveraging the platforms, advanced algorithms, intricate illustrations, and eye-catching graphics. They can be effectively craft perfectly aligned with the desired creative vision and texts prompt. Now, we can use this for branding, for marketing materials, web design, digital Art. Really the possibilities are endless here. And Midjourney offers a wide range of applications. It's intuitive interface and intelligent features empower designers to enhance their productivity and achieve impressive results in only a fraction of time. Designers can now easily, more, sorry, can now easily explore rather Artistic and new directions. Experiment with different designs and Styles and push the boundaries of their own creative expression. With this approach, professionals can now elevate their designs to new levels of excellence. Now, let's also, let's also see a few examples. We're going to cover all of this in this course, so don't worry about it for now. But here we have interesting book covers. Now, these can be used as Mock-ups where designers can just nicely placed on top of the actual illustration of a Book Cover on the top, which is really nice right over here. Now. Alternatively, you can also create Mock-ups or images, right? So this is like the real physical paintings, where it's not only enough to place her idea or an image on a canvas, for example. But you can actually frame it as well. Which is very cool. So this goes a lot with if you want to, if you're selling a painting and you want to sell it on Amazon, you can create the sort of framed backgrounds if you want. Which is cool. And really same goes with digital applications, right? Like websites or apps, where we need to create mockups of things, which is really cool. This is one example that I made where if you've creating an app that shows the skyline of something, you can kind of have a boat, a computer, and a phone show that you see these in marketing materials all the time and this is what they use for it. We also have UI elements where mid journey is perfectly capable of generating different visual elements for the user interface, for it, Application, website, or even in Game, which has lot of PFAD. I mean, it's really, the possibilities are endless. After you've created this, right, designers would only need to cut out all of the elements and implement them in their application. Finally, in the realm of gaming Art, Midjourney offers unprecedented possibilities for artists to explore at unleash their creativity. Artists can now generate intricate and visually striking artworks, getting prompts and experiment with different styles, concepts, and compositions. Obviously, I'm not going to leave you all hanging here. Let's explore a couple of funding samples that we've created. So here, for example, Midjourney generated on our request different prompts for a gain or Game Elements and object. We have closely specify the style and other details so we can match the game aesthetics and design or the idea that we have. Let's see, I have a few more here that I created and check out this one here is an example of character or a game character concept design, which can be used by actual game developers, developers, and designers to create a 3D model based on this concept. I can't wait to show you everything that we've created because trust me, it's very exciting. So those were the just reuse Cases or three niches where Midjourney is already used by designers and creative studios. But this was just a fraction of use cases that are possible. Okay, just a small fraction. And I can't wait to show you all the different ways we can use Midjourney. Awesome 4. Midjourney Introduction & Gallery Showcase: Now let's go ahead and see what this Midjourney and how we can access it. Now. In general, Midjourney is a text to Image tool that uses AI to create images out of text descriptions and Prompts. Now, by itself, Midjourney is a small independent research institute and a tech startup based in San Fran. It is led by David holds, who was a co-founder rather of a leap motion. Now, currently, Midjourney's only accessible through a Discord bot that you can interact with on their official Discord Server, on some third-party servers or your own private Discord Server if you have one. Yeah, it's really annoying that how they use Discord, but whatever we kind of, I don't know why they're using it. Now to generate an image, you would need to basically Type and trigger a comment and discord, after which Midjourney replies and starts to generate the image. However, currently it requires you to have a valid Midjourney subscription. So to start using it, you would need to have a Discord account which is free. And preferably download the Discord application on your Mac or PC or your phone. You could also use your Discord app directly on your web browser if you wish, and install their native application on an iPad, Android device, or iPhone. But it's also important to know that if you happen to get a free trial for Midjourney, you could only use images generated for non commercial purposes. So basically what that means is if you're going to generate images and then sell them as part of something bigger. Use them commercially. You need to get an official Midjourney license to do that. Which basically means you need to pay for a paid subscription plan. You'll start that crazy. It's very, very affordable. Another important fact is that Midjourney Gallery is public, which allows other users to view or even if potentially use your images as Image References or for remixes. We'll get to that functionality later. But there is a possibility to hide all your generated images by obtaining a Pro plan. And we will also check and compare all the plants that Midjourney currently offers. Now, before we move on to the next slide, I just wanted to show very briefly what is Discord to people who have never heard of it. So this is the interface here of the installed Discord version. Let me just make it a bit bigger. Okay, perfect. You can access it through a web browser. There'll be almost identical to this. And its main purpose is to be a your serve as a chat or communication platform. Now, you can not only chat here, you can also trigger a different comments. Like for example, this imagine comment. So all I would have to do is type in imagine and then insert in this field a prompt. So that's how Discord generally works, and that's how Midjourney generally works. But Discord also has something called voice channels as well. We could activate a webcam and have a conference with multiple people. You could organize the conversation in different channels on your server. And basically there's a lot more to it. Obviously streamers use it all the time. But we will cover the most important parts I've discord relevant to Midjourney in the second chapter. And this one, Let's go to the midjourney website. Perfect. Let me just pull that up here. So the landing page or the first page rather is super basic. You can access the gallery and we're Showcase. Here's the documentation. That's really the Get Started button. To sign up, you would need to press join the beta. If you already have an account you could send, please sign in here. But for now, let's go ahead and select Showcase. Now, before browsing the Gallery, I'll just quickly go through Midjourney's mission and what the goal is of this project. So Midjourney's mission is to expand human imagination by enabling imaginative expression. Now what does that mean? Rather than creating an imaginative machine? The Midjourney team wants to strive to empower individuals to various fields, including artists and professionals, by providing them with the tool and technology that foster creativity Facilitate the concepting and the iteration process and boost confidence in areas where they may not feel proficient. Midjourney inks to be a catalyst for pushing boundaries of human creativity and exploration while also respecting and supporting the existing Art economy. That's what they say at least, but there's a bit, There's some controversy around that, of course. So feel free to browse through their Gallery and you will immediately see the diversity and different, an endless styles and ideas that people come up with. And all you have to do is to prepare a text prompt, which is really cool. And now while we're on the website, I do want to take a second and highlight the importance of this button or the Help page. Overall. Even after the extensive and daily usage of Midjourney, I sometimes still go to this page to remind myself of all the possible comments, parameters. And also check a few important updated pages like the Community Guidelines and see if something has changed, especially when a new version is being released like 5.2 or 5.3. We can just go to a version over here and we can actually see, see that in action. So I really suggest you bookmark this page immediately, because trust me, this will be your best guide. And tutorial. I only suggest to browse for really bookmark this page because you can learn a lot from this course obviously. But I'm telling you, looking at these Community Guidelines here, we'll also give you a good Lake up as well. This course will definitely give you the foundational knowledge and show you the different ways to use Midjourney. But these dots are very, very important as well. So I wanted to make sure that this is referenced because by knowing the official and recommended use of Midjourney will allow you to identify the quality of any tutorials that you have found online. Okay, perfect. Now here we see a short summary and important information on what is Midjourney. So feel free to pause the video if you want, if you want to learn a little bit more. But I will just highlight a couple of extra things to note. Midjourney has quite a notable set of advisors. For example, a gym Keller and Nat Friedman. And a couple of other facts. Basically, currently already millions of people are using Midjourney on a daily basis. And their public Discord has over, I think 16 million users, which makes the server the largest. The whole Discord server directory, which is pretty cool. Another thing to note here is that the underlying dataset that is used by Midjourney for their trading. The AI models consists of a large scrape of the Internet. Basically utilizes open datasets, Public date, public websites, galleries, things like that, that are used to train their models. Now, let's actually go back and check some more examples of the Midjourney Gallery, the public Gallery. Let's go to Explore and then let's check the top. So here we have the top images. This is the most popular or the most rated. Really. It's extremely simple Prompt, but see how powerful it is just bedroom melting into the ocean. And it can create an image like that which is very cool, uses a special model and we'll go through all of these parameters later. But just a simple prompt like that can produce such an interesting image. We can see how Midjourney kinda combines these two things together, which is cool. Let's see another example. Now. This is more of a Fashion look and also pretty photorealistic. Wow, this prompt is way bigger though. We can see that it was generated using the fourth version of Midjourney. Not the latest version, but still extremely accurate. Now, this is the example of illustrations that Midjourney's capable of doing. Also a version for. And this is a very simple and short prompt. This is also another nice example where we can see the blend of Artistic Paint Style, modern day superhero Batman with the past decade really, which is creating a really interesting and surreal effect like this would be something that you would see painted in the 1800s. But it, but you also see Batman at the same time. Out. You'll see a lot of things like this in the gallery as well. Because most people come here and use Midjourney exactly for this, to unleash their imaginative creativity and try to generate something that's well, very creative. You could also see the prompt here, and it's also relatively simple. I would really recommend you guys to browse the Gallery, which has a lot of PFK-1 to do really. I think you might need to register and sign in to see the actual Prompt, but to see all the options here and even be able to save the image, I think you need to register first for Discord account and then a Midjourney count, which we will do in later videos. We'll do exactly those things in chapter two, but the Gallery is a floodway to spur your inspiration. Before we do that, I want to briefly show in the next video some alternatives, some alternative Generative AIs that are very similar to Midjourney. And that can also be used to generate images out of a text prompt. Awesome 5. Other Generative AIs: Let's briefly explore the alternative Generative AIs that I can recommend at the time of this recording, this video. So the first one is the Art breeder, which is quite good at creating different artworks in the painting style. It's not really photorealistic, but still a cool painting style. Okay, So let's open up a few artists and C, in terms of their aesthetics. So in terms of the ascetics at quite close to dolly or the earlier versions of Dolly, which is another Generative AI. And there's also, I believe, some social aspect to this platform as well, where you can currently obtain followers. Right here, yeah. You can follow other people as well. Early Midjourney does allow you to follow, but you never know who actually followed you and this information is headed to you. Feel free to explore the community and Images. Their pricing model is quite simple, so they do have a free account. In case you just want to try it. If you like, you can always upgrade. Another option than I do recommend. It's a pretty popular option is Canva. Now, you can actually generate images directly in Canva out of a tax prompt, which is pretty cool right here. You can see here that it's like a short tutorial or an example, a video of how it's done. So it appears that you would just see to select the style that you wanted to generate in. And just add your text prompt, which is pretty cool. And they also support Painting in different styles, including the 3D style or photo Style. I think there are more styles as well that just don't show you here. The good thing about Canva is you can edit this image directly on the platform. And of course, apart from that, you can create postcards and all different types of media that Canvas supports. So it's quite convenient if you worked with this type of graphic design. Anyways. Let's see here, but also to be fair here that all the alternatives that we have in this list, in my opinion, are currently miles and miles behind. What Midjourney can do in terms of quality and interpretation of complex prompts. But the silver strike, I want this to be a comprehensive course. Next one is lexica. It's called Lighting Lipscomb Art. It's not exactly an image generator, but you can use it to generate artwork using these stable diffusion engine. Now, the main purpose of this project is to be like a Google for the Prompts and all the artworks created with different stable diffusion models. So if you want to get your hands on stable diffusion, you just need to register here. And you can try generating a few images out of your text front. Now, of course, it would not be so accurate at the moment as mid journey because Midjourney is still pretty amazing and probably the best one. But lexica is still pretty capable. So I personally think that the latest best sale of diffusion models are second-best behind Midjourney. So even if you use stable diffusion, Midjourney is a lot better. Usually stable that fusion also requires a bit more expertise. Even custom train models to achieve the same high level of quality that you can achieve in Midjourney. Now let's check the other one which was called crayon. And you can also generate directly here as you can generate images. Here, we just nice. And they have quite an easy and simple pricing model as well. This would be the monthly prices. But you can generate images here for free. But of course, if you generate it for free, they would be public and also watermarked as well. Now you can feel free to try that out if you want and check it out in your next project. The next one we're going to check out is what I personally like it very much, which was called Dolly. And this Generative AI was invented by the same team that developed chat GPT, which really spun this entire AI world, right? In order to start using it, you would need to buy some credits here. But let's scroll a little bit and see. And we can see that Dali is also more Art focused as well. But as well capable of producing the photorealistic images. Of course, with, of course belt with the same type of accuracy. And when it comes to Art works like in Painting Styles and things like that. But it's quite good in some aspects, even better in Midjourney, which we'll kinda go through this course as well. It's also worth trying, in my opinion. So now let's check they'll very last one which is called photo for this project, offers quite a lot of different photo editing tools. Many of them being now backed up by Ai and specifically Generative AI. You can also directly generate images here from a text prompt. And as you can see here, they have some nice aesthetics as well. And the realism is actually pretty good. So both with Dolly and photo, realism kinda gets comparable to Midjourney. To be honest. The feature that I really like here as well is that it's quite good at generating Art out of a photo or a, or an image Reference Midjourney. Although it does that, it's not really suitable for this type of conversion. So if you're looking for something like this, that instead of Midjourney, foe, war is really worth looking into. Now, I have to more professional tools which are only suitable for advanced users. The first one is the actual stable diffusion. It's more of an interface to the engine. You need a client, you need to download it to actually do it. This tool, Automatic Double 11, can actually allow you to generate very precise Content using stability or using stable diffusion as the engine. And even use constantly trained models that were created and trained by other users as well. Now we're going to go into the details here because this does get very advanced. But this project does have substantial community support and very detailed documentation on how to install and use this tool. So let's go in and check the second one on this list here. This tool is more user-friendly, but it's also much more expensive and is paid, right? So stabled effusion, automatic 11 is actually free technically, you just need the compute resources on your computer and needs to be powerful enough to run the engine. Whereas replicate ones than Engine in the background. And you basically have to pay for that. Right? In with replicated, you can actually train your model as well or use pre-trained models directly through the website with a friendly UI. So it allows you to use stable diffusion without having to actually train generated models within. Nice thing is also, I mean, they do charge per second of CPU and GPU times so it can add up. So it's definitely one of the more expensive options that are out there if you're trading a model. So it's also worth trying if you're an advanced user or as you become one, stable diffusion, automatic 11 and replicate our good, good uses there. Okay, So that was the introductory first chapter. Now let's move on to the next chapter where we'll install Discord and start using Midjourney. Awesome 6. What is Discord?: Hello everyone. As we talked about in previous sections, Midjourney is operated through something called Discord. Now, this section in this course is all about how to use Discord if you've never used it before, how to set it up the different settings with Discord, navigating Discord, everything like that. This section goes into a lot of detail for Discord particularly, but not that much detail into Midjourney itself. So if you have any familiarity with Discord, please skip this section because there'll be pretty boring to you, right? If you've already used a little bit of Midjourney that you don't need to go through this section at all. If you've already have experienced with Discord, again, you don't need this section all you can skip directly to the next chapter. I've done a lot of comments about this, that this is to detail this section is too much into Discord. I mean, the reason why I created this section is because I wanted to do, sorry, I wanted it to be a comprehensive course. So that's why I'm giving this warning at the beginning where if you already know Discord or if you've already done a few things in Midjourney, skip this section. It's not going to be useful for you and go to the next chapter where we actually go through Prompt building and Image Generation. Okay, perfect. With that, let's go in and get started. 7. Discord Account Setup: As we already know, we need a discord Account to start using Midjourney. So in this video, we'll go through the sign-up process and will also install Discord application into Windows. Okay, so let's go ahead and get started. And let's go to the Discord website. Now. I'll quickly go through what is Discord and then we can proceed to the sign-up process. Now, Discord is simply a communication platform and application that allows users to chat, voice calls and video call with others. It was originally created for gamers, but since then it has an expanded by various different communities and organizations for communication and collaboration. The Discord app can be installed on desktop computers, Windows and Mac or Linux, as well as on mobile devices, iOS and Android. It provides User-friendly interfaces for interacting with others and offers a wide range of features and Commands to enhance communication and Community Engagement. Some key features of Discord app includes servers and channels. That it basically operates on a server-based Structure. Users can join servers that are like dedicated spaces for specific communities or topics. And within each Server, there are channels for different purposes, such as tax channels for written conversations, and voice channels for real-time voice and video communication. Biosharing discord allows users to share files, images, videos, and other media within servers and direct messages, making it very convenient for collaboration and sharing resources. And also bought integration. Discord supports the integration of bots, which are automated programs that can perform various tasks or provide additional functionality. Now, Midjourney BOD is a great example of how bots can expand the functionality of Discord. Now, let's go ahead and check out some of their servers. You can click on Discover on this page and you will find the most popular channels here. Now, you can also search for a specific topic. For example, you want to search communities that are related to Ai. So you can just type in Ai and then you can just join. You should read the description and really join any community that she liked. So let's go back now to the most popular channels. And what is interesting here is that the top three most popular and biggest communities are Midjourney, open Ai and blue willow. And these are all related to AI technology, which was really cool. The Midjourney currently has over 16 million members and open Ai is the company behind well, chat GPT. Then finally, blue willow is similar to Midjourney. It's a text to image generator where you can interact through the bot. Now, Discord also has a a paid version called nitro. So let's quickly check it out. Now, note, you do not need a paid account to use Discord. But these packs, they offer extras. As you can see here, you will get a higher upload limit of 50 mb. It will be able to use custom emojis. But honestly you don't really need that. Especially if you're a nod, a heavy user of Discord and not using it on a daily basis to collaborate with their colleagues or with your teammates. So we'll just focus on a free account. Now, let's go to the download section and let's go through the installation process first. Now, it may take awhile depending on your internet connection. So let's just wait a little bit. Download is complete. Let's go ahead and launch the application. It will download the updates on initial launch, and then you will be presented with a login or a sign-up screen. Now look, go through the process of registering a new Discord account. You just need to click here register. You just need to provide an e-mail address. For just do that over here. Will need a username and make sure it's unique. Then for the password, obviously make sure it's a strong password. Not easy to guess that save it in a secure place. So let's keep going and we'll set up our date of birth. These don't judge me. Perfect. So I should say here that you should take a look at the Terms of Service before you sign up for everything, You Should Take a look at the Terms of Service. I know most people don't, but just wanted to make sure you guys read this out and do that fully. I've just for liability purposes. So assuming we read all of this information here, Let's go back in. It's also good to check privacy policy as well. Now, this is not required here, so we're just gonna go ahead and click Continue. Obviously, if you select it, you will get optional updates from Discord and some tips and special offers. I always prefer to have an unchecked. Okay. Now let's try to register. Hate these things. I swear, I guess I probably could train an AI to pass those things, even though I know that's not the Use Case here. Okay. And we're in now, sometimes you might have to actually put on your phone number to receive an SMS. They usually do that if you try to sign up more than one account. So just if they do that to you than just you might need your you might need to provide your phone number, but it's safe. Discord is a pretty reputable company. That's it for this video. Let's go through the next video. We'll discuss how Discord works. And we'll go through the interface and important Settings 8. User Settings and Login at Discord: In the following videos, we're going to learn how to use discord in the browser. And we're also going to check their Help page. And I will also show you how to set up your profile and other basic settings inside the Discord application. So let's go ahead and get started. Now, before we proceed to the application, I just wanted to quickly show you how you can use discord within your browser. All you have to do is you have to go to discord.com website. And on the top right corner, click Login and just insert the email address and password you used ones creating the account. So let's click Login and we're inside, which is great. Now, this means you can access Discord or virtually on any device that has a browser in it. However, bear in mind that large servers like Midjourney can slow down your the virus by a big extent. So I do not recommend using the discord inside the browser. So let's go ahead and close the website and the launcher application. Oh wow, it actually launched. Sometimes I'd asked to update, which is very annoying. Whenever it does update, make sure you let it update. I've had since uninstalled and reinstall and many times because I skipped the update. But anyways, so what we see here immediately is two areas highlighted by Discord. It wants you to click here. This is a search bar. Here. You can search for servers, channels, or messages. Here, just click escape to close it. And here you can create your own server if you want as well. We will follow that process later down in the course where you'll be able to create your own private server with the Midjourney bought. And at Midjourney bought to that server and have all the messages saved in privacy directly on the server, which is nice. Now, if you'd like to add a friend, you can click this button here. You probably need to click here first, and then add friend. And then if you know a username of your friend, you can just type here and send a friend request. Now, what do you have all your friends added in? Then you can find them all here. You also need to select friends. If you ever blocks someone by mistake and want to unblock them, you'll be able to find them in this tab. Perfect. Also, don't forget to check the Pending tab because sometimes your friends might not, or it might be sending you a request. And if you miss notification, you can just search for the Pending tab and search for the request here and then either approve it or not. You can also find all your notifications and the inbox button, the and reds. Or if someone ever mentions you on a server, you can find everything here. I believe this is the friend request count. So if you have pending requests, you will, will be able to see this number higher than 0.1. Very important button, year is Help. So never underestimate the official help and documentation because usually simple things can be troubleshoot it by just going through the documentation or finding the related article. Now, one of the most important pages in health Discord is getting started. If you're a beginner, of course. Beginner's guide to Discord here on this page, you can dive really deep into every aspect of Discord. Now, honestly, to use Midjourney, you don't really need to have a scientific knowledge of every aspect of Discord, but I wants to be a comprehensive course. And when we do later concepts with Discord and Midjourney, you need to know this, but for now it's fine. But if you're ever lost, you can always find your answers to specific Discord questions here. Another important issue to know that if for some reason you are unable to fix something, some technical issues with Discord, you can always submit a request rate. Here are Effects. Now that we've gone through the Discord Help page, Let's go back to the application. And here, let's go ahead and let's check Settings. Accessor Settings page. Just find this button, user Settings Click on it. Let me zoom the window a little bit. And the first thing you see here is that it asks you to verify your email. So what I suggest is before we continue, you verify your email and, you know, it's it's, it's something that you need to do before you actually join any Server. So might as well do it right now. Now, if you don't see an email from Discord asking you to verify your email. You can just click. Where's my mouse here? You can click this button again to verify your email and I'll just resend you another email. It might also be in your spam folder or in your trash folder. If you'd also, if you still don't see the email me, you need to make another account or something because it stopped working. Right. And that reminds me to just verify my e-mail Salton back quickly. Perfect. Perfect. Now what else to note here is that you can customize your profile. For example, you can change your avatar image. You can change your display username, which is different from the username that you used to login. Let's click this button and check it out. As you can see, you can set up your display name. So my username is my Midjourney course. So let's imagine that I want people to see me as Mg course. I can save this. Now to change your avatar image, click this button here. And let's select an avatar. Perfect. Let's go ahead and click Save Changes. Now, next thing up is the privacy and safety. Now, I recommend leaving those two filters intact. In short, those just mean that any messages coming from non friends will be filtered for explicit images or potential spam messages. One important setting here as to allow direct messages from Server members. Right here. To be off. That's just my personal recommendation because when we're going to join the Midjourney Server, you will see it has over 16 million people in it. And this setting basically allows anyone who is part of the Midjourney to send you a direct message. So this can mean that you might not receive spam once in awhile, which is nice. But another important setting is to enable message request. Perfect, so evil Mr. Griffin, Server members, you may not know to have this part off as well. Perfect. Now what this means is that people who are not in your friends list will not be able to send to you friends messages. Okay, well then there won't be able to send your friend requests rather. Listen can get handy on, especially on big servers because this can also distract you sometimes, for people will try to advertise their services and try to direct message you. Okay? So if you're concerned about your privacy, which you can do here, let me, let me find it here. 222. You can deselect this text box here, which will basically stop sharing with Discord servers. Telemetry. Basically data that they store anonymously on how to improve their services, is just basically data on how you use Discord. Perfect. Let's just scroll here. And another important feature that you have to remember here is that let me see here. Oh, there we go. Next time. Okay, So the next sorry, sorry about that. Actually, nothing too important for it. Remember there, but the next thing is authorized apps, which we don't really need to look into. The next one here is devices. Here you will find all the sessions and all devices that you ever use to log into Discord. Also, you will see here locations currently there have been redacted. So if you see something suspicious says here or locations that you don't recognize, which you can do here as you can log out of that session by clicking this button, just entering the password from your discord Account. Perfect. We can see that they've logged out of the session. But if you do notice some suspicious activity, I do strongly advise you to change your password immediately for your Discord account and also set up multi-factor authentication or a two-factor authentication. And we will get to that in a second. The next tab is connections. If you ever used another account to login and connect to your discord Account, you will see all of those connections here. Friend requests, Settings is, this is very important. So you can allow or disallow everyone to send you a request. I recommend leaving this all on. This means that people who know your username can find you and add you as a friend. But next tab is nitro. They've ever decide to upgrade your account. Then you will find all the settings here and all of the upgrade options. Also in this section, the next tab is the Server boost. Boost, our like parks. The boosts are mainly used to donate to your favorite Server. So how it generally works is that if you are a neutral member or if you purchase a Server boost separately. Basically that means that if you give this boost to your favorite Server, you can give it to your own server or some other server. This will basically help the server some special functionalities depending on how many boosts were received. So you can explore this page further if you're interested. We can just skip some of these for now. Let me see our billing. That's the tab where you will have all your payment methods, Settings if you ever want to use it for a nitro or a Server boost. One important note is that this is not related to Midjourney payments that is separate. This is the Discord payments that you are that are used purely for your own private Discord account or a server that you have made. Now let's move on and go to the end. Let's keep going. 9. Discord Setup Completion: In this video, we'll complete Discord Settings, just notifications, and we'll tweak the audio and video settings. So let's start from app settings and the Appearance tab. And here you can just change from dark theme to white and vice versa. You can see how the chat would look like. If it would be white, I prefer to work because it's much easier for the eyes. You can also change some other appearance settings here as well, like fonts scaling, or a space between message groups and etcetera, which is nice. Now I'm going to refer to leave it as it is because that's the perfect balance between size and appearance. The next tab here is accessibility. If you ever need to increase contrast or enable some other accessibility settings, you can always find them here. Another one, which is pretty important is the voice and video tab. So if you ever encounter an issue with a microphone or you don't hear a sound, that you can always check your input and output device here. Also, you can check the volumes. I prefer to leave it as a default because then your system, or in my case, it's Windows, will basically just choose your default microphone or OPA device selected at the time. You can also test your video, came here to see if it works. One important thing to remember is that Discord tends to automatically determine your inputs sensitivity for the microphone. As you can see, it lights up green when I talk. So what I prefer is to disable that automatic option and to lower that a little bit and have it around. Well, it really depends on your environment of microphone. You can always go back and check the mic and do a small test. I prefer to have it on manual because sometimes when you speak loud, for example, because you have background noise or anything like that, you speak louder. Or if I for 10 min. But then when the background noise drops and you start speaking lower than what happens is that your voice is actually getting caught. Because the automatic inputs sensitivity doesn't recognize that immediately. So it's really better at Type it manually set up. Perfect. Let's go ahead and check the video Settings. So we can select here at the camera. If you have more than one. Which you can also do is you can blurrier background or choose some other interesting background from here, or even even a custom one. But for that you will need nitro or a paid account. Unfortunately. Another important thing to note is the advanced part of the video and the audio setting. So as you can see here, I've hardware acceleration enabled by default, which is important. Basically sometimes on a little older graphics card. This can cause some issues. So if you ever see that the video is distorted or your webcam doesn't work properly or anything like that, you can, what You could try to do is you can troubleshoot, just disable hardware acceleration. But normally you would not need to do that. But just so you know, that's the first thing to check if you have any issues with whatever we're doing. Another thing which I prefer is to have noise suppression on standard risk is something like an advanced algorithm to suppress the background noise. But sometimes it can be much too aggressive. Well, obviously depends on the environment that you have. You should test both. But honestly, the way I use it, you should make it standard. If your friends or colleagues, every note is that your voice cuts off where they cannot hear you properly. Then you can check here the noise suppression and also the input sensitivity, which I showed you earlier, right here. Perfect, Yep, right here. Also Also make sure that automatic gain control is enabled. This is important to keep your voice sound level balanced, so that's not too loud. Also, not too quiet. And basically the rest you should leave as is. Now only you can refer to those settings if you actually experienced issues that you cannot fix easily, then as you remember, you can go to the Help page of Discord and try to troubleshoot that with the guides that they have there. Okay, let's move on. In the text and Images tab. For our purposes, I recommend leaving everything here, as is. We don't need to change anything. Let's check the notifications. And basically what you can do here, you can disable desktop notifications. I usually prefer to have it enabled, but it's of course how you prefer. Or you can also optionally disabled the sound of those notifications by just clicking here or choosing a specific sound to disable. Which you can also do here is you can enable the text to speech notification that would basically read all the messages arriving to a channel. You could have it read it out loud. If you ever need to do that, you can basically enable it here. One important setting that we should trigger here is the e-mail notifications. If you're not interested in receiving any e-mails from Discord, then you can check this tab, which is in notifications. And that's the lower part of it. I prefer to have those off, but of course, make sure you read whatever you notification means and you can customize it to your preference as well. Let's move on. The next tab here is key binds. Basically these are like hot keys that you can use to speed up typing or some specific function activation. I don't really use it, but if you are a hotkey user, you can always check he bins right here. You can also change the language of discord in the language tab. Now let's check windows settings. Now, this is actually an important tab to check because here, which you can do is you can enable discord to start when you log into your account in Windows. So if you wanted to start automatically, you can basically have this enabled. I always have this enabled, which is nice. Okay, Now let's check the streamer mode. Well, now this is out of scope for this video, but just let you know if you ever to decide to stream a video game or anything else or steroid with your friends. You should check this tab first and customize all the settings there. Now, let's go ahead and check the advanced mode. So one important setting here is the development mode. So sorry, the developer mode. I have this enabled on my account for now. But for now we will leave it disabled. But later on, we'll need to refer to this. Because if we wanted to extract any details from messages or channels that were in. So for the rest, you can just leave it as it is. And if you noticed, there is also a hardware acceleration setting in your Advanced tab that is basically just a duplicate from the voice and video tab. Okay, now let's quickly check the activity privacy. So in this top, what I suggest doing is disabling everything you see here. Because what this means is that Discord will automatically share the notification tab. Well, really in the tagline of your profile, your activities, and not all, but some activities like watching a movie or playing some game. It will detect that and then share that in the tagline in your account. So it'd becomes visible to all your friends and also non friends, but other people that you have Part of your servers that you are in. So yeah, if you don't have the intent of sharing that which really nobody does, to be honest, please check this tab because Discord enables this by default, but everyone always prefers to have it disabled. Perfect. Okay, the next tab is your registered games. Here you will find all the games installed on your computer. If you don't see one, you can just add it here. This is needed to just privacy settings per game. So if you have inactivity, privacy, if you have those switches on, this means that when you actually play a game, it will be displayed in your tagline under your username and discord that you are playing that game. If it's registered here, you'll be able to adjust that per game, that privacy setting. So the game overlay tab here, you just adjust how you want notifications to appear inside the game. The Discord notifications for now, we can just skip that. What's new is the Discord use tab. And Hype Squad is a special Discord community where people basically, well market Discord. It's kind of like an affiliate, but not exactly. So for marketing or hosting some events on your server related to Discord itself, you receive a special badge and some other perks. Okay? And then finally you have the logout button. Okay? So I think we've done everything we need to do to successfully set up discord. We've gone through all the settings, the app settings, the profile settings. Just one thing I did promise to show you is how to set up two-factor authentication for your account. And you can do that right here. You should always enable two-factor authentication. It's very important because it help you secure your account, because you have a Midjourney subscription on your discord Account now, which means that it's pretty valuable, right? Obviously, what this means is that if you ever decide to, to login to Discord from another advice, Discord will send a special light confirmation e-mail. And instead, which you can just use as a two-factor authentication, then it will just prompt you to enter in a two-factor authentication for your account. Instead of confirming your Login through e-mail, which was a bit annoying. So it's just basically an important security step. So I encouraged that you enable it. For now. It's just the good thing to do. Perfect, awesome. Let's move on to the next video. 10. Midjourney and Server Setting Creation: In this video, we will sign up for a Midjourney account using our Discord account. And then we will navigate and explored Discord and Midjourney servers. We will learn how to navigate their and other basics of the Discord user interface. So let's go ahead and launch a browser. And then we'll have to Midjourney's website, which is obviously midjourney.com. Now, search for a sign-up button over here. It's called join the beta. So let's click it. And then it will be up directed to our Discord application. As you can see here, will be offered to join a Midjourney Server. In case you use discord inside the browser, then you will stay on the same page. But then you will click. You'll basically be asked to login to your Discord account and you'll be presented with the same window there. Let's go ahead and click Join. Will then need to confirm that word, not a robot. But who knows these days with a privilege generation of AI. So let's go ahead and try that. I hate these things so much. Now. It didn't convince that wasn't able to convince her that I was human. Let's try it again. Okay, and we're in. Now let's go through the user interface of Discord. The first thing we'll see here on the left, so the top-left button with the Discord is your list of direct messages. Now currently it's empty because we have a new account later on. If you will have any direct communication with someone or with your friends, you can find them all here. The next one goes this Server, and you will have all your servers here. You will be able to group them. Also, if there's, if there's too many. With this button over here, you can actually create your own private Server. That's a free function. And we'll actually need that later. With this button, you can actually search for communities the same way we did on the Discord website. Okay? So if you click the server name on the top, you will be presented with a few options. Server boost is the option that we've seen on a user Settings pages as well. So if you've purchased boosts, you can donate them to a server. Basically showing your support and helping the Server to have extra abilities like full HD video, sorry, full HD video. Hire file uploads available to the server. As you can see, Midjourney has over 983 boosts. Just shows you that community of Midjourney is quite large and also very supportive. Okay, let's check out the next option. Invite people so you can have your friends and Discord. You can actually invite them to the server by just clicking here and copying the link. Even if they are not part of Discord, they will be able to sign up and then automatically join the Server with this link app directory. And we can skip that for now, but we will not need it for interacting with Midjourney anyways, there's linked roles. Here. You'll be able to see all the roles that are assigned to your account on this server. I've hasn't. Club basically means that the user has generated 1,000 or less images with the Midjourney bot. So when you generate, I think there's a 2000 Club, there's a 5,000 Club, there's a 10,000 club. You'll be added to the respective roles, which will open up special channels on the server. Perfect. Now this option show all channels, mean that basically ticket, it will unveil the channels that are hidden by default. Or if you've hidden some of them manually before. I just wanted to take that for now, notification settings. So this is the most important part when you join a new server. So let's click here and then you set up how you want to be notified about the events in this particular server. So let's quickly go through this with a solid means. The next option is to show is to basically hide Muda channels. I usually have it selected because sometimes you want to mute some specific channels We will get to that on how you do it. But just for now, remember that if you have this option selected and if you happen to mute some specific channel, maybe because it was too annoying or too many notifications, etcetera. What will happen is that it will actually just disappear from this list, which is sometimes convenient. With this option, you can adjust how people see you in Midjourney. Your nickname basically. Our basic nickname is, well, MJ course that we've selected during registration. But let's change it to MJ superuser because that's how people will see us in the Midjourney Server. Perfect. Now it will not change our username. Obviously that will stay the same. But the display name will be changed for this server only. Another important page here is privacy settings of your Midjourney Server or any other. So make sure you have direct messages switched off and activities, basically everything switched off. So what direct messages means is that you have this enabled. It means that anyone on the server, even if they're not, your friends, will be able to send you messages directly to you, not through a channel, but, but directly to you. The activity status. This means that it's the activity feature that we've checked on user Settings. So if you have a game running, or sometimes a video player running, and if you don't want to share that activity, and this to be visible to users on the server, then there should be disabled. Now, before we dive deep into all the channels and how to use them inside Midjourney Server in Discord. Let me quickly show you a few things here. On the bottom part, you can see your account. If you click what options you have here, you can set up your public status. So either you want everyone to see you as an online, or you can set up, for example, invisible and then you will appear to everyone like being offline. You can also set a custom status as well. So you can either add an emoji and then a short text. Perfect, Let's just try that. And let's add a smile to this. Then we can also set up here when we want her status to be automatically be cleared. So let's say today, our status says on line. Let's save it. And that's how people will see us in our public Rooms and also your friends as well. Now, one thing I do suggest though, is that as soon as you register and sign up to a new server, switch your microphone off. This is just precaution that by mistake you were not added. Or if you joined by mistake to some channel where there's a voice communication allowed, then you will not clicking this, you will not accidentally expose everything what happens in your room, which is good. So this is just a safety precaution. And then obviously as soon as you're ready to talk, you can just unmute it, press it again. And then you can talk. By defaults, I prefer to have this muted. Okay, with this button, you can go to that user Settings page we should have gone through in previous videos. Perfect. So another important thing I wanted to show here is the Server settings on this page. On this page, notification settings for Midjourney Server, which you can do here is adjust notifications for this particular server. Any Settings here will not affect any other settings you might be part of or general notification settings of discord. So let me quickly remind you where that without as being set. So if you want to change general notifications, you have to go to your user Settings and then Notifications tab here. And then you'll be able to set up all the notification settings that will affect every single Server. Let's quickly go back to Midjourney notification settings. So by default, you will receive notifications for mentions. Basically, I mentioned is when someone adds to their text message and at, and then tags or username or a role that you are a part of. For example, let's check our roles here on the server. So we are part of this role. So if somebody mentions that, then we will also receive notification. Let's go back to your notification settings. If you want to mute the Server completely, just click here and then enable everything you see here. So basically this will make sure that you will not receive any kind of notifications anywhere, even on your mobile device. Let's turn it back on. However, what I do suggest doing is suppressed everyone in here and mentions because that is, this is often used by other users or by announcements that you may not want to receive. And you have to look for 34 days joining the server. If you receive too many notifications from everyone or mentions, then you can just mute that. I prefer to always have been muted by default. But then at the same time it's a trade-off. You might also miss some important announcements from the server. Okay, perfect. Then we also allow other, any other mentions which are usually, for example, a role or a username. When somebody mentioned to you directly, if you want to disable More fall notifications, if you happen to install Discord on your mobile to than just click that here. And let's go ahead and save that 11. Going Over the Midjourney Server Part 11: In this video, we continue to explore the Discord interface. And we also dive deeper into a Midjourney Server structure, it's channels and how to use them. Let's start by launching a Discord. And the very first thing you see here in the Channels List is a channel called home. This. This is an auto-generated channel. It's actually still in beta. This is a new Beta feature of Discord. What it does is it basically selects, let's say, the most popular message or messages that are currently gaining in popularity. Which you can do here. If you don't like some particular message or some Style that you don't want to see in your highlights. You can press not interested button. Then select what you want to hide. You want to hide this specific channel or a user, or just the specific post. Let's say I don't want to see anything from this user, and that's basically it. Let's go to the next one. The next is events in which you can do here is you can check what events are upcoming. If you see here, there are two events scheduled. One starts in 49 min, another one tomorrow, 09:00 P.M. local time. And what this means is that if you press the interested button, you'll be notified as soon as this event starts. And you'll be able to join into this room, which in this called is called stage one. Now, apparently, it's not accessible. But as soon as the event starts, you will get a notification and you will be able to join. Let's go in and close that. The next channel that we want to check, it's called channels enrolls. Here. If you click browse channels, you can actually go through the full list of all available channels in the Server. By the way, not all servers. Half of this, but Midjourney does. What you can do here is you can read the description of the channel and get an idea of each channel. You can even view that as well. Let's try to track that so you can see the content of the channel. And if you like what you see here, you can enable it. And then it will show up on the list. Perfect, for now, let's just leave it as it is. And let's continue on to the next one. As you can see here, those channels are grouped in a folder called info. So what You can do also is you can have dedicated Settings per channel or per the whole folder. So for example, if you don't want to be notified from any of these channels that you see in this folder. You can just right-click it on the folder and then you can mute that folder as well. Which basically means that all the channels within that folder will be muted automatically. Or you can just simply select specific notification settings for that instead. Now let's check the announcements. Now this is quite an important channel for Midjourney. As you can see, David H, who is the founder of Midjourney. He is quite active in the community and he often posts different news and announcements of Midjourney and the progress that they're doing. So every once in awhile, I suggest to check this channel and also see if there's any important announcements that you might want to track. One important part that I want to mention is that you can override notification settings per channel. So let's say that before we have selected to suppress, although notifications hear from everyone in these roles. Let's say we still want to receive all the announcements from this channel. Well, what we can do here is we can right-click the channel. We go here. And this basically means that it's a standard setting that you have per server. But we can override all of this by saying that I want to receive notifications or all messages. Just one quick thing, don't forget to check is that you don't have this channel unmuted. So Now, even if the role is mentioned, you will still receive notifications from here. That's how you override the notification settings that you have set per server. The next channel is called the recent changes. It's similar to announcements, but it's more about technical aspects of the Midjourney bought. Then we have a status channel. This is also quite an important channel. Basically, every time Midjourney's experiencing some issues or downtimes. And this can happen because it's project is still currently in beta technically. Every time that happens, a short update is posted here. So if you're experiencing issues with the Midjourney bought, what I suggest you do is to check the status channel because there will likely be post there. Now if you don't see any proceed announcements and there's a major outage or something like that. Which you can do is you can go to the actual channel where you can channel with support. And we'll get to that and then you can ask for help there. So those three worthy announcement channels, it's a different channel type. As you can see, you're not allowed to type any messages here. The next one is being rules. In this channel. You can, You cannot also Type any messages. And this is basically just the general rules. Nice strongly suggest that you check this before you start using any Server because you don't want to be banned, that will be really bad. The next two channels, they are standard texts channels. But those two, as you can see here, we also can't write anything in here. So getting started is just a quick start guide. A very small one. But as soon as you join a new server, they usually have this channel here. I also strongly suggest that you check it out so you can get an idea of how to start using this Server as well. And then we have the welcome channel. It has some important information. We will not go through that in detail. But I suggest also, as soon as you joined the Server to go through it, read and this might be helpful. It might help, help you to get started as well. Obviously, we will go through this in much more detail, but it's a good welcome page. Let's go into the next one. As you can see that this channel is located under the support folder. So if we click here, we can see that we are also not allowed to write messages because this channel is actually locked. So if we scroll up a little bit, we can see that people were still chatting not long ago. And this is related to that. Midjourney recently introduced a new support for billing issues. And basically all those billing issues are being handled through email. So you don't need to go through trial support channel here. So if you've ever happened to experience an issue with payment that I strongly suggest to use this e-mail and communicate through the email. Because people just tend to go to this channel and use it, which is why they blocked it, right? So it's better to just e-mail them if you are having any billing issues. Okay. Perfect. So those were the first few channels that it's worth going through. Awesome. 12. Going Over the Midjourney Server Part 2: Now, in this video, we'll continue to explore Midjourney's Server Rooms. And we'll also try to broadcast stream and see how that works. Now what you can see here are the newbies Rooms. Basically, those are mostly used by new comers who just started with Midjourney. And basically you can already start Prompting right here, right in this room. After we will complete our Setup of Midjourney Account, there will be more Rooms revealed where we can actually prompt. So as you can see, there's a button called jumped to present. Let's click that. And those will scroll us all the way down. And let's try to prompt for something. And let's see what happens. And we're going to be a little bit meta here because I like to have some FUN. So we're going to type in slash, imagine. Okay, so let's go to type that in. And then I'll ask us for a particular prompt. Let's say right here we want to Prompt Midjourney itself. Okay? So we're going to type in Midjourney. Perfect. Let's try that. And then we're presented with some Terms and Conditions. So let's go ahead and accept that and see what happens. And we probably need to try the prompt again because this isn't fresh new Server or fresh new account, whether. Now let's try Prompt again. Mid journey. As you can see here, we've got a reply that says we need a subscription. Basically. Unfortunately, Midjourney doesn't offer a free trial. So that means in order for us to use Midjourney, we need to have an active and paid subscription. And we will get that shortly. For now. We will do is let's continue browsing through all the other channels. And then we can continue to a Midjourney website where we will obtain a subscription and start using it. I want to make sure we go through all the Server channels. Is I want this to be a comprehensive thing on Discord. As you can see, there's more newbies channels in here. All of them serve the same purpose. Those are basically used to prompt your images. You can use any of those if you want as well. Of course. We'll actually up to a subscription to make our own. But for now, let's move on. There is a general discussion channel. And we'll basically obtain access as soon as we add Discord bot into our Discord account. And then we have a very important Shot which is called Prompt chat. In here, you can actually get help with your prompts, which is really cool. So if you want to create a very specific image, but you are unable to reach your desired results. You can head to this channel. There are usually people here that can help you with your prompt, which is really nice. Now, usually you can distinguish a genuine Midjourney moderator if they have this little green icon there. You can also, you can always check that if you want. This doesn't mean that the general public cannot help you. But at least you can distinguish between the official moderators and not. The next useful channel is called Prompt FAQs. Let's browse through it and it's basically a knowledge base for different commands and parameters. You can filter this or you can even browse the whole tutorial if you want. For example, let's check this one for creating stylized selfies. And year. It explained step-by-step on how to achieve from this photo, this kind of artwork, which is very cool. Now you can also search. And I also suggest before asking anyone in the community for help, just check, simply just check this FAQ because it will usually Visually all the most common questions we're already answered here. And there are located right here as well. Okay, in the Showcase folder, you will find different channels where people share their best works. And you will be able to do that to a daily theme is a channel where every day there's a certain team announced and people try to basically match it as close as they can and produce the best artworks that they can for that team. So this was a basic Midjourney Server Overview. Basically in the following chapters will learn how to benefit from those channels and how to use them properly. But before we continue on to the next video, I just wanted to quickly jump into this broadcast and to show you how this works. So before we joined, I defind all the sounds so I can explain what's happening without being interrupted. And then we'll listen a little about what actually happens. So let's just click here. And as you can see here, you can raise a hand. Certainly it's disabled in this particular broadcast. But basically, you can raise a hand. Moderators will see that. And then they can invite you to the speakers. Then you can ask her questions and basically interact with all the speakers here. Okay, let me indefinite and let's listen a little bit. Interesting to share this bit, bit, bit. So you know, the hair pattern that you have on your head, the swirl, they call it the hair whorl. There's another name for it called triglyph. Midjourney thinks they're hedgehogs if you use that word. So that was just a quick intro of how a broadcast would sound like. Now, don't forget that if you want to enter into the speakers by raising your hand over here, you can definitely do that. If you do that, but don't forget to unmute it as well. If you see a full red microphone, this means that it's muted from the server side because we are now located in the audience. But as soon as we enter into the speakers, make sure that microphone is unmuted. Okay. After you finish listening, you can either press this button, disconnect quietly, or you can also check this section here. It always shows you if you're connected to any voice or video channel, you will always see this here. You can just simply press this and you are disconnected 13. Subscription of Midjourney and Comparison vs Fast & Relax Modes: Now let's go in and complete our Setup. Let's move with midjourney.com and select sign-in on the bottom right. And you will be presented with this window here, which basically asks us for authorization of a Midjourney pot. Now, before authorizing Midjourney BOD or a or any other brought, you should carefully review the access rights that you are giving. Because for example, there'll be able to access your username, avatar. It also have access to your email address as well. So you're giving permission to access to join servers on your behalf. And you're also giving it the ability to read your member information like nickname, roles and etc. and also to update your connection and metadata for this application. So just be a bit careful here. So let's go ahead and click Authorize. And now we're assigned to our own private Midjourney account. Okay. Before we continue to explore our account, Let's purchase a plan. You can either do that by clicking here or going to the managed sub, which is short for a subscription. Now as we see here, we have three plans available to us, the basic, standard and Pro plan, with Basic being the smallest one. It only gives you around 200 images a month, which might sound like a lot, but trust me, this will get used up very quickly. The standard plan gives you 15 h of Fast generations. And the Pro plan gives you double the amount. To understand better what this all means. Let's head to the Help and FAQ page and explore about how this is all being calculated. So let's go and select user guide or rather subscriptions. Then go to Fast and Relax Modes. Now, as you can see here, Midjourney uses powerful graphics processing units or GPUs for short, to interpret and process each image. When you purchase a subscription to Midjourney, you're basically purchasing time to access those GPUs. This basically means that every time you trigger an imagined comment or other command that produces an image, a large cluster of GPUs as working in the background to produce that image, taking power and time. And this is where the calculation of allocated minutes and hours comes from. Now, Midjourney provides to work in Modes, fast mode and relaxed mode. The standard and propylene subscribers have access to relax mode. While the basic plan can only access the fast mode, which consumes the allocated time, but it produces images much faster. Now, let's go ahead and check the plans. As you can see here. Standard plan and the Pro plan can have unlimited Relax generations. While the basic plan only has limited generations of 200 a month, which is the fast mode. So if you're subscription plan has access to relax mode. This means you can virtually create an unlimited amount of images when you are in a relaxed mode without consuming the allocated post hours. However, jobs are relaxed mode are queued based on the current system usage. During high demand, a single job may take more than 10 min to complete. One thing to note here is that your prioritization in the relaxed cube depends on your own usage. So if you use Relax Modes frequently, it will likely, it's going to be likely that your job is, we'll take even longer to complete. However, this priority, this priority resets every month when you're subscription renews. Now let's check what actually affects the amount of Fast hours consumed. As you can see, reading a variety, sorry, I Variation out of already image takes the least amount of GPU time while upscaling a GPU or an image rather takes more than average. Other factors that affect GPU time consumptions. You can see here. One of it is Aspect Ratio. So if we stay with the default square aspect ratio, this will just consume the average time. But if we set some tall or a wide aspect ratio, this will consume much more GPU time. Also, some models, like test models, they consume more. We will not be really using that for production We will definitely try those out. Another factor that's important here is the quality parameter. By default it's set to one. I usually prefer to use it on quality two, which is the maximum quality. However, that obviously costs more time. And another parameter, which is the stop parameter, we will also not be really using that, but this basically directs Midjourney to stop at a certain level of completeness. Let's say if we have a parameter of stop 50, that when Midjourney completes an image on 50%, it will just return the result without completing it fully. Now let me quickly show you how that would look. So let's go into the Help page in a user guide in Parameters. And let's find the stop parameter. And as you can see, perimeter of 50 would look something like this. Whereas if the image will be completed fully, it will produce a completed image like this. And that's how it looks like on perimeter attend, which doesn't look like much. Now, let's go back to our account and subscribe to one of our plans. For the purpose of this course, I will personally subscribe to a standard plan. So I will have access to 15 Fast hours and then limited Relax generations. Before clicking subscribe. Just to which a building type that you would like to subscribe to? Would you like to would you like to pay monthly, which has a slightly higher cost? Or would you just like to purchase a yearly plan? If you're just starting off with Midjourney that I of course, advise you to purchase the basic or standard plan on a monthly basis. Also, bear in mind that these prices don't include any taxes or value-added taxes. And depending on your location, you might be subject to that. So let's continue. I click subscribe. Also, depending on your country, you might have different payment methods, like with a credit card or some alternatives. So we'll just go through this form, filling all the details and purchase the standard plan. Now, after you I filled all the details, just scroll down. If you're a purchasing this on behalf of business, then you should probably take this box. And then with the additional details, you can also provide a promotion code. If you have one. When you're ready, just click subscribe. And you should be set up with a confirmation page which basically confirms that your payment was successful and you're subscription has now been activated, which is very exciting. Now, just close this window and here you can find details about your subscription. Now as you can see, we have about 15 h remaining in our billing period and the date where we have the next payment. Now one thing to note here is that a new button appeared by more Fast hours. Now, this is available to any paid customer. So as soon as you run out of hours, if it happens to that there are 0 h left in this month, you can always click here and purchase additional hours. Just don't forget that those are Fast hours. The relaxed mode hours are unlimited and available to standard and pro users. So if you have relaxed mode, just to use that, I think it's much simpler. Now, let's go back here. Now, if it happens that you encounter an error during payment that you can add resolve yourself. Or if you have another Billing inquiry, which you need to do is send a support request to this email, which is right here, billing add midjourney.com. Now, if you have another question not related to billing, then you'll be able to ask those questions directly in Discord on the Midjourney Server in the channel called member support. So now that we have a subscription plan, Let's go back to our Account page and see what's changed. As you may have noticed, there are two new sections in your account page called rank pairs and Settings. Rank pairs. This is where you can earn bonus fast GPU hours. All you have to do is choose between two images that you like more easy enough. And this helps to train upcoming new models of maturity. It's generally recommended to choose an image that is more aesthetically pleasing than the more photorealistic or based on any other criteria. However, It's entirely up to you. So you can also shape the direction in which Midjourney development is heading. Below, you can find the full prompt that was used to generate the image which is the same in both windows. In the Settings menu, you can adjust how the explore page looks like. Just play around with different settings and see what works best for you. Personally, I prefer to leave it at the default setting. 14. Going Over New Rooms, First Prompt, Variation and Upscale: Now in this video, we'll check what new channels we have access to with our subscription. Will try your first prompt and will learn how to rotate the generated Options, grade variations, and Upscale them. So let's begin with launching a Discord. And let's go to Midjourney Server. And the first thing we want to do is go to this top channels and roles and into browse channels. Now, let's see what we have here. I can already see a new channel here called micro polls. I would disabled the welcome channel. Getting started is up to you if you want to leave it on, on member support, that's a very important channel to have in your list. If you have any issues with your subscription or your account in general. Not know what prompts. So if you need help with Prompts, there's a special channel here are called Prompt chat, where you can ask for assistance with your prompts. But if you have issues with your subscription, then basically you can ask those questions here. Now, what I would also do is I would disable all the newbies channels because they're usually getting spammed quite heavily with Prompts. There's a lot of people Prompting there. So discussion is up to you. If you want to engage the community, make sure that you have Prompt chat enabled for the feedback. If you ever want to submit a bug report or anything like that, you can enable that channel. Prompt FAQs. Make sure you have that. Showcase. It's up to you. I usually like to browse through the latest version, Showcase to see what people are coming up with. But usually they don't submit the Prompts with them in this channel. So it's just basically to get inspired what the latest versions can do. I would disable the paint overs. Of course, that's totally up to you, but this is just what I recommend. The bunch Showcase will leave the daily theme. I usually don't participate in that, so use that off. So those channels can be useful if you want to get inspiration, see what people are prompting for the right now. However, I do suggest to disable all of those general channels. Maybe just leave one because they actually produce quite heavily load for your computer. Because there's so many messages floating through all of those background messages from being digested by the CPU. So basically it's recommended that you maybe only have one or two channels enabled. I'm telling you it will be, they'll be a lot faster. I would even disable everything else. Because if you just want to get inspiration, you can just use the Gallery on a website. Okay? Also you will see all the prompts there. We will observe Gallery in more detail in upcoming videos. For now, let's just disable those channels. Also in here. Those are all related to the voice communication. There are a couple of texts channels. So what I suggest is to leave VC text. Because if you communicate in one of the voice channels and you want to drop a link, that is a channel to use to do that. And also, I will suggest leaving the stage one. Because sometimes the Midjourney creates a broadcast event where you can join and listen to import an updates and announcements. You can also requested join the conversation and ask your questions as well, which is pretty cool. Now, let's just leave one voice channel active and the rest we can disable it. Alright? One more thing we can do is to collapse the category, for example, the information channels. And in this way, we actually created a very clean interface where we only have the important channels in our Overview. One last thing that we can do here is collapsed the group of channels. And now as you can see, we have only the important channels in the list. And now it's time to try our First Prompt. Let's go ahead and find the channel where we can do that. So we can either try them in the newbies channel, but those are user prompting or let's just temporary enabled one of the general channels. Okay, perfect. Let's go to general 16 and do it there. Perfect. And let's go ahead and try our very first prompt. So let's scroll all the way down. And as you can see other a lot of messages appearing here all the time. So it's not going to be so easy to track or messages here. However, that's just for the purpose of this video. Just to try our very first prompt in here in the general channel. So in order for us to trigger an image generated by Midjourney will need to send a command called forward slash. Imagine. Let's try that. And then provider or prompt in this gray box and hit Enter. Now as you can see here, a public channels are very crowded. So will lead to scroll up to find our prompt. We could also see the progress. So this is our message here. It is highlighted with a yellow background, so you can actually distinguish which is nice. And then one thing to note is that this message will be visible by others. So everything you've seen this channel is Public, apart from example, when you trigger an info comment or some specific comment that has some private information about your account. Those comments, we'll have a blue background and those will be private and only visible to you. I'll show you how it looks like in a minute. Okay? So as we see, our message has gone from here. It means that Midjourney has actually completed the generation. So let's go ahead and find that message. Alright, so this is the result that we've thought. As you can see, all four cats are doing some kind of dance move either Midjourney sees dance like that. And we also see that all four of them will look basically like photorealistic images. Now, this will not always be the case because we haven't provided any details or any styles. So Midjourney just selected that randomly. Most of the images that are generated by Midjourney are usually photorealistic, especially if we're talking about version five or 5.1. Now, let's see what else we can do with this image. So first of all, what you notice here is that we have basically four you buttons and for B buttons. So U1, U2 stands for the number, it says for the image. So that's first, second, third, fourth. And you stands for Upscale. So let's say if we want to Upscale this image, the first one, let's, we need to press this button here. Or if we want to, for example, we liked this image and we want to see other variants that he can provide the similar image. And it's just going to be slightly different. We can press the V2 button or if we don't like any of those, but we'd like the Style. So for example, we've managed to hit the style with this prompt that we're looking for. But we wanted to see alternate versions. Then just click this, which is called the Re-rolling button. So let's try that. And you will also see this button changed to blue. Now, sometimes might not be the case. If it's not done, just try to rest again. Basically try to re-trigger the Prompts. Sometimes it doesn't work for whatever reason. Okay, let's go ahead and find our new version. It should be highlighted with a yellow background so you should be able to follow it. Hello, right? So this is another four options that we've got. As you can see here, we actually received a completely different look. So all four cats are not photorealistic now, this more it looks like a painting style and looks like a mix of photo realism and Cartoon. So basically this just says to us that if we don't provide or do not specify specific style or a theme, that mid journey will just selected randomly, right? So that's basically what's thoughts about. Now, let's try to Upscale one of these images. So let's say we like this version specifically. Well, actually before upscaling, let's see how their version look. As so we like that one, but we want to see some variations. Sometimes you can see like artifacts or missing parts of an object, for example. So you can just try to make a version which is also called a virile. We've selected the fourth image here. So we should see variations of that fourth image. Let's scroll all the way down. And there it is. Okay, I think it's finished. Now, View roles usually take the least amount of GPU hours, just so that, you know, upscaling is the most heavy, but not for the version five and the future releases. This upscaling will take GPU time, and certainly it doesn't. Okay, wait, we've found the wrong image. Scroll down. Here it is. Now as you can see, all the images here are the same style. So we have everywhere that cat we're doing some move. All the cats up blue hats. So most of the important things are the same, but the postures are a little bit different. We can see differences in the background. So basically that's how B-roll looks. Let's say that we like the cat the most. And now let's try to Upscale that. So that'll be u2 or the second version. Now let's scroll all the way down again. And as you can see, Upscale worked almost instantly because there are no real generations or a GPU processing happening in the background. That image is just being cut from one of those four Versions. You can also fabricate that. That would also reflect in your Midjourney Gallery and your profile. You would find all the fabric images. You can also make variations which is similar to virile, where you can open that image in Midjourney Gallery. So basically on a web, or you can simply save the image somewhere as well if you want it on your computer. In the upcoming video, we'll create a private Discord server, which will help us avoid all of these random distractions that we see here. And it'll make it a lot easier for us to use. Awesome 15. Private Discord Server Setup: In this video, we will set up our own Discord server and we'll add the Midjourney bought to it. And we'll create multiple rooms to help us sorting with our prompts and images. Now, let's go ahead and start by launching Discord. And now let's find this button here with a plus that says the server. So let's click on it. What you see here, those are just templates. We don't need any template, so we just click this button. We can also skip this question two. Now, just type any name that you like. And let's go in and select an avatar. And let's create the server. And that's it. Now, we have our very own private Server and we're the only user in that Server currently. Now, let's go to the Midjourney Server. And let's find the Midjourney bot, which should just be under Midjourney BOC group. Right-click. And we're going to select Add to Server. So basically these are just permissions to your own account that you are giving. And here we choose a server that we just created. Let's click Continue. We actually see permissions that we give on our own server. So not to our account, but permissions for the midtone bought on the server. You're going to review it and then authorize. And let's confirm that we're not a robot. That was quick. But now let's go back and we see our robot Midjourney here. Now, let's try it and let's see if it works. So let's trigger a simple info command. Perfect. Okay, This means that it works. Okay, perfect. Now let's create a few texts channels and then see how we can use those to sort our prompts and images. To create a channel, you can just press here the plus and the category. So these are just two categories, tax channels and voice channels. You can also create it from here as well if you want. But before we do that, let's actually create a category and we'll call that prompt. We can call it something else like images or remedial, but that's fine. And then inside there we will create a new channel and let's call that photorealistic. Okay, perfect. And here it will basically try prompts that can produce photorealism. And then let's just prepare another one that we will call vector, which stands for basically vector Style Images and Prompts. And then we'll create one more year, which we will call a test or draft. There we go. Here you can play around with different new prompts and different techniques without spoiling the other channels, which is nice. Let's see what options that we have on our server. So two important tabs here. One is invitation tap, invite people, and the other one is Settings. So invite people. If you click here, you can just basically copy this link and send it to your friends. And then even if they don't have a Discord account, they will be able to join to this link by creating an account and then joining the server automatically. And you can also set an expiration as well, which is nice. For example, I can set it to expire in seven days. You can also limit how many people can actually use this link, which is a good security feature as well. Okay. Now let's go to Server Settings. So this is the basic server information. The name, the logo. We don't need to change anything here, but will we do need to do, unfortunately is configure roles. Now this is an important security staff because sometimes invite lengths are being reshared by friends. And it could be that someone you don't know can actually join the server. So in order to add a two-step verification, basically for anyone who is joining, we need to clear all the permissions from the role. Everyone. As you can see here, they have supervisions here. Let's just click this clear permissions Perfect. What this does basically is that when someone joins the server by default, because this role is assigned by default to everyone who's joining, they will not be able to see any channels or use them. In order for them to actually start using the Server. A moderator or an admin, or a, or a server host will need to assign that role manually. So let's create a roll and call it. Oh, sorry, wrong type display. Let's call this role guest. We can select the color as well. Let's save changes. Now what we can do now is we can set permissions for guest role. And what you need to enable it is to View so they can view all the channels. Otherwise know without ovule to join the channels. You don't need to enable anything related to management or audit log. Also do consider if you want to allow them to create their own invites or allow other people to the server. You're going to allow them to change the nickname, but not to manage the nickname of others. And in this tab, texts channel permissions. Let's just allow everything apart for management roles, grit. So let's get those. But we do need to allow to read the message history, especially use the Application comments. Otherwise, they will not be able to interact with the Midjourney about. Now, those are provisions for the voice channels. Let's allow them to connect, speak. Use videos and activities for this. For this, you're gonna be what this means. I usually prefer to disallow to be a priority speaker and basically that's it. So we've just set permissions for the guest role. So when a user joins, which you can do, let's just go back here and I'll show as soon as your friends joined the Server, which you can do is you can go into Server Settings, go to Roles, find the guest role, and then in managed members, you can assign roles here. Another way to do that is actually from the user management type in members. You can also assign roles here. And what You can do also, you can actually assign multiple roles. And this way you can actually create like multiples that you've actually assigned multiple roles here, which is nice. One role, for example, it can allow users to view texture animals interact with them. Another role and allow for example, users to speak. Or you can also allow to view different channels. I will show you how to do that in a minute. Now let's see Other important Settings. So in apps and integration tabs, you can actually see all the bots and Integrations that you have on the server. You can also observe what permissions the Midjourney robot has. On the bottom. You can see what's allowed and what's disallowed. And you can see the list of comments that this BAD adds to your server. Doc comments rather Commands. You can also see all the important activities that happen on the server. For example, when the user joins the Server or when somebody takes a user or other important actions, you can see them in the audit log. This is also a good idea to check once in awhile to see what usually happens on your server. The last tab is the invites tab that I want to show. So here you can actually disable or delete invites if you want users to stop joining with a specific link. So just disable it and that's it. Okay, perfect. Now, one last option I wanted to show is that if you want to hide some category or channels from users, you can edit the category or a specific channel and Information tab. You can make it private. Which means that the roles and users would have to be added manually. So you can have the old role if you want. For example, if you want this category to be visible by guests, but maybe a special role, or you can add individual people right here as well. Just don't forget to save the changes. And basically that's it. Yeah, Discord is a pretty wonderful tool. 16. Prompts Help and Exploring Midjourney Gallery: This is the final video of this chapter. In the next chapter, we actually begin. Our Prompting practice will learn how to create good prompts, and we'll focus on creating photorealistic images. So in this chapter, we have already learned how to create the Discord account and use applications. We've created our Midjourney account and subscribe to a plan. We've gone through the basics of Discord, user interface and navigation. And we've also explored Midjourney Server and useful channels. And we've even tried our very first Prompt and prepared a private Discord Server. Now, before we continue on to the next chapter, I do want to show you how the Midjourney Gallery works. And we've tried a couple of handy channels on a discord server. So let's start as we always do and launch at Discord. And let's go to the Midjourney Server. And we need to find a channel here called Prompt chat. Now, if for some reason you don't see this channel, you can go to this tab over here. And then basically just browse channels. And then you can just search for it, which is sometimes easier. It's called Prompt chat. And make sure that this tick box is enabled. Let's go find this channel here, prompt chat. And basically this is a very important Shot for beginners. If you need help to achieve a desired result, to create a specific object, for example, to mix two themes in one image that you're going to ask for help here. It basically asks experienced users and usually they're very responsive and very helpful. Now let's close Discord and let's go to the Midjourney website and see how the Gallery works. You can sign in right here. Now let's go to the Explorer page. It will we see here, it's basically all the images that are generated by the community. The public images in which you need to note is that all generations, regardless of where the images were generated in private, public servers or the Midjourney Server. It doesn't matter. They will all appear on your profile right here, which is accessible to everyone. And they will also appear in the community feed. They can also be searched. So in case you want to hide those images from the public, then you would need to obtain a Pro plan, which does cause $60 if paid monthly. Because unfortunately that's the only package that has this stealth Image Generation option. When you obtain that, you would also need to generate a special to trigger a special comment with Midjourney called Style. Now let's go back to the explore page and let's see how we can actually benefit from this community Gallery. One of the most useful things that you can actually see, the prompt that was used to generate that image, which is very cool. So if you open up, you would see that full Prompt here. You would also see all the parameters that were used. If you want to try and reproduce that image, you can go here into the Options and you can copy the full comment, which is nice. So let's go back and discord. Let's go back into our Server in the test room and let's actually try the prompt. So let's type in imagine prompt and then just paste it. I loved the private surveys just so much easier to use. So now we can see how this image is being generated, which is cool. And you can crack the progress there. It's almost done. So this is the result that we've got, which is very similar to what we saw here. Of course, it's not going to be the same. It will never be the same because the state probalistic model, while you look at the same style of the same object. So this is just a quick example. How would we can get inspired or basically try to reproduce the images that we see here. Now, let's see what else we can do with an image. Let's take this for example. Well, first of all, when you favorite images. So let's favorite a few images here. This way you can actually create your own collection of favorites. It might not update immediately. Let's just refresh that. Okay, well eventually your favorites should all appear in this tab, which is nice. Another thing you can do is you can follow people. If you'd like what they're creating. This way, you can actually create like a personalized field. And let's see how that works. So let's follow this user. For example. Let's, Although this user here. Now, if we navigate to top follows, we'll see here is that all the images that were prompted by the people we just followed. If we go to the top, we can actually see the best images from those people that we follow. The best in terms of highest rated. Another cool feature is that you can actually select multiple images like this with this option here. And you can basically either added to your own collection or you can download them. But okay. Let's figure this out. And those are really the last topics I wanted to share with you in this chapter again, which was a very, very well comprehensive course on Discord. But now let's move on to the next chapter. 17. Introduction to Image Generation and Prompts: Hello everyone and welcome to the third chapter where we start our imaging practice for sizes. Now, this will be the largest chapter of this course, as it serves as a first practical chapter really. Let's now proceed to review the outline and the upcoming topics in this chapter. So in this chapter, we're going to learn about what exactly is a Prompt and the different application techniques as well. This is really important. We're also going to learn how to do a prompt. And the best says around with prompting. And we will learn how Prompts should be structured by taking a few examples and then examining the actual results. We're going to take a closer look at each part of the Prompt where we have a main subject or a theme. Then we will start including the surroundings, stylizing the result with a theme or certain keywords. Then we're going to try and apply certain well-known aesthetics of different famous artists or photographers who work with images, paintings, or videos. Then we're going to learn how to use Weights and Mulltiprompt all these parameters which will be very exciting. And then to Stylize with certain Artistic Styles or by combination of some styles and descriptive keywords. We're also going to explore how certain keywords in the Prompt effect, styling, colouring, texture, and other aspects, will learn how to improve per Prompts, incense, start enhancing them with teams to get a certain aesthetics for a scene. We're going to explore how certain keywords in the prompt, vector styling, color, mean, textures, and other aspects of the results. We're going to try and apply certain themes to enhance our results and apply well in create a Visually Appealing scenes. We'll also learn what are seeds and how to use them. We'll also learn how to try to replicate the same object or character in different scenes, which honestly is probably one of my favorite things to do. We're then going to practice and dive deeper into the advanced Prompts structure, where we will Weights and combined. Prompts. Also along the course will be Re-rolling or Re-rolling. The results. Learn how to get the best out of our prompts. Will explore Other Versions of Midjourney, like version four and Nietzsche. We'll also go through all the functionality, Commands, and parameters. So you've fully understand the actual functionality of Midjourney's and how the different ways you can use them to manipulate the prompt to achieve your desired results. Okay, perfect. Now, let's go ahead and get started. 18. Prompt Basic Composition and Structure: This video, we will start our practice by exploring the basic structure of a Prompt. We will then explore the impact of describing the environment on the final image. Now, let's go ahead and open up our browser. And we're not actually going to head to Discord Russia, going ahead to the documentation for Midjourney, let's open up getting started. Next steps and Prompts. Okay, Perfect. Let's scroll down a little bit. And this is a really good visualization for you to know here. You can see here, prompt in Midjourney consists of three parts. The first one being the image, basically a link to an image. The middle part being the actual texts prompt. The third part are parameters. So parameters, how an image generates, they can change things like the Aspect Ratios, models, up scalars and many other things. And basically, parameter always needs to go at the end for the Prompt where the image part, the link, don't actually need to include that the image prompt is optional. And those links to images, they serve as a reference image. But we'll get to the Part later on in the course. Now let's scroll down and we'll check one important part here. So we've checked how technically the prompt looks. But also a good prompt should always have a description about the subject, Medium, environment, lighting, color, mood, and composition. The first three parts, I would personally consider the most important. So of course, the subject, what you want to see in the image. Then in which Style is it a photo, a painting, or illustration? And then finally, the environment. Others are important as well. However, I do need to point out that these are the minimum of what you need to specify for a good prompt. By specifying everything, you will actually get the result that you were looking for. If you do not specify something, for example, you don't specify a mood or a Composition. Then basically Midjourney will most likely pick a random option for you. Now let's go to the next page and explore Prompting. Let's see how by specifying a specific medium or a Style, we can actually get pretty different results. So as you can see from this example here, where we're creating in some particular style, a cat. And we want to change this, for example, for our block or print or graffiti. That would then become a graffiti style cats. And then you would get that desired results or the needed results. So let's explore Other Versions. For example, the watercolor cat would produce an image of this style. Or a pixel Art Style cat would produce something like that. We're actually going to try this in Discord as well, but it's good to get a basis like this first. For now, let's just observe what Midjourney's capable of. Let's scroll to this part that time travel. Those are very basic examples. However, it indicates if we specify cat from '80s, it would actually understand that and try to produce an image that resembles that decade. If we scroll lower, we can also see that Midjourney can understand emotions. Now, of course, we're talking here about a very basic prompt when we actually start to make long Prompts with more than 50 or 60 words, sometimes even longer than we would need to use special techniques to actually tell a Midjourney to focus on specific emotions. If that word would just be somewhere in the middle of a long Prompt, most likely Midjourney would just skip that. Okay, now let's open up Discord. Let's go to our server. And let's go ahead and start Prompting. So let's start in a test channel first, and then we will move on to a photorealistic channel. So let's start with a very basic Prompt. Let's go ahead and prompt a cat. So you type comment or forward slash typing imagine. And then you want to type in your prompt. And we're going to write in here, a cat in a forest. Okay, Midjourney has finished generating our image. Let's examine the results. As we can see here, we've got one result and Photorealistic Style, and three more in a painting style results. Now, let's enhance our Prompt and add a median. Again, if you don't specify one Midjourney will just select it for you randomly, which is not always will be want. Now we want a photorealistic cat as a medium. So let's select that and see what happens. We're going to again type in, imagine. We're only going to type in a photo of a cat in a forest. So now we actually have a medium as well that we're providing. Now, let's examine the results. As you can see. Now, since we've specified photos, literally the word photos. Now, all four images are rendered in a photorealistic Style because we have supplied the type of the medium. Now let's keep enhancing our prompt. We went from a cat in the forest to photo of a cat in a forest. Now let's do, let's be a bit more descriptive. Okay. We want to now see a ginger cat in a vast forests. And let's see how that will affect the actual image. Okay, photo of a ginger cat in a vast forest. So we've added it into words, ginger and vast. And from our first iteration we've added photo as well. Alright, now we see a photo of a ginger cat in a vast forest, which is perfect. And we definitely see a more deeper forest than what we saw before. Which is great. Okay, Now let's go ahead and enhance this Prompt even more, okay? And actually add more details about the environment. Okay? So we'll agree to type in photo of a ginger cat in a vast snowy forest during winter. That's my bad at it. Let's try that. Okay. This is a result that we've got. Now. We have a ginger cat in a snowy forest during winter. Basically, Midjourney understands that very well. But when we actually describe the environment to more and more. So let's get back to where very First Prompt here. So this is where we started a cat in the forest. Then we've specified the Medium. So we described that we want a cat, a photo of a cat, a photo, that's the medium that we described, which type of cat that we want. And finally, we added more details about the environment and that's the final results. Sexually Upscale one of these images and see a little bit more details. Let's pick the fourth one. The Upscale should be very quick. It should be in an instant. Opening. Perfect, that looks really good. Let's examine it a little bit. As you can see here, Midjourney understood every word of what we asked. So we said we wanted a photo. So basically this is a very photorealistic Image. Then we started, we want a photo of a ginger cat. Then we explained that we wanted to be in a vast, I think the deepness of this forest in the background is actually the word vast here. And then we wanted a snowy forest during winter. So again, subject, Medium and environment, we clearly have specified it. That doesn't vary basic prompt. However, this tells us that Midjourney, especially in the current version, understands the language really well. And how was not really the case before? If you've Comparison three inversion for it was definitely struggling some time. So it's definitely gotten a lot better, especially if you put it a good Prompts structure. The next chapter, Let's go through good and bad examples and see why one Prompt works and another Prompt does not. And he to create a good phrase and reposition words if you don't get the desired results 19. Prompt Composition Enhancement: In this video, we'll continue to enhance our Prompt with ginger cat and we'll re-examine the effects of additional words in the result. Let's go ahead and get started. Let's launch Discord and start prompting in our new photorealistic channel that we created earlier, right here. Now, let's go ahead and play a little bit with word positioning. So now we put a forest in front of a cat. Let me go ahead and type the Prompting to show you multiple. So now the cat is behind the forest. Basically, let's check our previous prompt. So previously we were including a cat in a forest, right? That was the subject and environment. So the forest was going after the cat. Now it's the other way around. And also we've added in a new object, a rustic wooden cabin. Let me just add the cabin. Perfect. Now let's submit the prompt and will examine the results if the new wording that we've added in has any effect on the result. And also if we can actually fit two objects in a single image, which is both the ginger cat and the rustic wooden cabin. And let's see how that will play out. Alright, so we've got our results. Let's check them out. And perfect, as you can see here, Midjourney perfectly understood our Prompt and it rendered everything what we asked for. So we have a cat, we have a wooden cabin, and we have a winter forest. Now, let's try the same Prompt. But what we will do now is we're actually going to change the wording a little bit. And we're going to put a cat in front of a forest. And we'll see if this has any effect on the final image. So really the words are the same. We're just rearranging the order of the words. Let's try that now. But you may have also noticed that I added a comma after the first part of a sentence. And we'll get to punctuations a little bit later. But basically this also has an effect on the final image. Alright, It's almost done. I can already see a very big difference between the two Prompts. Let's open this image. We can see here that the cat takes a larger part of the image and it's clearly the main object of the image. So here you can actually see how Wording or the position of the words rather affect the final image. You may have noticed that the rustic wooden cabin itself was not the main object. It stayed at the back of the prompt. However, in the second prompt, we moved the cat into the beginning of the Prompt. And this directly affects the image result. The conclusion, everything that you are closer to the beginning is prioritized than something that's at the end of the prompt. So positioning and the Prompt does matter. Now let's try a totally different Composition. So we have a photograph of a majestic ginger cat gracefully navigating through a pristine snowy forest landscape. And let's see what kind of results that we have here. Almost finished. Now let's examine this. Now, the biggest difference that I notice here is that we can actually observe at an, a movement. So if we check the previous two images, we can see that the cat is just sitting in the snow. Right? And let's check the prompt. But here. Yeah, so Over here we can see you that it's class gracefully navigating. So it's actually doing something. So the key thing here is when you add action into the Composition, Midjourney basically tries to animate or like picture how the animation would look. So the cat navigating or the animal navigating Midjourney really understands this, which is pretty much what I asked for, right? So in your composition, you wanted to have animation or your subject animating as well. Now this is just a cat what, and we could have put in something like navigating, running, or hunting. Now another difference that we notice here, if I compare it to the previous yeah, perfect it down. It's a very slight difference. But what I can notice here is that the cat here has, has more for, and I believed that has something to do with the word majestic and the context of a cat. So most probably Midjourney you References the word majestic in this context with a wildcat or maybe even a lion. And another feature that this Prompt had is the word gracefully. So gracefully navigating, not just navigating, which I believe also adds a little bit of beauty to the movement. And I believe they'll whole image itself as well. So Midjourney, by default, usually produces quite aesthetically pleasing images. But anyways, it doesn't harm to add an adjectives or features into a prompt like the word amazing, graceful, aesthetic. Depending on the mood of course and the senior want to create. But you always have to be careful and not to overdo the Prompt or overstuffed. There's a common term called overstaffing the Prompt with these words, because otherwise Midjourney might get lost and will not understand what you actually want. So again, the composition matters, the ordering of the word matters, and also animating the image also matters as well. Now let's try one more prompt. We will add a cat in action and try to put my cat into a specific environment. So that would be our next Prompts. So as you can see here, we've chosen a cat playing, well, sorry, who was playing with Snowflake Surrounding by towering pine trees in a snowy mountain forest. So let's go ahead and check the results. And as you can see here, there's not much action going on here. So we did say cat playing in the forest. But it looks like here that the cat is just staring at snowflakes. And also we don't really see any references to a mountain. Always just see here is a forest. We did get the ginger cat and we did get a snowy forest and a pine forest. But there's no relation here to Mountain or playful unfortunately. So basically in the next video, we're going to learn how to actually achieve those two points. How to achieve a specific environment and action. When you don't want to Midjourney doesn't do it for you. Okay. So let us for this video, Let's go 20. Prompt Composition Emphasis and Wording: Hi everyone. In this video we'll actually learn how to achieve the desired action in a C and how to place an object this specific environment through Emphasis and Wording. Okay, let's launch Discord. If you remember from the previous video, this was our last unit where we have prompted for a photo of a ginger cat playing with snowflakes and surrounded by towering pine trees in a snowy mountain forest. But when we opened up the image, will we saw that the cat doesn't actually play with the snowflakes, unfortunately. And there is no reference to any mountains anywhere. Alright, so let's try this prompt. Now. Again, you can find all of these problems in the PDF file in this, in this course module. Now, this prompt as a photo on a slope of a rocky mountain during winter with the ginger cat in a playful action. Playing with large snowflakes and forest, with towering trees are Patriots and the background. Now, as you can see here, we've actually changed the prompt quite a lot. But we're basically keeping the same C, but the structure is completely different. Now let's wait for the result and then examine it. Let's open it up. And now what we can immediately see here is that first of all, is that the cat is now in action, which is great. The cat is not just looking at the snow, right? The cat looks like he's actually playing with the snow. And that's exactly what we've asked for here. Let's actually Upscale a couple of the images and look at it a little bit deeper. I would prefer it to Upscale these two there. The second and the herd one. Perfect. And let's see here. Let's actually go ahead and Upscale another one that we had where the Prompt didn't work or it didn't provide the playing action. So let's just Upscale the second one here. Now let's compare. So in previous images, we actually prompted for a cat playing with snowflakes. So ACAT an actual surrounded by pine trees in a mountain forests. But we don't see any action here. We take this image, we do see an action. In the second option, we also see an action. So the cat does look like he's playing. Now, let's actually make for More Actions. And we'll see that in all of those four actions that there'll be playing. Now the first, second, third, fourth image there. Yeah, let's actually, let's actually re-evaluate them again and see. The nice thing is you, we've also got a slope here as well, which references the Prompting. We've got a winter forest and we've got pine trees. So basically we've got everything we've prompted for. And similar to hear, the cat is clearly on a slope. And as you can see that there are trees that are actually going on an angle as well. And those are pine trees. It's winter. That's a snow-covered mountain, which is great. Now see, we've generated for more options. We've got some interesting here. Actually. We've actually got a human, which is probably, it's just part of the noise, I guess, because it probably associated playing with playing with a human. And things like this do happen all the time. But notice how all four of them are still playing. They have actions that are playing, right? Whereas before the cat wasn't doing anything that was just staring at snowflakes. You do expect some variation like the humans sometimes being there. Which is why when you do variations, you always get something slightly different. Alright, perfect. Now let's go back. And we can also see in this option, snowflakes, they are also, let's say, more visible. Now let's examine those to Prompts I gave us those two different images and find out what is the difference. Okay. Let me actually copy that Perfect. Now let's examine it a little bit. So let's go ahead and take the First Prompt. And with the First Look, there's actually nothing wrong with it. So if you would tell the sentence to a person, I believe a person what probably imagine something like this, right? The, the mountain, the forest, the cat playing, etcetera. Now, this is why we need to understand that there is a difference to how Midjourney interprets the prompt. Okay? So what happens here? The way I see this as Midjourney is just like a small child who is only learning and trying to understand the world around him. So if you would spell this sentence to an adult, I'm sure than adult would paint a correct picture in his mind. And it would closely match to what we imagine. However, to a child, you will probably need to spend a little bit more effort to explain the same situation with more words and more detail. More detail is the key to really Prompting correctly with the Midjourney. Okay? Now, in fact, what happens is that Midjourney doesn't know which details are more important to you. Is the count more important? Is the mountain more important or the pine trees and more important, replaying more important. Basically the First Prompt just got merged, let's say in the Ai imagination, into a playful cat in a forest, right? Whereas in the second Prompt, since there's more detail, we actually changed the word laying into a playful action. And plague basically means that we actually took rewards that to say the object is performing some sort of action. So the first word, playful, just tells what Haydn of action. Let's say the character or the type of the action than the word action itself, which is very helpful to Midjourney. Action is very helpful to Midjourney to determine a specific action or create or inaction or this combination, always use the word action in Prompts. And then we have a third verb here which is playing. So what we're really doing here is exaggerating less. Okay? Now, you wouldn't speak like this to an adult roommate. You wouldn't tell an adult a playful plot, playfully playing, or anything like that because they would be like You're crazy, I understand that he's playing, right? But when you have more objects in the prompt, because we have a word, for example, surrounding tower pine trees. Those are all separate like tiny prompts that are getting merged inside the artificial neural network of Midjourney. It's really a good idea actually to start from simple prompts like the bad prompt we have here. So maybe bought parameters, swapping out the track name. It's probably something else like basic prompt or something like that or base prompt. And then if you see that, like here, like here, you do have almost everything. You do have almost everything with the base Prompt. But the thing you don't see here, then you can add more keywords that explain a little bit better that particular action or object that is missing from the image. We saw we were missing playing. So we added playful playing, playing action playing, and then reach that, right? So this was an example of how to achieve the results that you expect. If some parts of the image are missing, all you need to do is you need to enrich and retro those missing parts. In this case it was playing. And that's really for this video. In the next video, we'll actually learn if Punctuation dashes, things like that, actually mean something. And we'll learn also what noise means as well. See you in the next video. 21. Noise and Punctuation Part 1: Hi everyone. In this video, we will explain how basic punctuation affects the results. We will go through a couple of Punctuation examples that can actually change the scene. And additionally, we will learn what noise means in the context of a Prompt. Ok, so let's go ahead and launch Discord and go to our ultra amazing Server here. Now let's actually switch back to a test channel for this. And let's try similar looking Prompts with only minor differences. And we'll see what comma does here. The first would be Elvis Presley's pinged Palmer road cafe. Okay, so let's trigger that. Now. Actually before we examine the results, let's trigger to more Prompts. Again, you can find these Prompts, by the way, in the PDF file at the beginning of this module. Now let's try a third version. So as you can see here, I've removed all the Punctuation. There's no commas here. So now, as we can see here, Elvis Presley's is like a plural. So we could pick it up probably expect, maybe multiple levels. Presley's breathing logically, but let's see how it turns out. Now, let's see how this affects the image. Okay, perfect. We have other images. Let's start from the first one. So we can see here Elvis Presley everywhere on the four options. And Will. We can also see here is the PIM resembles Elvis Presley's decade and the style overall right? Now, let's go ahead and check the second one, the second prompt, where we move to the comma. Now, as you can see here, that there is a difference. It is light, but it is different. You can see the same theme here. But we don't see Elvis Presley on every image. And if we compare it to the first image, it actually has more Elvis Presley Style than the second one. But just to be sure it because there was lots of noise, because we don't specify here. Let's just roll it and see what it gives us. Just to make sure that change that actually is doing something and not just producing noise. It is not just changing it because of noise. With noise, yeah, I mean randomness. If we remember from our previous videos, if you don't specify additional details in your Prompts, those will be picked up randomly by Midjourney. Perfect. Perfect. We have the results. Now, let's examine those two. And let's start with the first version. So we started with a comma right after the word themed, which makes Elvis Presley themed Roll Cafe. And will be noticed here is a very similar style to the first version. Let's go ahead and compare that. Perfect. Now let's check the version where the comma is after the word road, which makes this Elvis Presley's steam road. Then we have a cafe as like a separation. Let's check it. Well, what I can definitely see here that the road is not an extra object in here. So a human would understand this. They would understand the sentence that it's obviously Presley Tina road, and then a cafe is like another object. However, Midjourney did print actually a little bit different results. That's because of the noise, will again get to that in a second. But overall, we're getting cafes in both situations. Both cafes are in Elvis Presley's Theme. Just with one slight difference here. It here it also has a cafe Overview like from the outside. And here we only have indoor shots. So let's check those two options. Actually here we also have a snapshot technically from the outside. But basically there's not much difference between this prompt And this Prompt. Well, let's go ahead and check the third one where we don't have any punctuation at all. And just for the clarity of this experiment, let's actually do a Re-rolling of that one as well. Let's see Where is it? Let's see, Let's see the alternative version of the ones with no punctuation. Let's check it out. With this Wording. Elvis Presley's should actually logically expect multiple Elvis Presley's, right. But in this option, in this version. And also where it was it here, we're getting one-person or no person at all. But what we do get here, we do get any Elvis Presley themed and, and also a cafe. What this tells us is that in reality, that all the punctuations that we're used to, that is grammatically correct way of saying these things doesn't mean anything to Midjourney apart from an extra seed to the noise. Now, I will trigger one last prompt for this video. And then we will check what noise means. Let's just add a question mark. Actually, I will make this prompt and we will immediately trigger another one. And let's just add a random string to the end of it, and it doesn't make any sense. But it's just a random combination of letters and numbers. Sometimes you to get a security trigger. So if there is something funny in your prompts or the BOD detects that it could be against Terms and Conditions. For example, forbidden words and things. You could get this warning. Let's try to appeal. Amazing, okay? Or appeal has been accepted. If you ever see that warning, don't worry your account is in pain or anything. Just try to appeal if you didn't try to do anything funny. And usually it's accepted. Those are usually those prompts are usually tract. As you can see, it started generating, it continued the Generation which has good. So let's examine this prompt, the result of this prompt. So we have a question mark. And this affected the results by a little bit. We still have a Elvis Presley themes cafes. We just have in one, but it could just be a random selection or noise. But basically just changing a sentence, meaning with Punctuation. That would mean a different scene or topic to a human. But the key thing is I'd mid journey does not care about that. It does not care about that. He doesn't care about Punctuation. And let's check this one here, but the random letters and numbers. So we're getting a little bit, Let's actually do a Re-rolling this one just to make sure if this is different or not. This looks to me like a 3D game, like a shot from a video game, which we actually didn't have technically in the previous results. What I can see for sure is that the random phrase contributed to the overall noise, right? So basically just produce something random noise is basically your randomness. So this randomness in every Prompt, I'm from two sources, okay? The first one being the seed. In the future video, we'll learn how to actually use this parameter so we can specify the initial noise. And the second source of noise is, well, your actual prompt itself. So basically, everything that Midjourney cannot explain or interpret. This contributes to the overall noise or randomness. Okay? Now feel free to experiment with random noise. This can be things like those random letters we put in. This can be things like punctuations or random combinations of other characters. And as you will Re-rolling, Re-rolling, you find that it becomes more noisy and more random. Really, the main thing here is at Punctuation doesn't affect Midjourney as much as you think traditionally. Perfect 22. Noise and Punctuation Part 2: Alright, in this video, we'll continue to explore Punctuation. And additionally, we will briefly look at how to use Mulltiprompt using a double colon, which is a functional Punctuation inside of prompt, which is different than a comma. We'll also share a small Punctuation tip that is not documented in the official Help page. Okay. Now let's go ahead and launch Discord. Let's go to the Midjourney Server and to Prompt FAQs. Now, let's go ahead and find the Punctuation page. We need an article called Punctuation. Perfect, Here it is. Let's brief. You'll learn what types of Punctuation Midjourney actually understands, not Palmer's. You'll need Punctuation Midjourney understandings are a double hyphens, double columns, curly braces, and spaces. So space is the most obvious one. If you want to separate words. And for Midjourney words are basically like tokens that are then used to look for correlation with the words in the underlying dataset. Now, another one, which is the dipole hyphen, which is used to specify parameter. There is established set of parameters. And also each version of Midjourney actually has slightly different set of parameters that you can use. Then it comes the double column, which basically D limits and basically is used for things like Weights and also Mulltiprompt. But while this later, There's one more curly braces. So those are quite advanced Techniques and we'll definitely come back to that. But later in this video, what I do want to try your with all of you beautiful people is to show Mulltiprompt doing and using the Punctuation double colon to do that and how that affects the resulting image. So this would be a very simple Overview of Mulltiprompt. We will of course, dive deeper into this topic. Because using Mulltiprompt, we can actually achieve pretty consistent results. And tell Midjourney what to prioritize. Okay, perfect. Let's go ahead and try to simple prompts. And this was actually in the intro video of this course. So one would be hot dog. Let me try to imagine. So hot space dog, one of my favorite, favorite things to get from Costco. And then the second Prompt, Let's divide hot and dog with a double colon. Just like that. And let's see what type of difference we actually get here. Now one thing I do want to show here that usually don't get very often, but sometimes this happens. As you can see, for a brief moment, it was paused. So when you see this, instead of a percentage or you're waiting to start, our pause. Pause basically means is that it's an acute and there is some maybe technical issues that are happening. This can also happen with servers are overcrowded, so you have to wait for your turn in line. Basically, nothing to worry about. Just a small hint. Okay. Perfect. Let me wait until this continues. Now, if sometimes it is paused and nothing happens for a minute or two, then what you can do is you can just retrieve it, the prompt as well. Now, of course, if you need it fast, if you're on a relaxed mode, then things are absolutely normal. That things do take awhile because relaxed mode, it takes a lot more time to complete the image. Alright, perfect. Cus non-intensive civic Book mixed up. But anyways, here we see a hotdog and that's how we would really magic, right? So now let's see how this double column actually affected the image. We can clearly see here that the word dog is now a separate. So it's now not related too hot in terms of the logical understanding of it. Okay? So in this case, the word hot without giving any other further contexts, interprets basically as Midjourney thinks and feels But this might also be random noise as well. But the thing Here's a Midjourney. Does it connect the word hot and dog together? One important thing to hear is that by using a double column, we've actually prompted, we've actually triggered to Prompts. We triggered a hot and then we triggered a dog. And basically we've asked Midjourney to combine the two and use that to merge. Again, don't use commas. You can't do with hot dog, you to do hot double colon dog. It's very useful to merge a specific object or seeing with a theme. For example, if you have a very complex theme with lots of detail in it, than for example, if at the end of the Prompt you just have details, then those details might actually get lost by Midjourney because they don't have a specific priority. Remember, we said that Midjourney prioritizes things first. If you separate that Theme with a double column, that would be like a separate prompt. Let's try that right now to see the effects of it. Let's take one of the previous examples. And instead of something using commas or putting that the very end, instead, we will use a double colon and see how it prioritizes things in the second prompt after the double colon. Okay, So we'll take this prompt and let's extract the Rocky Mountain during winter prompt. And we'll add it as a separate, like a second Prompt. And we'll use it to describe a theme for the image. And let's see how that works out. So I'm just going to remove that. Okay, Perfect. I've just added a double colons there. So we basically moved it an added a double colon after the double colon is specify a clear theme. Technically, it would be more correct to spell Rocky Mountain Theme or which are Theme separately. We can also add more than one double colon as well. You can have three prompts for Midjourney, for example, to merge together. So let's see how that will look like. Okay? Now perfect, Let's check it out. And you can see here what's different than the ones before is that all these image, we definitely see mountains very clearly in the background. Here we see a mountain. Here. This is a very large slope. And here we also can see it clearly. I mean, we can see that this is not really technically a cat, more of a wireless cat. But what this did to a prompt is it basically told Midjourney that we want this all as one Prompt. And we definitely want to see this part and we asked it to merge the two together. Now let's compare to the Prompt without the double colons. Here. Only maybe in this image you can see the little bit of mountains, but let's check the other options. The other articles don't really have mountains. We just, we just have a little mountain over there, but otherwise we don't really see the mountain image because we didn't add a double colon. We didn't add a second Prompt. Whereas here we can clearly see that. Now before we complete this part, I do want to share one simple hint on how to create a strong relations between two words. So let's go to the test channel. Now let's try the simple prompt here, which is a wide angle mountain photo. And basically it's a very simple example of how you can increase correlation between two words. So the word wide an angle can exist separately. But if you wanted Midjourney, that these two should be understood or spelled really like a single word, like wide-angle, which refers to a wide angle lens or a wide angle camera picture. Then it's a good idea to add Daesh or hyphens between these. That's only just one dash or hyphen knock and multiple don't do multiple because multiple dashes means something else. Because multiple dashes interpret that as a perimeter. Now, let's wait for the image to form and then check it out a little bit. Yes. As you can see, the first and the third image, especially those are clear examples of a wide angle shot. Alright. So now let's just a little hint quick and to increase correlation. But again, the meeting This video is how to use the double colons because that's really what you should use instead of a polymer. And to put more emphasis on things later in the Prompt. Awesome 23. Basic Parameters: Aspect Ratio: In this video, we will explore important and useful parameters and examine how they affect the results. Let's go ahead and start by launching Discord. And now let's go to our Server. Now let's create a separate channel in the Prompt category, and let's call it parameters. Now the reason why I like to do this, and I suggest you do this because this will serve us as a reminder in case we forget the effect of specific parameters, we can always go to the messages in this channel and then see the results of those parameters. Now let's navigate to Midjourney Help page. So let's start the browser. And let's go ahead and go to the User Guide. Now, under this section, user guide, we find a chapter called parameters. And then here we can actually check the parameter list. Okay, Perfect. Now, as you already know, Parameters always go at the end of the Prompts. And they always start with double hyphens. You can also include multiple parameters. So let's scroll down and check the first one, which is Aspect ratio. As the name suggests, this perimeter at modifies the Aspect Ratio of an image. And you may have noticed that Parameters have short naming. So basically Aspect that only be spelled with two letters AR if you want. This is obviously short for Aspect. Okay, let's go to the Parameters channel and let's trigger a simple image generation. And we will include some specific aspect ratios. Now you can use any prompt for this test. My Prompt but I will be using is a photo of a yellow armchair standing near a table in a living room. So again, Subject action started subject Medium action and environment. Let's generate this without any parameters. And now let's generate the same thing with an Aspect Ratio. And let's try three column two. So that would be an Aspect ratio of three times two. So the first number three represents the width of the photo, whereas the second number stands for the height of the photo. Alright, now let's go ahead and examine the image. So that is the first one, as you can see here, without specifying the Aspect ratio, we're getting the square image by default or one-by-one. In the second image where we specify the Aspect Ratio, all options are wider. So we got four options and every image is wide. Now let's Upscale to see it a bit closer. So here you can clearly see the effect of Aspect Ratio parameter on the image. So this perimeter is very useful when you need a specific aspect ratio. For example, if you're generating an image as a background or a wallpaper or your desktop computer, for example, or even a smartphone. And then for a smartphone, you would indicate a tall Aspect Ratio. So not something like this, that three by two, but probably two-by-three or something like that instead. So let's go ahead and try to create an image that would suit mobile devices, for example, as a background. Let's try a specific ratio for that. Okay, let's type in our Prompt. And less use Aspect Ratio four by seven. So this Aspect Ratio will create for us an image which is tall. And while it's creating. Let's actually go to the Midjourney Help page and let's go to Aspect Ratios. Perfect. So you might be familiar with Aspect Ratios 16 by nine, which is usually the standard aspect ratios for most full HD or four K screens, like monitors and TVs. And now typically this is also common for mobile screens, but the other way around. So nine by 16 because it's portrait. However, Midjourney states that aspect ratios will be slightly altered. So when we specify, for example, Aspect Ratio of 16 by nine, which equals this ratio, it will be changed to seven by four, which has a slightly different ratio. If you see hanger. And our Aspect ratio of four to seven is just a reverse of this one here, 16 by nine. So let's look at the results. Perfect. Alright, let's Upscale one of them to take a closer look. Alright, and what we see here is a tall image. And if we were to rotate that to Landscape, that would be typical or near to 16 by nine aspect ratio. Okay? So that's how Aspect Ratios effect. Now, before we go to the next parameter, let's check out a few important notes here that you need to know about Aspect ratio. As you can see here, not every version supports any Aspect Ratio. So version four is limited to one or two or vice-versa. And as far as I know, version five has basically unlimited aspect ratios. So you can specify 100 to one, for example. And that will produce a very long Image. We'll talk about Versions later. But don't forget that a non-standard Aspect Ratio is consumed much, much more GPU power. So especially if you try some extreme Aspect Ratios, this could exhaust your fast GPU hours very fast. Or if you're trying relaxed mode, that could be a very slow generation. Perfect. One more thing that I want to note here is that Aspect Ratios only accept whole numbers. So you can specify, for example, one.39, 21. Instead you should specify 139 to 100. Okay? So in case you want to get some inspiration or try some common aspect ratios, you can go to this page here and just browse to the bottom. And here are the popular options for you. Okay, perfect. Now let's go back to our perimeter list and we'll see you in the next video. We'll talk about more Parameters. Awesome 24. Basic Parameters: Chaos and Repeat: Let's continue to explore basic Parameters and we'll try the fact of the Chaos parameter and the Repeat parameter. Let's go ahead and start Discord. Then let's go to the Midjourney Help page. So let's go to Parameters, Parameters list. So in the previous videos, we've checked that the effect of the Aspect Ratio parameter. Now let's go to the Chaos perimeter. And let's go ahead and check the theory quickly in the dedicated page. Now, the first thing to note is that Chaos can be simplified as the only the C letter. So if you don't want Chaos and the defaults Chaos is zero. Now let's scroll and see how it affects the image. So when we don't specify a Chaos or we specify a zero value, we're getting more consistent results. That's really what Chaos determines between the four images. There are more consistent. In this case, we have a watermelon owl hybrid. And we can see that most images don't deviate from each other. But when we increase the Chaos value, let's check, for example, the moderate option. We already see here that with Chaos values of 25, also we have possible values of zero to 100. So this is just a quarter of the way up. We can already see much more diverse Options. Let's compare, for example, the first image. So that is the Chaos value of 25. And they all look very different, different styles. It's all everywhere. We do see both an owl and watermelon, but it looks like there emerged differently. One case, the L is on top of the watermelon. The other one, it's an owl made of watermelon and summons photorealistic Painting. Sometimes it's different. If we go back up here, we can see that they don't really deviate too much when the Chaos is zero. We can see that also there's not many objects in the image and even the style of the outlet South looks fairly similar. Let's scroll down and check one of the extreme values. Technically not 100, but still a very high value of 80. Let's check it out. So here we have very different, different styles. Let's check another one here. And even here we got some even that are 3D like, like to photorealistic, but even the Photorealistic ones, they look very different to each other, especially when compared to the Prompts with zero Chaos. So that was a quick overview of the Chaos perimeter. And now let's go try it out. Let's go to discord. And let's actually use one of our previous or some of our previous Prompts. Well, let's take hot dog for example, and the Parameters channel. Let's go ahead and imagine it. So we'll try a hot dog without any parameters first. And then let's trigger it with Chaos of 80 or rather 100. Actually, no, let's keep it out at okay, we've gotten her first results. So this is the hot dog without any Chaos level. Really, it's without any parameters. And obviously we know that the default Chaos without any parameters is zero. So those are more or less and the Photorealistic Style. All of those, and we see a hotdog everywhere. Now let's go to the option where they Chaos is much higher and where the callus is at. We also got a hot dog in three options, but the first one is completely different. I think this resembles a dog. Those could be sausages or something, but I guess you'd never know. Basically by specifying a honey Chaos value, you're letting Midjourney go wild. And imagine this scene that you prompted for basically whatever you want it to create an, an, an absolute random fashion. For example, you also got a hot dogs like this one here where the hot dog is yellow. Never seen a yellow hot dog in my life. I don't know what the purpose of that was, but it's definitely more random. So these four images are definitely more random here. And it's quite unusual. Now I personally don't use it very often. I don't see it. Many use cases in Midjourney Gallery by other users. However, by specifying a moderate or low Chaos values. Especially when you have a more specific prompt, when you don't just have two words, but you specify a scene with details and you want to randomize that result, then Chaos is very helpful. And then the Chaos parameter can get it handy. And I actually want to show that here as well. So let's try another prompt. Let's take some examples from our previous Images. Let's take from the Photorealistic. Yep. Okay, let's take the cat. Alright. So let's actually specify multiple parameters here. Let's specify that we want the Aspect ratio of three to two, which is wide. And then we specify Chaos of let's say 40. And with those two things, Let's see what happens. This is a much more specific prompt. We have quite a more for quite a few more details. Specify it, right? So we say we want a photo. We specified a rocky Mountain. During winter. We said a cat specify an action. And we have a few other objects in here as well. Okay, let's check the results. So as you see here, Midjourney drifted from the initial idea a little bit and started to include other objects, basically randomizing the image even more. Also, this looks like a photorealistic illustration rather than a photo. But it is quite close. But anyway, even, even like house value of 40 is quite high, we still see the cat everywhere. We also see the mountains, the winter. So basically, most ideas, Midjourney did technically keep, right? But this is a good example if you want to diversify and get more extreme ideas of the image. So let's say randomizing a bit more with specifying the prompt. Again, I really use Chaos. If you have a more specific Prompt. I usually like to experimental with Chaos have 20 or 30 if I don't know exactly the image I want, and I want a lot of images with extreme ideas. So it's a way to generate more interesting results really. Now one more useful parameter that I use width Chaos and that I want to show you is called the Repeat parameter. So let's go back to the website. Year it is Repeat. Now what this basically does is it tells Midjourney to repeat the Prompt multiple times. So for example, if you want to specify Repeat to than it will basically do the prompt for an image once with a repeat parameter of two. But in fact, Midjourney will start generating the second variant immediately, or it will put it into the queue. So instead of retyping the prompt, for example, five times, if you want to receive five different options in a row, then you can just type this perimeter. And he gets very, very useful when you also specify Chaos as well. Because Chaos diversifies the image to a big extent, especially with moderate or high values. Then it makes sense to prompt a multiple times, to repeat the prompt, to see all the possible options that Midjourney can do. Okay, I hope that makes sense. Will actually go through an example as well. But let's go and check the Help page or it oh, yeah, Sorry. It's important to note here that Repeat is only available for standard or Pro subscribers. And it is only available to add this parameter while you are in Fast and GPU mode. So that's important to remember. So let's go back to our Parameters channel. And let's go ahead and try the perimeter really quickly. So I would say here, let's let's try to repeat it three times. It will ask you for confirmation that you are sure that you want three Prompts of this. Sorry that you want three results rather. Just confirm. And as you can see, it will immediately trigger these prompt three times. So instead of copying and pasting, you can just do that instead. Now we can just wait for the result. So we should see 12 results by the end of all of this. Now it's worth noting here that standard, basically every plan has a limit of how many parallel jobs you can run at one time. You can check that information on a subscription page in your account, on Midjourney website, you can just compare subscriptions and check how many parallel jobs you can run. Every extra job, we'll just go to the queue and we'll start generating when one of your parallel jobs has been completed. Okay, now let's quickly browse the results. So we've got three images. And as you can see, they are very different images if you compare them. But it's nice because the Chaos is high and we have 12 images will be looking and look at all of those 12 different images, which are slightly different than another. It's important to note that because it's being randomized so much that we don't get every single detail of the prompts that we asked for. So we don't see cat, for example, playing on every image. Sometimes the catcher sets. We also don't see mountains consistently. For example, for example, those two images there, which not really any mountain Aspect I can see to that. So it's interesting. So that's Chaos and Repeat, hopefully that was helpful. In the next video, we'll talk through other basic Parameters and look at their Effects. Okay, Awesome, Great work 25. Basic Parameters: No: Now we will explore a very useful parameter called Know, which basically means to exclude certain things from the image. Not just objects, but even to change an image Style which usually goes along with certain scenes. So let's, as we do, go to discord and will launch the Midjourney Help page. Perfect. So on the Parameters list page, you can find this perimeter here, which is no or negative Prompting. So for example, if you're a prompting for some indoor photo or Interior Design, but you don't want any plants. So this is a simple example. You just include at the end, no parameter and the word plants. Now in here, you can actually specify a whole Prompt, not just a single word. And that would work as a negative prompt. So let's click and dive a little bit into the theory and how that works. So the no parameter is basically just a shortcut for negative weight. Now we've quickly touched on this topic on our previous videos when we tried to produce a hotdog by separating those two words with the double colon. Right? We already went through this already. Obviously, if you don't specify any number, that basically just works as a Mulltiprompt. So it basically just assigns the same weight to both prompts. So we're prompting for a dog and a hot separately. And then we're combining the two together. If we scroll down, let's find the node here, which is negative Prompting. Perfect. Okay? So if you were to specify a number after the double column, so you would say this word. So to lip, read as an example. But you would specify a minus or some number. You would indicate that the red color, for example, or anything related to the red, should be missing or should be negative. Basically, it will not be completely removed from the image. But there would be definitely a lot less variance associated with the word red. And that's where the No perimeter really comes into play. As it says here, the no parameter is basically the same as awaiting part of a Mulltiprompt to a negative or negative 0.5 value. So as you can see here, this is the example of Prompt. Vibrant tulip fields. Separate double colon, red double colon -0.5 will be the same, or a shortcut of that would be to include the no parameter and basically have the same word red. Okay, Now let's go to or Discord and actually try a couple of examples. And this is probably my most FUN parameter because you can really see the different sailor. Perfect. Let's go to our first example here. The photo of a yellow armchair standing near a table in a living room. We did prompt for this earlier on as well with different aspect ratios. For now, let's just leave the Aspect Ratio by default. And let's actually Prompt without any parameters for now and see how it looks. Okay, so we have the results. Let's open it up. So all four variants, they look quite similar. And as you can see here, I also see also plants and all four options. And I also see a painting as well, or a picture on the wall. And all four options. So there's consists something consistent in every image. Now let's try to actually remove an object from these options. Let's say if I don't want any picture on the wall. Now, let's go ahead and take this prompt. And let's add double hyphens and no and Painting. Let's see how that works. While it renders also just a quick theory how this actually works, this negative Prompting. So basically it doesn't just prompt the very same image under the, under the bonnet and then just try to remove an object. Rather, how Midjourney or any other similar Ai works is that it searches in the dataset, in the underlying dataset for images that have photos of yellow armchairs. But the photos don't have any pictures or Painting described in them. Okay. Perfect. So let's see how that looks like. And what we see here is that we don't have any paintings. But you also may notice is that the actual composition is now very different. Before we had a gray wall, we had plants look quite similar. But now the old Composition of the actual image has changed. Just because we didn't want to include paintings on the wall. The backgrounds are very different. Certainly there are no, there are no images, but there's also different walls. Now let's say that the wall CRM or mostly glass. Now, why is that? Because it's not possible or let's say for example, glass walls or Windows. You cannot really hitting anything on glass walls or on Windows. So basically that's how it works. So it's important to understand that because when you ask to remove certain objects, it will actually change the image completely and might even result in a different style. Here it shows us windows because objects can't be hung on Windows. Let's try one more example. Let's try a seating Buddha on top of a mountain. That's triggered that. And let's see how that looks like with any exclusions, without any negative Prompting first. Okay. And now we tried to exclude anything that refers to realistic. So it's, the note perimeter is not just for objects, but it can also be for themes like 3D render or realistic. Okay? These images look pretty realistic, right? They do look photorealistic, but they do not resemble the actual photo. More like a render. So let's see what effect this will have on the image. So let's just trigger that. So again here, the key important thing is we're not removing a specific object with this negative prompt. Rather, we're changing the style or we're telling Midjourney to not have this Style, to not have this realistic 3D render Style of the image. So my suggestion would be to just specify the style that you want. But this is just an example of how you can affect the final result with negative Prompting. Again, not only just objects, but also styles. Perfect. So here we have the results. As you can see, the images are quite flat. So basically they're like drawing or painting style. And there's nothing related to photorealism or 3D rendering or anything like that. Cool. We can see that compared to this one here, how different they aren't from just some negative prompting. And that's our negative prompting. You can remove objects and Styles. And it will definitely explore this topic a bit deeper when we start working with Weights. But it's important to know this shortcut of no parameter, which can be very handy when you just want to remove a certain mood, Style or an object from an image. And the next video we'll continue to explore parameters 26. Basic Parameters: Stylize and Tile: This is a thiol, a video about the Use of Basic or most prominent parameters. And here we learn the Stylize and Tile parameters. Let's go to discord and to the perimeters channel. Now, before we start exploring, whoops, there we go. Let me just let's open up the Parameters section in the Midjourney Help page. And in the Parameters section. Where is that Over here? Style. Sorry, not Style, which is different actually, I Stylize. Perfect. Okay. Now that we're on the dedicated Help page, let me read the official explanation on the parameter. So the Midjourney bot has been trained to produce images that favor Artistic color, composition and forms. The Stylize parameter influences how strongly this training is applied. Low stylization values produce images that closely match the prompt that are less Style, they're less Artistic. High stylization values, images that are very realistic but less than E2 to the prompt itself. So basically, stylists parameter modifies the outcome by making images more Artistic and basically by enhancing the compositions. So let's see what options we have here. So the default value for a stylized parameter in version five at least. And version for actually is 100 and the range is 0-1 thousand. Now in Other Versions, it's slightly different or it's not possible at all. Like in this version, which is niji, which we'll get to later down in the course. Let's scroll the page a little bit and check how a Prompt illustrated fix would look like with the standard stylization. And that's by the way. Let me see here, I think version four. And that would be lowered Stylize outcome. As you can see here, this doesn't change the image too much. But if we compare, for example, these three Versions, and let's take the higher stylization of 250. You can actually see that there are more Artistic in a way. So it actually quite works differently than the type of the prompt. For example, if you're a prompting for a photo realism or a drawing or a vector, the Alpert use different results. So you would need to experiment and see what works best. I prefer to use it as default version, which is Stylize 100. So basically I don't specify this, but it can get handy in certain use cases. And this is a comprehensive course. So let's go ahead and try to practice this and see the difference from our usages. So here's my prompt that I've prepared. So it's an exotic wallpaper of Art of Buddha surrounded by a pink flower petals. Let's trigger this, which is a standard that's not specify anything. So really the default parameter that it's using a Stylize one-hundred. Now let's trigger to more prompts. And now we specify Stylize. Lets say one. And let's specify another. Or 1,000. Perfect. And let's wait for the results. Alright, they are ready. Now let's start from the default stylization. Now you can see that all four options here, I have plenty of detailing and the texture. We have shading, shadows, everything. Now let's check the Stylize one option. There is not much difference here, but you can see that it's definitely less textures and let's Composition. In other words, it's more simple. We can also see shadows here, but there's just a little bit, maybe less aesthetic and less texture Okay, Now let's see, Stylize one-thousand. I can already see here that these are much more like HDR type Options, way more shading shadows. Also the flowers are a bit more diverse and detailed. They look like more an actual photo. We even got a special option here with an Interior Design and other objects, but that could also totally be a random, of course. But overall, what we see here is that it does affect known as much as other parameters. But higher stylized values would be useful when you want to enrich a little bit of Composition and Artistic vibe into your photos. Okay? You can experiment with different prompts. You can use one of our previews once and then examined the results. And probably you can even stick to a certain stylized option that you like. But now let's move on to the next parameter here, which is more exciting and you'll be able to tell a difference here, which is the Tile perimeter. Now as before, Let's go through the theoretical first. Let's open up the dedicated page for Tile. Let's go First the Parameters list and have an overview. So the Tile parameter generates images that can be used as a repeating tiles to create seamless patterns. What this basically means is that you can create texture. And I will show you how that works. I will open up Photoshop then, and I'll show you how you can actually Tile that generated images, for example, to create a wallpaper texture. Now, we're talking about fiscal wallpaper, like what you see on walls. Let's open up the dedicated page and you'll be able to see how this parameter works. This parameter works with the most Versions of Midjourney. So here's an example. Let's open this. So Midjourney would generate this image, but then you would copy that and stick the same image onto all sides of it. It will generate a seamless texture like this. That's really important point about this perimeter. You can actually create textures and wallpapers. Let's go back to discord and try that. Let's try you Prompt over here. And let's try an illustration of chocolate milkshake with a straw, whipped cream, and cherry. And we want to make a wallpaper out of it. Now, let's trigger that without any parameters first. And now let's also try that with the Tile parameter. All right, here's the results. So this is the Tile version which reads something like a texture or repeating patterns. And this is how it looks like without Tile. It's just an illustration of a milkshake. Nothing exciting. Now let's launch Photoshop and let me actually Upscale one of these things and we'll show you can actually create a texture with that. Let's go with the first and the third option here. And let's see how that will look like on a texture. Okay, I have loaded up this image in Photoshop. Now, let me show you how would look like if you stack up in image on top of it and all around it. So I'll just do that right now. Copy and paste down. I'll copy and paste this image for the rate by copying pieces. Image to the bottom right. Wp says image for the bottom-left, copying pieces, image to the left, and now the very top as well. And look how beautiful that texture looks. You didn't have to do anything, you just use the Tile parameter. It's really cool. As you can clearly see, it created a perfectly seamless texture matter of a single game loops. What does amazing. Now, we don't even see the hut or the image edge anywhere, which is how powerful this can become. Now you can try this on your own with different objects and compositions. But this is a basic example of how you can use Style to create a seamless textures. Just OneNote I want to add here that I don't recommend using the Photorealistic Style. It works best with vector or illustration style. Look, whether or not too much texture or details in it. You can also try lowering the stylized option if your results look to oversaturated with different tiny details that could potentially break that seamless tile pattern. It. So now we've gone through all of the basic and most parameters. Later on, we'll explore how to combine these, but that's it for now. Awesome 27. Advanced Prompts: Prompt Weights and Mulltiprompt: Okay, now let's explore what Prompt Weights is and how we can help us highlight or minimize certain parts of a resulting image. Let's start by exploring the official documentation on the Midjourney Help page. And now let's go and find another user guide section in advanced Prompts. And we want to Multiprompts. Alright, so let's quickly check the basics. So we've already learned how adding a double colon to a prompt. It tells me the Midjourney bought that it should consider each part of the Prompt separately. And we've already seen this with an example of a hot dog. Where do you tried this with whatever previous videos, right? Where we basically added a double column between the word hot and the word dog. And basically if we are adding a double column without any number after the double column, This basically stands for Mulltiprompt and you can add more and more and more. So after the word dog, you can still keep adding more and more and merging all these different concepts and Prompts together. Let's read the official explanation. So it is possible to have Midjourney bought, considered two or more separate concepts individually by using colon colon as a separator. Separating Prompts allows you to assign relative importance to parts of a prompt. That's the most important thing here. If we scroll a little bit down to Prompt Weights. And let's go into examples here. We can see here that if we add a number after the prompt, after the double colon separation, we tell Midjourney to use a certain weight. In this example, we increase the importance of of, of the hot here. And that will be the and basically we double weight the hot versus the word dog here. And that's why it results in a much more dog image that's hot instead. Basically two means is it's twice as important. Let's check Other examples. From here, we can see that Weights are normalized, meaning that for example, if you specify a hot with a way to net dog with also a way to. Now we're just mean the same thing as hot double colon demag. So it's really, really relative weights. Okay? Now if you want to specify a weight for other prompts like this would be the second prompt, then we would need to consider only the important prompt. So if we wanted to highlight the cup or let's say the first part. Then by doing something like this, specifying, let's say a higher number. We're First part and then a lower number to the second part. That would not be needed. We can only indicate the important prompt, right? So you don't need to do to one, you can just those two. Apart from the situations, we can actually lower the impact as well. We've already gone through when negative prompt means as well. Where a no parameter can be used as a shortcut or a Prompt of -0.5. Okay, and let's take this as an example. So two lips red. We could highlight the tulips. And then we could actually lower and basically remove the red color from that as we saw before. Okay, now, this is a very draft example and this particular one doesn't make too much sense. But when we have multiple Prompts, three or four parts there, we can actually use whatever those parts as a negative prompt, basically supplying a negative number that could go at the end as the last part. And then with the first few parts of the prompt, we can play around with different ways, basically asking Midjourney to focus on certain parts of the screen. Okay, so now let's summarize all of this and then we'll put it towards actual practice. All words in a Prompt have a default weight of one, and you can increase that weight by adding a double colon and a number after it. Now let's continue on to the next video and we'll actually start practicing with Weights and Prompts 28. Advanced Prompts: Prompt Weights and Multiprompts Application & Examples: Let's go ahead and practice Weights and Multiprompts, and then let's examine the results because that's the only way to really know these things is by practicing them. Okay. Let's go ahead and launch Discord. And we'll go to our server. And then I actually suggest to create a separate channel for this topic. Let's name it, Multiprompts. Okay, we're going to try this Prompt over here and then you can find this in the full library of Prompts. We're going to try it without any Weights. Vintage armchair of Pablo pests, Costco's Style. Okay, let's trigger that. And this is our results. Now, all four Options look pretty good. And we can see the P Paso Style and all four of them. Now, let's go ahead and try to divide this prompt. And we divide a vintage armchair. And in pyplot, in Pablo Picasso style. That would be two separate Prompts merged together. So let's check how this would work. Now, we'll definitely see more of the Pablo Picasso style, because now there are weighted equally or boots. It's now let's examine this. And what we can see here is what happens when you split that and tell Midjourney to weight each part equally. The vintage chair and the Pablo Picasso. I can imagine that if we wouldn't ask only for a Picasso style without specifying the Medium or anything else. Weak mid get something similar to that. But because we also specify an armchair, those two parts are now merged with equal priority for each part of the Prompt. They're the two parts right there. But now let's influence, sorry, let's increase the influence of the Pablo Picasso style, but leaving the armchair also in the image. So how do we do that? Again? What we're gonna do is bring its type in. Imagine. Let's add a double colon, and then we'll add in a number after that. Because we want to increase the weight of the Picasso's style and we add the number two. Now just remember that you should not have any space between the double column and the number. If you have a negative, that would also look like this without being any space. Okay, let's try with the weight of two. Again. Now what we're doing is we're waiting. Pablo Picasso style more weighted than armchair, actually double weighted. So let's see how that worked. We could definitely see here a much higher influence of Picasso style to an image. We can still see an armchair. At least I would think that that looks like an arbitrary. So everywhere we can see an armchair. But the Style on Composition is completely different. Feel free to experiment with other objects. So you could try in Pablo Picasso style to imagine a particular scene or object. You'll only need to add an extra weight to the Pablo Picasso style to receive a similar outcome. Now let's try it next prompt. In this example, what we're doing is we're basically requesting a scene of a photo of a French restaurant. And then what we're doing is with a Mulltiprompt is where populating that scene with people. So for example, here we want to see a couple enjoying a french breakfast. So let's trigger that. Weight of Mulltiprompt usage can be very, very handy when you want to populate certain scenes with object. What you can use, you can First describe the scene with one part of the prompt and then multi colon. And then the second Prompt describes the object, or for example, at Theme also. Okay. Alright, so those are the results. And here we can see very French vibes and we can see the couple in every image. So the Mulltiprompt worked well. However, I can see a couple of options are not really a photo. So how can we fix that? Let's just add photo or photorealistic as another part of the prompt as a third option. But let's try one more thing. Here is the prompt Fold-over French restaurant couple and drink bezels. And we asked for a rainy day. Okay? By setting the theme and changing, this should change the composition a little bit. So I expect a more Outdoor option doing this now. Again, Mulltiprompt, it means you can have more than two prompts. You can have 34500 prompts if you want. Nice. So this is the type of composition that I want. We can definitely see raindrops on all the different options here. We can see the French vibe and we can also see that tuple everywhere two. Now, let's try a little bit more rain. And we specify weight of two after the rainy day. And let's see how that works. Again, what we're doing here is we're assigning more weight to the rain Prompt than the other Prompts. Okay, perfect. Let's check the results. And now we can see definitely the influence of a rainy day Mood a lot higher than we saw in the previous photos. Also, we can see pictures, especially this one. The rod even looks more wet. We can also see umbrellas and every option as well, which is obviously umbrellas are related to rainy days, which is awesome. Speaking of umbrellas, speaking of obelisks. What if we don't want any umbrellas in the picture? Well, let's enhance our prompt. So what we'll do is we'll keep rainy day there with a weight of two, but we'll add an umbrella with a weight of minus one. Okay, let's see what that will do. Now we have for Prompts, we're waiting rainy day would double the weight and we're waiting umbrella with negative one. Wait. Let's see what happens here. Okay? But I can notice here is that the rainy Mood is still here, but the actual ring is gone. Especially with this option here. So it's more of overcast weather. But Will, we can definitely see here is that the umbrellas are all bond. So that is one more way that we can achieve a rain. Sorry, there's one more way we can achieve a rainy Mood. So we looked at rainy day colon, colon two. Let's check another option we have here. And again, all these prompts can be found in the PDF. Instead of saying we want a rainy day, Will we can do is say sunny day with negative weights minus one, right? Because that's the sunny and rainy are opposites of each other. So instead of saying rainy day to, we'll say, we're saying sunny date, negative one. So let's see how that works. So it definitely shouldn't have anything related to sun or a well that might not be rainy, but it's definitely shouldn't have any sunny aspects to it. And let's check the results. And what we can notice, for example here is that the option in this option here is that behind the windows the weather is not really sunny. And here we don't see any windows. But what we can also see here is that all the scenes are indoors. Because what happens is we lowered the weight not just from sunny, but also from the ward day. And I would say that day could be referred to an Outdoor in this context. So that's why it made him more indoors. So there we go. Those are two very good options. But Mulltiprompt thing is probably the most things that you'll play around with because you can use so many different things here. Now those were very simple examples that use Multiprompts and Weights to achieve certain results. But now we're ready to use this techniques to achieve even more desired results in our next video. Awesome. This is very exciting guys 29. Parameters: Seed: Now in this video, we'll learn how to obtain and use seeds. Will learn what they represent, and now they can contribute to consistency in your image Series. Additionally, we will explore a special Prompts structure which can assist you in placing the same object or character into various scenes, potentially creating a storyline or a storybook. Before we begin with Prompting, let's check the Midjourney Help page. And let's go to seeds here. So it's located in the user guide in the Parameters category. Let's first read the official explanation for seeds. The Midjourney bot uses a seed number to create a field of visual noise, like television static, as a starting point to generate the initial image grids. Seed numbers are generated randomly for each image, but can be specified with Seed or same Seed parameter. Using the same Seed and prompt will produce similar ending images. As you can see, Seed is a whole number between zero and this number. And there is also a slight difference on how the seeds are used by different models. So with previous models up to version three, and including test and test peep will produce similar Composition colored in details, but not the similar or identical image. Whereas the version 4.5 will produce nearly identical images. However, with time, I did find that version five or 5.1 does indeed produce nearly identical images. But version for when you Prompt with the same Seed, It might still deviate. L produce similar but not identical looking images. Okay? So as it says here, Midjourney will randomize the seed every time, which is why you get different photos for every Prompt. Here is an example of three images that were generated using the same Prompt. But without specifying the seats, which is the default way of prompting. That would be the first time prompt, that would be the second, and that would be the third. They all look a bit similar, but different Composition backgrounds and the object style is also a bit different too. But with the latest version, as you can see, that would produce identical images. So when there is a Seed parameter at the end with the same number, of course. Now let's check, or you can actually obtain the seed number. So the first way would be to go through your own Gallery on Midjourney's website. Now let's say you want to try and reproduce this image. All you'd have to do is go to Pompeii and then choose Seed. And then wait until it's finished actually copying to your clipboard. So another way would be to add a reaction to this message. So let's say that we want to get the seed from this illustration. All we need to do is attach a reaction. So just go to Add reaction. Search for envelope on. I think this is the correct one here. Now, if you don't see a message or anything happens after that, then you probably need to change your Server privacy settings. On your privacy Server. In the privacy settings menu, make sure that you have direct messages enabled. Those basically allow you to message you directly by other members of the Server. And as you can see, Midjourney bot is one of those members. So let's try again. Let's try attaching an envelope. And we should see a direct message coming from the Midjourney bought on the top-left. Look at and perfect, Here is your Seed number. Now let's go back to the Parameters channel and let's try to Prompt and include a seat parameter. But first, let's try this Prompt without a Seed number. Looks higher. Fresh, new Prompt here So we have here we have a photograph, Man viewed from the back, facing away, hunched over Laptop, a few crumbles, papers, balls on the floor, cinematic, dim light. And we've specified with a specific aspect ratio. We'll get to the full explanation in next chapter as to how to use these lights. But for now let's try this prompt. And then we'll go into extract the seed number. Alright, so here's the results. Now, let's say we kinda like the aesthetics of the first version here. So let's Upscale it. Perfect. Now, let's get the seed number. We added here. Let's copy it. Now, just a small note. You could also get the seed number from the grid version. And it would be the same, at least on the latest versions. Okay, so now let's take the very same Prompt, but we'll add in a Seed parameter. Results have been dash, dash Seed, and then the taste. Thus also try to change the prompt a little bit as well. This man, we're going to prompt for a woman viewed from the back in this scene. So let's try this, which just start off with the same static as the one before. But we've changed the Prompts slightly. Alright, it's ready. Now if we compare sexual, move this part here, Okay, Perfect. We're now, if we compare the second version here, the second option. And let's say that's the first one. To this option here. You can notice that there are some similarities, at least in the aesthetics. So here we have some kind of a button to shelter or underground room. At least we get that kind of feeling or vibe. We do see some windows here, but here we got a very similar Mood and P, and the same for the four other options as well. So there are similarities in terms of aesthetics, haulers and scenes. And that's because we use the same seed. So in this way, you can use seeds across different images. If you have similar prompts which described the theme, the Mood, the lighting, and everything else. Than with the same Seed, you can try to achieve similar looking images with similar aesthetics. Perfect. Okay, now let's check the final prompt, which is a special one. And what we'll do is we'll actually, let's remove the link for now. So this would be here, the structure. This type of Prompting. Here, we're trying to achieve a persistent storytelling. What we have, let's say a specific object or a character. We want to place that character in different scenes or scenarios. So you would need to include a link to an image with your object or character. Now what Act has a image Reference? Then you described in this fashion a Setting. So let's say this character is sitting on a house garden in the evening. And we'll have a cat here as an example two. Then by splitting the Prompt with double columns. Next, you specify an object or Describe your character. Again, here we have a ginger cat. Next, let's go towards the story. The story here is combining here. Where basic combining your, the Setting, the character, and we're adding in some action or a scene. So for example, let's just say it's sitting. And then it's also important to raise the weight a little bit of this particular pump because again, it's coming afterwards. Then at the end as the last prompt, you Describe the styling or the theme. Here we have a photorealistic Style with a cozy Barton Theme. And we also have a wide aspect ratio as well. So that would be the structure that can potentially help you to achieve that. Of course, you have to experiment because it really depends on the setting that you have and the character. So let's check our character. So this is the image that I'm providing Midjourney. Here we have a ginger cat So we could expect something similar in terms of the posture and the actual object. So it's quite a furry cat with a pig tail. Perfect, so let's add the link to that cat here. And let's try it now. This is exciting. Okay, perfect. Let's check it out. And as we can see here, we see a very similar looking cat. Of course, if you want to achieve a very close likeness of an object or the character, you would need to re role or B-Roll. The option that matches the closest, the Seed and the object. So that's the night setting or the evening burden. Now let's use the same Prompt. But here we change that, that this is a sunny day and this is a summer day, cozy garden Theme. Everything else should be the same here. Let's see how this affects the results. If you could actually get similar looking object but just in a different setting. So it's very, it's very useful if you want to have the same object in different settings. You can also see that in the First Prompt, we also changed it to Setting in a house garden in a sunny day. Here we have House garden in the evening time. Okay, Now let's see the results. And here we can definitely see a very similar cat or a character. Also, the setting is quite similar, but now it's quite bright and it's sunny. So let's open up the previous one. We can definitely see that at least on free Options and evening scene. And if we check the second prompt, we can see again a similar scene, similar objects, but the Theme and the Lighting is completely different. So try to experiment with this type of comp, sorry, this ever prompt Composition. It's also good practice to have the object or character that you want to use in your storytelling to be in different angles and different emotions. So if it's a photo of an animal or a cat, then it should be a like a full body photo. But if you want, for example, to prompt for a CAT standing or walking, then it's best to use a picture that where you have a similar position or posture of the character. So trying to experiment with this and the seat, because in the Seed we saw how the same photo and a different had a man at the beginning and a woman at the other settings. So try to experiment with both the Seed Options and the consistency of the type Prompts. Awesome 30. Stylization: Medium: Alright, from this video we'll begin to practice stylization, which is a very important part of any prompt, sets, the tone and overall styling. They can help produce very specific results. Okay. Before we begin to practice, let's go to Midjourney, Help documentation, and remind ourselves. So getting started, next steps and explore Prompting. One of our previous videos that we already looked at this page. And we checked on how a graffiti style a cat would look like, and how the styles, you can actually change it to different types of styles. You can't just set the style, but also things like the tone as well by specifying the specific decade, for example. Or even cleanly just playing with the color like this over here. For example, like this saturated or black and white Style, or the acid green, or some specific color combination. Now you can browse this page a little bit more. But for now let's start our practice and let's launch Discord. Perfect. For this purpose, let's create this dedicated channel and call it Styles. Now with this course, we're focused more on a photorealistic type of content. So let's begin with a photorealistic prompts and then we'll briefly check out a few more popular options. So this is a prompt that I've prepared. It's very simple, but it's a photo of cherry blossoms in Germany during the spring season. Awarded Painting Photography. I was like to put that because it makes it a very aesthetically pleasing. So here we emphasize in the photography. So we specify this twice. We say it's a photo, but then we also specify award winning photography. Everyone, photorealism, that's a cool trick that I've learned. Okay, let's observe the results. Actually. To check out the details a little bit close, Let's Upscale a couple of the Versions. Let's Upscale to and Upscale a three. And what we see here is a very beautiful photo. To reach that level of aesthetics, it's a good idea to include things like Award winning or amazing, or some other keywords to help Midjourney understanding the Style and the vibe that you're looking for. Okay, here's the second one. So those are good examples of photo and award winning photography that helped us to specify the style or let's say highlight. I liked the idea w1 to receive a nice and beautiful photo. Now, let's try one more photorealistic Style. This is the prompt. So this time we want a fisheye style of photo that captures the urban landscape with tall skyscrapers. Skyscrapers and captivating Style in captivating noon Style at night. So here we're prompting for a photo, but we specify a special type of Lense. Let's wait until the results and then we'll examine it. Alright, let's check it out now. So here we see here a specific photo style that will be prompted for with an extremely wide angle. And on top of it, of course we see the neon, the skyscrapers. Everything we prompted for. Those are the Photorealistic different styles. Before we move on to the next video, let's browse a few more options and a few more styles. Let's start with a 3D render, which helps to produce images that are more like a 3D game or a PC game or a PlayStation game. That looks like in this particular prompt. This is another style you can use, something like Unreal Engine or 3D render Specify go horror Theme as well. Again, we specifying the Unreal Engine with 3D render. And then we're also Yep. And we're also specifying 3D as well. So again, this is an Style you can add to your prompt and make it 3D. And then we're also going to specify photorealistic render. But we're not specifying that it's a photo, we're specifying it. So render, which makes this different, which makes it more like a 3D game. Let's actually Upscale. This scene indeed looks like a game, like a 3D game with very photorealistic graphics. Let's go ahead and check the other one. Same yet, this doesn't look like a photo. So it's just another example of how you can alter the style of your image by specifying specific keywords. In this case, it was 3D render and Unreal Engine five, and photorealistic render. Now let's quickly try a couple of more. Let's see how a street graffiti of cherry blossoms would look like. Alright. So we see here a nicely drawn graffiti of cherry blossoms. All images are in photorealistic Style. So it's like more of a photo of graffiti. Even though we didn't specify that we wanted a photo of the graffiti. Midjourney has chosen this by default because in the underlying dataset of Midjourney, I'm guessing most of the graffiti, these are basically photos of someone's graffiti on the street. And one more keyword of street actually helped to choose the style. Okay, now let's try it like totally different style and vibe. So we want a vibrant and lively urban landscape in a comic cartoonish style. Let's see what that gives us. Alright, I'm perfect. Yep, that is a perfect example of a comic and Cartoon Style. It's bright, it's colourful. Some of them look like they're hand-drawn. So nothing referred to a photorealistic Style or any other Style. So basically it's quite simple. Now let's try one final version for this video, which is a charcoal pencil sketch. And what we want to see here as you want to see a series of footprints in the sand. So we actually have three keywords here. So it's charcoal, pencil and sketch. All of them Describe quite well what we're looking for. And again, repetition here is the key, like how we solve with a 3D render one before. Let's open this up. Perfect. That's another example how Midjourney can interpret the prompt and generate images in this style that you're looking for. Now, you should feel free to experiment with all the different styles. We went through, photorealistic ones. We went through charcoal, we went through Cartoon page. You can also check the help page for inspiration as well, but we'll go through a lot more as well in subsequent videos. Awesome 31. Stylization: Artistic Styles: Let's continue to explore stylization. And in this video we'll learn how to apply a certain artists style and how to check if Midjourney has been trained on it. This is probably the most exciting video here. Now before we continue to practice, I would like to show you one very handy document that I've linked down in the video course notes below. Just find the link in the description of this video. So here we have a public document. Now, before we continue to browse here, make sure you select the current version. So here we're at version five. So basically how you can use this document is you can check and browse if Midjourney actually knows this specific artist's style. If we scroll a little bit, we can see that mid journey has been trained on very different styles. Photorealistic styles, for example, on photography, very different Painting type styles, and even some 3D render or 3D artists type styles as well. The way this works is that you actually make your prompt and ask Midjourney to create this image in the style of a specific artist. Now, it's possible to rephrase this a little bit. So in this example we have a Prompt of a cherry blossom in Germany. So we add desk prompt before. That's how it looks. If we don't specify any Style apart from the fact that it's an award winning photography. This is a photo. But if we scroll down, now, we can add to the end of it in the style of acos Mayor. Okay, so let's trigger it. And let's check the list four of artists and let's try to find this article. Artists here, I pass major city. Where is it? Where is it? Let's go ahead and sort the list alphabetically here. I'm just going to sort the sheet so you can actually find it. There we go. Perfect. There he is. Oh whoops. Clearly don't OCL blue sheets as well as I thought I did. Perfect. Okay. So we filtered this list and we found our artist. So this is the Style photography of acos major, and we should expect this type of photography here and perfect. It is. Yep. So we can see it's a very Type, middle minimalistic type Style. And then we can also see here that's a bit hazy. We can still see the cherry blossom. So we see pink color as we see the tree. But we can clearly see a big difference from this results compared to the regular photo where we didn't have a specified Style or we added photography type style instead. Okay. So specify the artist Midjourney has been framed on these artists will make sure you specify them in your design. Let's try one more artist. Let's keep the scene the same. For now, we'll just keep it the same. And now we're prompting for a photo of cherry blossoms. But if you notice here, we've added in Award winning underwater photography in the style of this artist, Jason daycares Taylor. So let's go to the list again and find this artist. Perfect. You can see that this artist focuses on underwater type images. And it's been trained on this object. Midjourney has been trained on this artwork. Perfect. And we've helped Midjourney to understand it a little bit better by adding an underwater keyboard as well. To be honest, in this prompt would also make sense to add words like minimalistic and hazy, and simply highlight Well, we're looking for in these particular styles. Okay? So we did this in this prompt. And we've got quite an interesting results. We do see that people that we didn't prompt for that is mainly because the artist is mainly photographer and an underwater statues. So let's try our next prompt. And again, these are a little different styles that are fed by the actual artists themselves. This time, we're prompting for graffiti or street graffiti of street blossoms in the style of Banksy. Now we already did that street graffiti of cherry blossoms. But now let's see how the style of Banksy will affect the underlying results. And bank C is a very famous graffiti artist, if you don't know. So I actually don't even know if this less technically has baked see, but basically it's just so popular that it should have been trained on images of Banksy. So if an artist is well-known, it doesn't necessarily need to be trained on, doesn't necessarily need to be on this list. Okay, it's not. And that's perfectly fine. Banksy, such a famous artists that Midjourney will just know Banksy Style. Unf banks is not on the list of artists since beings he's so famous Midjourney's able to understand if Banksy Style. Let's check the result. And yes, I can confirm that this is traditionally the banks, these Style. And you can see here that's a street graffiti. We also have our cherry blossom on the street graffiti as well. So let's put the Google obeying see, and you can see what I'm talking about. Perfect. Let's go to Images. And yes, those are the famous artworks of Banksy. So you can see it's very similar. We can even compare your Midjourney did understand this very well. Just proves that you can use any famous artists style. However, there's a good practice to stay within the medium of an artist. So if an artist as a photographer, then it will produce best results if you prompt for a photo in the style of that photographer. So much of what we did for the underwater example. If the artist is a painter or some other type, than working with Prompting within the same Medium would definitely produce the best results. Now believer, it doesn't hurt to experiment. And sometimes experimenting with switching the Medium can also produce a very interesting results. So let's, so we should try that. But first, we'll actually try to, more Prompts. We'll do over here is priced them surreal expressionism. So surreal express lesion in oil painting style of a volcano eruption in Hawaii. So here we do not specify the specific artist, but this is also a very known Style. So expressing Islam and oil painting. And we also specify which kind of expressionism as well. Okay, let's try this and we will examine the results. And let's actually Upscale one of them. As you can see, this is a very good example of oil painting and expressionism. We also see some hints about surrealism, of course, and the volcano. So feel free to use that file, that Google doc file and experimental with other Painting Styles. I now let's try our last prompt for this video, which would be a Fashion Photography, which has its own style. Fashion Photography, full body shot. A postmodern female Fashion. I'd where studio setting. So we're basically also highlighting that we do intend to get a Fashion and studios style of photography. So just as here, we asked it to create a specific oil painting style with surreal expressionism. In a very similar way. We requested here Midjourney to create Fashion Photography, which has its own aesthetics. And that's a lot of headway are definitely, definitely took that and imagination. I think it's a very good example though, how you can impact the styling of the image, right? So we could just prompt for female with hardware and not specify any styles. But because we wanted for the image to be in a more fashionable Style, we indicated that with multiple keywords, like Fashion Photography than female Fashion, studio setting. And all of these things working together, redid the aesthetics that we see in this, in this photo. Okay, Awesome. 32. Stylization: Combination and Enhancements: With this video, we complete the exploration of how to apply different styles. We will try to combine two styles in one and learn how to use Mulltiprompt to combine these two styles together. Let's continue in Discord and in the styles channel. And obviously, we've been prompting for an image with the right here with the acos major Style, which produces a Hennessy and minimalistic, photorealistic Content. Now let's try to combine. And we prompt for a. Let's see here fisheye style photo of cherry blossoms in this style of acos major. We also prompted for a fisheye image with the skyline in neon style in previous videos. But now what we're doing here is we're actually combining two different styles together in one. Without the use of Mulltiprompt, we specify one Style in the beginning, then a second Style at the end. And we've got a pretty neat result. Awesome. So we can definitely see characteristics of both styles. We can see minimalism from the major Style. We can also see the fisheye and the cherry blossoms as well. So that's a very basic way to combine a tear. I do recommend that if you have to Styles to have one in the beginning and one in the end in this fashion. If you want to say an artist, it's better to put the artist at the end because that has a high influence than the medium description. You can put at the front. You can experiment with both ways. But I found this to be working the best. Now let's try another prompt. And in a similar fashion. But instead of specifying an artist, Will we have to specify here is a Fashion photograph. And at the end, we want to do it in the style of the Burning Man Festival. If you don't know what that is, you can also Google and check out the aesthetics of the festival as well. It's basically white famous festival that takes place in the desert of Nevada every year. We bought a result. Let's actually Upscale one of the images. Let's take the third one and let's check it out. Cool. So this is a very good example of combination. So we have a Fashion Photography. We have a Nordic looking man, and we also got the Burning Man style dress and the aesthetics overall. So those two examples can show you how to combine the style of the photo or style of another medium and a style of a specific artist or another tab type of well-known aesthetics. For example, at Burning Man in this case. Okay, cool. Now that we have that set, Let's try a totally different way of combining styles using a Mulltiprompt. Okay, So with this Prompt, we're asking to create a thicker that stands near a volcano. We do not specify any kind of style here, we just specify the scene. Then by splitting the Prompt, we're going to start next one where we asked for the monopoly Style. We're going to check that artist as second. And then we're actually provide a neither Style, a second style called Amy guacamole. Hopefully I pronounced that right. I'm sorry, Amy. If I didn't, Let's prompt this and let's check the artist in that document. Let me just get the document open here. So let's filter that. We're sort Let's clear the selections first and then type. Okay, I like monopoly. So that's how the alec monopolies Style looks. So it's kind of graffiti and very colorful Type Style. It's also a little bit comic with color splashes and a certain look overall. Now let's check the second artist. So here it is, Amy guacamole. Now, this is also a painting style, but with certain aesthetics. It's more of a color, sorry, curly color. It's definitely not graffiti. Which basically means that these two styles are quite different. And let's see how Midjourney combines them together. Cool. Let's open it up. So basically, what we've done here is combined two totally different aesthetics. So I don't see any image, the graffiti or let's say exclusively the graffiti style or the second imi Style. But it is interesting to learn how Midjourney actually combine those two styles together. I can see some further details, for example, in these clouds. So this looks like this specific characteristics that are coming from the second Style, which is Amy. And we can definitely see plenty of colors everywhere and aesthetics that hint to an alloc, monopoly Style. All right, Let's try another one using Mulltiprompt. So we saw how Mulltiprompt can, can combine to styles together. Here we're asking you to read a street graffiti of cherry blossoms in the style of Banksy. And that we're going to add a second style called Pixar style. However, if you notice, I do not split the style of Banksy, so that all serves as one prompt. And then I want to mix it with the Pixar style. Now, Pixar stands for a studio that creates 3D Cartoon movies. So if you ever seen a Pixar movie, you know what I'm talking about? Frozen finding nemo, that kind of stuff. We don't specify any weights, so we're just going to use the result as a Multiprompts that are going to be equal the weighted. And now let's wait for the results. Okay, let's examine the results. And again, in this case we see aesthetics in graffiti in some of the options, right? But we also see the Pixar style, a cartoonish style as well. We did get one of the images of the graffiti on the Wall Street images. So let's the suction you who for prompt because we want that graffiti style to be more prominent, whereas in this case only showed up in one photo. So instead of saying just Pixar style, we say pixel Style graffiti. But let's see what that gives us. Because we want, we want graffiti to be in this photo as well, not just Pixar. Okay? So that really fix the issue. So it's important to remember that when you are doing a Mulltiprompt, the second or third prompt is being understood separately, and that Midjourney basically combines those. So if you're asking here for something that is not related to graffiti, but you want to receive at graffiti than Midjourney, wouldn't know. Um, but if you specify that or for example, if you don't see some other objects, you could add this to the second prompt. In this case, graffiti are Banksy Sal was appearing on both Prompts. That's important to remember when using Multiprompts. Otherwise, you can also try to experiment and add a third part of the prompt and then adjust the weights a little bit as well. Let's examine the results here. We can definitely see that all of those options are graffiti is. We can also see there are cherry blossoms. And we can see that it's Pixar or Cartoon Style. And we can also see that there's a Banksy Style in there as well, which was the style that we merged in. This is very cool because now we've learned how to merge multiple styles together, like in this case, Banksy and Pixar, and get the exact image that we want to get. This video series, we've learned how to stylize your images and how to combine multiple styles into one 33. Stylization: Theme: Now that we've learned about Styles, PMM is another way to stylize your compositions. In this section, we will explore how to turn any prompt it into a specific scene, giving you a special Mood. So let's go to worry Discord here. And we'll continue to use our Styles channel. And let's start off with a simple prompt. Okay, so it's just a photo of friends walking along the coastline. Very simple. Okay, let's examine the results. So it's a very Basic Composition. All oxygens are in photorealistic Style. And we can also see that the weather is good. And there's no rain or snow or anything like that. But now let's try to add a Theme into this. So we keep this Prompt, but we add in another one. And by doing this, we're creating a Mulltiprompt, autumn rainy day Theme. And let's see what this Gibbs again Theme and Mood here are used pretty interchangeably. Let's see what the henry De PMMA gives us. Okay, Let's open up the results. We can see here that the weather has changed and it's apparently raining. It's wet, it's autumn. Their leaves everywhere there was multi-colored. There's others like orange and yellow here. The dress of people has also changed slightly as well. And that's by just adding that other Mulltiprompt for that Theme. Now, let's increase the effect a little bit by adding a weight and a couple of other keywords. Perfect. So at the end of the second Prompt, we're going to add in sad and romantic theme. And we're going to up the weight 1-1, 0.5. Again, if you don't have the weight, it's one, then we do always recommend starting with lower weights. I mean, if you go straight to the weight of two or three, this can actually break the Composition and create some very unpredictable results. So try to increase the weight slowly until you get to the actual results of what you're looking for. Here. Are the results. Nice, very romantic. They're holding hands. You can see that we could definitely see a higher influence of an autumn or romantic theme. But what we also see, for example here, is that the coastline that we had is gone. We can also see the photo style that will be prompted for at the beginning has been diluted a bit. It's not really photorealistic anymore. We see some hints of painting. Let's try to fix that. To fix that basically you can just add the more photography keywords into the second Prompt or another medium that you're Prompting width. If you're going to go back to another cell, you can add it as another keyword. But let's try another different prompt. And I really want to show you this prompt because this says prompting for a Movie still. So that's usually produces images with a movie Type feel. The word Movie still it's been trained on that. And most of these are usually photorealistic. So it's very similar to a photo. But later we'll add a special PIM. Okay, so first we need a Movie still for that. So let's just try that without any Theme at all. So it's a group of people sitting at a table. I like that. I like this and type of Movie still Action. Love the Art style here. And here you can see the influence of the keyword. All the options are colored graded. And also the composition looks at bit different. If we were to just ask for a typical candid photo, this looks more of a staged or directed seen, like how you would have in a movie. So now when we do now is, let's change this theme here to a sci-fi theme. Okay, Perfect. With a closer look here, you can actually notice that all the options are like modern day Options. I mean that the time in those scenes are mostly, are mostly modern day. Or maybe for example, this is a little bit like passed from the past. And now we want to change this to a sci-fi. And let's see how that affects the image. Okay, Perfect, and this worked perfectly. Now this did change the composition a little bit. So we don't see any option with a closer look. But then of course, all the not specified options are being randomized as far as you now. Now let's try to add a bit more sci-fi, and let's increase the weight of that sci-fi theme and see how that affects the image. Perfect, Let's check it out. We can immediately see the influence of the weight. So we have a lot more details and are relative to the sci-fi scene or the sci-fi theme. However, we did lose a little bit of the movie effect. Some images look more like drone images than what you would see in a movie directed or acted way. And once again, to fix that, you can just try to add Movie still or relative keywords in the second part of the Prompt, or at a third Prompt called Movie still an increase the weight. Let's try a totally different example now to finalize the P message here. Okay, so here we are prompting for Chinese lotus blossoms in the style of Gong Bei. We're also going to add in maritime game. So let's try to prompt for that. Alright, let's open it up. As you see here. We did get everything. We asked for. The PIM, the Style, also the objects, the lotus blossoms. And this was without specifying it as a double prompt or a Mulltiprompt. We've just added at the end, the merrier time PIM. This also works great when you have a simple prompt. However, when you're Prompts start to grow and become more complex than it's much better to separate the team or the style. However, you should, of course, always experiment and test works best because every scene, every topic, or even different mediums, they all produce different results. Some work better with Multiprompts, some without. And now let's actually increase the effect of the maritime PIM. And we'll also do a little trick and increase our stylization. So let's separate the theme and let's add extra stylization to it. With this, we're basically enhancing the overall aesthetic to be more maritime Theme as maritime team is its own, is its own prompt. And we're also stylizing like crazy the 750. So let's see what happens to limit. Awesome, awesome, awesome. Let's competitors. So we've lost the style of gang boy here, right? Because it's very obvious here and after the maritime Theme has a much higher priority than gone boy. Also, by specifying the stylization, we've also got a more textured and more detailed image. The Stylize works really well with paintings when they are not minimalistic, when there are more details and more objects. So it's always a good idea to increase it and check out the results. Perfect effect. Now we've understood how to apply themes to our images 34. Crafting Visually Appealing Images Part 1: In this video, we'll learn how to enhance our prompts to create more visually appealing and aesthetic scenes. Now let's go to Discord into our photorealistic channel. Because we'll be focusing on photorealistic Content in this video. Now, let's start with a portrait photo in the style of how low Reversi. Now this is probably my favorite hikes for creating photorealistic portrait Type photos. You can obviously check online that fatale of the photographer, but mainly he focuses on Fashion Photography. And Midjourney has basically been trained on his type of artwork as well. So this will be our very basic Prompt. It'd be want a tall image because that's what a portrait is. Okay. And after that, we'll start enhancing our prompt to make it more visually appealing and more. So let's first get the outputs. Okay, let's check those portraits. Okay, let's actually Upscale one of them. Let's take the first one. We can see is that that's a portrait with the side view and a half body Shot. And we can also see the specific aesthetics that are related to this photographer. And basically that doesn't look bad. That looks pretty good. Well, let's see what happens if we enhance the Prompt a little bit. And let's explain our idea a little better. So instead of just prompting for a portrait in the style, we will add some interesting background or Theme. So let's add in volcanoes background. There will also specify that we want a strong and elegant aesthetics. We want a sharp focus, because here, I think gets a little bit blurry. We also add the award winning portrait photography, and we'll also add minimalistic. Okay, so now let's try that again, Writing more detail on the Theme and Mood to create that more Visually Appealing scene. To get some more inspiration, you can actually go to the Midjourney Gallery. Let's go in and sign in. And you can check the hot or the top rated images. There. You can get some inspiration here. Some of the Prompts here are very simple, but most of the top rated Prompts have very specific keywords that help to describe this particular scene. So you can grab some inspiration from here. But I advise you to not just copy and paste the prompt, because sometimes people do tend to over Describe the prompt by adding too many keywords. Some of them even conflict with each other, but honestly it is a good place to start and if you are a bit lost as to what to create. Okay, Now let's see what changed. Now. Immediately, you can notice that you have a much higher and sharper image. We also have the aesthetics and a bit more creative seen. Unfortunately, we didn't get a volcano here. But if we do want that, we can also add a Mulltiprompt. This has volcano, and then you would definitely get a volcano. But the idea here is to enhance the aesthetics and make them image more aesthetically pleasing or more Artistic in a way. Let's Upscale a couple of options. An awesome, we immediately see that it better represents the style of Palo Reversi and then Fashion Photography in this option. Feel free to experiment by adding more dedicated keywords to this particular scene or Style. This was just a very basic example. But now let's try to create a more complex prompt where we also add in more elements about the mood and Theme. Let's add any specific decade as well. Okay, so let's add in multiple lines. We also have been in multiple times that we want photorealistic Content. We add a film, we add a cinema Shot. Finally, detailed. That's quite important. If you want to have a more detailed image, then there are multiple keywords like color grading, post-processing, and post-production, that are responsible for color grading and balancing the overall style of the image. This will result in more Visually Appealing scenes. The keywords, elegant, post-production, post-processing, depth of field. And you'll see that actually worked as well. There should all hello, increase the photo realism and quality over all of the image. So let's see how that works. Let's go ahead and Upscale one version so we can see it closer. Perfect. So that's the result that we can get. So we can definitely see the 20s in America. We can see the jazz night. We see a little bit cinematic Mood, plenty of details from a finely detailed keyword and depth of field or DOF. Basically what that means is it creates a very blurred background and liquids you see on iPhone portrait mode with the main object in focus and all others like post-production, post-processing. Those are all adding up to the aesthetics, the Photorealistic aesthetics of the image. Here is another version that's Upscale it. Same here, plenty of details, very moody Type picture and with strong accent to the topic that we've prompted for. Now, one important note that I want to make is that we didn't specify a team separately. As you see, rather, we have added this theme details at the beginning around until here. So the 20s, the jazz night photographed in the '20s and the US, we basically tell this twice. Us in America. And everything else. Starting from here. All of this part of the Prompt act only as an enhancement to this very first part. We're specifying a very detailed way, the aesthetics that we're looking for, and what kind of defined to Style. Okay, Now let's try one more option. And this is where my favorite things to do as well with Midjourney, which is food photography. But we're prompting for is a grilled tofu with microgreens. And then everything else starting from starting from this part over here to the end is just a description basically to the aesthetics of what we're looking for. We do not specify that it's a photo of Grill tofu. But we do say multiple times here that we say photography twice. So that will really, that will lead to more photorealistic images. We also explained it, that'll be one natural lighting. We explain that we want post-processing, elegant and eight Kenya. The AKS stands for anything K resolution, usually images and the underlying dataset with high resolution and with higher-quality, they are being tagged with such keywords. So they're usually result in better quality images. Let's go ahead and Upscale and looked at it a bit closer. So it's definitely quite a high-quality image. Almost impossible to distinguish between the real photo and all those keywords that they adapt, which is very cool, very visually appealing. We're even specify things like eight K photography multiple times that again, add to the Photorealistic nature of it. So those were a few examples of how to enhance the quality and create Visually Appealing scenes by repeating keywords 35. Crafting Visually Appealing Images Part 2: All right, this is the second part of the video where we continue to enhance our prompt to generate higher-quality and more aesthetically pleasing images. Let's continue in discord in the Photorealistic channel. And let's actually, let's actually competitor. That's prompt for a more simplified photo. Okay, so first we want a photo of a young woman holding a citrus near face, a very detailed and eight K. Those are just a couple of keywords that we want to indicate that we want a higher-quality image. Of course, we have to understand that in real life, this does not generate an eight K image or a higher-quality if we wouldn't dedicate that. However, this does affect on Midjourney, on how it searches for those relative scenes in the actual dataset. And only picks the ones that are tagged as very detailed or an eight K. Okay, now let's go ahead and Upscale. Wow, nice. I mean, these are really good results without even putting any additional prompts into it, which is what we're gonna do, right? However, it might not have been while we were looking for it. We might have wanted the image to be more suitable for a magazine, for example, or a bit more creative. So that's quite standard with some orange trees in the background. But let's see how we can use extra keywords to affect the result that we want. So first of all, instead of the woman, we add goddess here. Okay? This should influence how the woman looks by giving it more like, more God-like look or goddess Type luck. Then we'll also specify that it's aesthetic. Instead of very detailed, we say hyper details. Then we'll also add create a photography and a couple of extra keywords of the magazine we want, like Lightroom, which is basically stands for Adobe image processing. That is usually used to apply filters and balance the overall look and polish it up. So we want it to be that type of image. The clean flow should give the image a clean and more liquid look. And the featured on Unsplash should be able to kind of up to the quality of the image as well. Unsplash is a large CTL with all kinds of photos and images uploaded by society and then rated. And usually the ones that are being featured are very creative and interesting and of high-quality. So those are the keywords that we're going to add in here to make this scene where Visually Appealing and more photorealistic. Okay, let's try that and let's see what the difference will be between those two Prompts. Nice. This is looking a lot better. Let's Upscale a couple of options. As we can see, the scene became much more interesting with more details in it, a bit more aesthetically pleasing background. Will we can notice immediately that it's a more polished or a balanced image. Others, they match very well together. We can compare it with this one as well. So that's quite basic image. Let me see here. Yeah, This one looks really good for a magazine type. This is definitely more rid of more modest like and it's more interesting overall. There's more happening of the image as well. And we've got a little bit more diversity as well, which is nice. But the main point here is it's a lot more aesthetically pleasing and Visually Appealing, right? Even if we just looked at the four options to the grade fever. And if we compare that version to this version here, which is quite basic, especially the bottom left photo. We can technically say it's a bad quality or anything like that. But definitely all those keywords add to the scene that we have here. However, which you can see here, that we did not really describe the scene. We focused more on the actual aesthetic aspect. And the rest. For example, featured unsplash. This was picked up and basically randomized. And other images that were ever featured on Unsplash, even with different topics influenced this scene. If we would, for example, pickup Something else, not the Unsplash, but for example, some other famous portal or Gallery that is focused on scenes of specific scene, then that would influence our scenes here. Even though we went in specify, for example, some specific scenes that the Gallery specifies. The Gallery really, really influences the photo. Now, it is a good idea to obviously Describe it here. Or will we did in the previous Prompt with the tofu, with post-processing Lighting elegant eight K. To get those photorealistic images that you want. Obviously, you shouldn't go wild, but you should add in those 456 words, which is helpful. Okay, now let's explore one more photorealistic seen. Perfect. So again, just stopping paste this from the PDF file that I've provided so you don't have to type it in here, but will be prompted in here for is an all white Fine dining in busy Michelin star restaurant in Monaco. So all white. It should give us a very bright image with bright white surroundings like white clothes, furniture. And then we've asked for Michelin star, which stands for Fine and high dining. And then we asked for Monaco, which is a luxurious destination. Then we also add extremely luxurious and elegant seen in a white Theme as well. So we emphasize a white very much, of course in very detailed light room, clean, flow, hyper-realistic, and Bu County. One extra thing we have here is a lifestyle that shouldn't affect the image though. But let's see how it looks like. Now, let's Upscale a couple of Versions and let's check the results. Nice. So we definitely see the restaurant is all white, everything almost as white, even the flowers. We can also see the French southern coast influence here with the palms and of course, the luxurious touch. Let's check the second same here. It's, has HIV. As a fine dining type. Everything is white, even these LEDs or the Lighting is white. So I was expecting, and I wanted to see these kinds of aesthetics and it's not that difficult. Again, with the things that we've learned here. You see to explain and find basically the three words that resemble this. Once again, feel free to check out the Gallery page on the Midjourney website to get some inspiration and to teach others how to do it as well. Because you can actually learn how others do it in that Gallery page. Now, we have a bonus here since we have still time leftover. Let's try the final prompt for this video. Actually the final prompt. But this time we will not be asking for a photorealistic image. Instead, we want an illustration, vector Style illustration of three sweating children playing on the ground, sizzling street, extreme weather, heat. That's basically a 2D vector Style. Now let's prompt for that. And let's also trigger a singular Prompt, but without such description. So we say that we want an illustration, not a photo, but we do not really describe what type of illustration that we want. Is that a vector or any other details? And let's then compare. The first enhanced version is ready. We can see that. We can see the orange like hot and red colors, which come from this extreme heat. And we see 2D or vector Style illustrations everywhere. Now let's wait for the other prompt to finish without such descriptions. Perfect. And as you see here, we've got some completely different results. This is just to showcase how important it is to specify all those important keywords. Describe the vibe, Mood, Theme, and aesthetics that you're looking for. 36. The Use of Re-rolling: In this video, we'll learn why prompting for extra Options or Re-rolling and a V rolling is important and how it can help you achieve the perfect result that you're looking for. Let's continue with our photorealistic channel and let's repeat some existing prompt that we've used before. For example, this one here. In the last video, we learned how all the extra keywords and affect the image and uplift the aesthetic feel of it. But what if we actually didn't get that exact idea? But we're very close. So we don't need to modify and we don't want to modify any of it, but we want to see and try to find a, we can vary a little bit to get the perfect results. Because sometimes people tend to modify the prompt if they're not unhappy with the scene and try to add more objects that they wanted, C and Et cetera. But we know that this can also affect other parts and the other aesthetics of the image. For a simple example, if somebody wanted to see more cold colors and the image, they could, for example, add a keyword like a cool white balance. But that would definitely change and affect the composition. Also. Again, the more you add, the more you modify the initial idea. So let's imagine that if we had the Style and aesthetics aren't 100%, but we can't see the exact seen it that we want. Just use this Re-rolling button. Basically, sometimes I hinted like 45 or six times. What this does is it just reruns the prompt again with a different seed. It's really nice because it actually takes that and really create set. Now, we rolls take more GPU time, which is a bit annoying, but that's fine if depending on what subscription you with. What I've done here is I've requested multiple versions, I think for extra Versions. And let's wait for them to re-render. The next time with them again, we've used both the Re-rolling and the viewer role Options. Now let's observe the variance. We have them already. Wow, so that is the first alternate version. You can see a lot of yellow color in this one because that's the one that we'd be rolled. And for the second variant, this looks very different than the previous variant. However, we didn't change anything. It's the same very prompt. It's just an alternative. Now let's check to more. And maybe you noticed that, for example, these four, they look a little bit similar in terms of aesthetics and overall Style. Same here. At least for, at least for these two. Especially here. Even the clothes and all of those ones look similar. So that's why it's important to make an option, an extra option, if you like the overall aesthetics and the scenes. But if you want to enhance it and you want to see different options. Okay? Yeah. And here again, we see some correlation between those two and those two. So the second and third here, between the first and third year, capable B-roll and Re-rolling is a really great way to do it. Now. For example, let's say for me like the fourth one here, Let's Upscale it. And let's say that we were looking for something very, very close to the Upscale aversion in terms of like taller, the clothes, the scene, and the overall mood. So we've managed to hit it. However, we can't really use this image because we haven't artifact here. Because clearly we have an extra hinder, which is funny actually. So that's basically just some random noise that's been added to this photo here. To fix that. We have one option here. Without the use of any external software or Photoshop or anything like that. We can Re-rolling this version multiple times. So after you Upscale it, you can select, make variations or We can go to the v4 option in the grid photo Style. So we want to see Versions of that. We can try this also will now see eight each alternate different Versions, little bit different versions, each one. Alright, now let's check them out. We can also see that the face is slightly different. There are others small differences here, but overall the look and feel are the same. So will we also notice are the hands there are in a different position? Let's check the other option here. It's really, it's really similar as the same here. It's nice being able to very specific image. Here, we can clearly see that there are extra hands or fingers. Now, will we need to choose here, let's say the perfect version. If we still didn't get that, you can choose more Versions and make more variations. But let's say we like this version, the bottom left one, the most. Let's Upscale that one. And let's look at that a little bit closer because it has the strongest picture of what we want. Okay. It looks like for this one, all the hands are okay to it also very good lighting as well. And it's obviously good at the hands are that there's no third hand. We also like the face of the model more or etcetera. Now, Will we can do here is we can make variance of this one and go deeper. So it's one step further. And now we make a variant of the variance and we keep enhancing. So you can basically do that multiple times to get to that photo that you actually want without even prompting. You can keep doing that again and again and again until you reach the very perfect results. So if we compare again, it was this one on the bottom left. We can clearly see there are more lights here. One of the orange is hot because the rest are not caught in the different photos. Also the decorations here, they look a little bit different, more detailed, more tiny details as well. We really liked this one the most. Now, if you see here, we have an all Versions. We have cut orange areas. So these tiny details from that variant migrated to further different variance. Now let's imagine that we want to select this one because it has even a higher light and also more lit and also to oranges instead of one. So let's go ahead and we will that one. So again, you can keep a View rolling and adjusting and View rolling and adjusting as many times as you want. Which is very cool. Perfect. So that we bought with our third V role. Now would say that the first variant looks the best to me. So what I wanna do is I just want to Upscale that version and then I receive the perfect variant for what I was looking for initially. So the idea behind here, and I just want to say in this video is keep a Re-rolling and V rolling images, focusing on the differences between them and then picking the variant that takes more boxes depending on what you want. So if we're on a hunt for a lighter, little bit lighter image, more details, maybe with more specific posture and etcetera. Then you can just keep a Re-rolling and selecting the image until you get the perfect result. And it's a very, very good strategy to find that perfect result instead of giving it more Prompts. Again, artists do this all the time with their own images. 37. Exploring Other Versions of Midjourney: v5.0 and raw: In this series, we will explore older in previous versions of Midjourney, compare them and check their use cases in our base. Now before we jump into discord, Let's go through the official documentation and see what Versions we have available now. Alright, and we Lense straight into Versions. So as you can see it on the time of the recording of this video, the latest and the default version is 5.1, but all the previous versions and one alternative. Let me show you old niji is available right now. So in order to choose a different version, you need a specify this parameter, double hyphen V, and then the number or the name of the version. So one model or version that I find useful now, apart from the thigh version, which is the best so far, it's version four. Let's read the description of the version five. That's actually so that was the first version 5.0 basically. And the main difference to the previous generation, version four, was that this model requires a longer prompt. So by default, it would produce quite basic images, even with more focus on photorealistic Content. Still, if you want to enhance the details and actually receive a more interesting scene with more details in it and etc. you would need to express that in your prompt. But the version four didn't really need that because it would be generated and randomized by default. All you needed there was to say what you wanted, for example, some object and a photorealistic Style. And then just a couple of keywords. And then the scene will be interesting by default. Let's actually see the version five from one of our previous examples. Okay, Let's check right here. So that's a perfect example. Of course, version five is more tuned to producing a more accurate images or objects. For example, hands. Version four was struggling a lot, especially with fingers or some other objects, especially where you would need an anatomically correct scene or an object. So that's the main difference in requiring longer prompts and basically providing more correct images. A little bit more aesthetic. But if we can also say that it's a little bit more boring than the version four. Or let's say less creative. Let's check the latest version description. So the main difference is that as the version 5.1 has a stronger default aesthetic, which makes it simple to use with a simple text Prompt or short prompt. However, it's still in terms of creativity, far away from version for where it was producing quite interesting images, even with a very basic Prompt. But the main part here is that it does produce a lot more accurate images and a lot less artifacts or unwanted particles and details. So there are a few other differences, but that would be the main one. Now, one cool feature that the latest version has is that you can remove all the aesthetics altogether from the result by adding a simple parameter, Style raw. Now let's try that real quick and let's see how this affects the result. So we take the very same Prompt, but the one before the Enhancements that we did in the last video. So that would be just a young woman holding a citrus near her face. Very detailed. But we add a style raw. Now, we will also compare. I will actually roll Re-read that version. So I don't always have to scroll up. Yeah, that was it. Now let's check the Style raw. What it gave us. And while we can see, is that we don't have any extra objects apart from what we asked for and Style. If we compare it to the very same Prompt and same version, but without specifying that we want a raw Style, we see that the image has been enhanced by mid journey with random details that Midjourney found to be suitable for this type of scene. So basically the Style raw could be very handy for when you're prompting for a photorealistic Style, for example. And you don't want anything extra to be added in the scene. Now, let's take one of our previous examples where we were prompting for a ginger cat and a winter forest. Let's actually take this one of a ginger cat playing with snowflakes. So this is a photo. But if you can see, it's, it's very HDR and a very beautified photo. And let's now see how that would look. If we would supply a raw Style parameter to it. All right, So it's finished. And there isn't a lot of difference with this example. But still it feels like it was not post-processed or that it does not have or that it does not have any special effects on it. So the Style raw works very well, especially with the photography, especially with people when you don't have too many details in it because you don't need to. So try that, experiment with it. And let's also try and see how this very same prompt would look like on the version five, not 5.1, but just five. We do not specify that we want a raw Style while it's generating. Let's check this image again. So there's not much styling going on. It's really just a very high-quality photo. And let's just compare that again. Yes. And here you can actually see the color grading in the first and the fourth version. So they do look like a little bit post-processed, which makes them a bit more pleasing to look at, but also less photorealistic. In this case, because we didn't really emphasize or specify how photorealistic we actually warrant. Okay, now let's check the version five. And immediately you can see that the results are very different to version 5.1 because five-point one tends to enhance with the aesthetics, which is good. If you actually specify in your prompt the Style and the Theme, basically which will produce the needed aesthetics. But if you do not specify or don't want to specify this for some reason, then it's better to use the raw Style or even to try aversion five, which looks like it understands our Prompt literally. So we only asked for a photo of a young woman holding a citrus near face. And the very detailed eight K picture, sorry photo. So here we have an absolutely identical Prompt without any extra parameters. But we definitely see a lot more aesthetics and a lot more details in here then in the previous version. So try to play with that. And also as a recommendation, if you're, if you're prompting for photography style Content and portraits, then try version five or the Style raw on 5.1 38. Exploring Other Versions of Midjourney: v4: Well, continue to explore different versions. And in this video, we'll try version four, and we'll also find the best use cases for it. Let's go back to Discord into our photorealistic channel. And this time let's take an unusual prompt and will exercise on that ultra-wide cameras shot of the Mercedes-Benz vision EQ silver arrow. You can Google that model. This is a very elegant concept model from Mercedes-Benz. But you also state the future year and enhance details with a sunny day of photo realism, color grading, photographic, photoshoot, depth of field. So let's just trigger that and see how the basic version, the default version 5.1 would luck. And while it's preparing, compare that with version 5.0, which is just v5. Don't forget the space between the V and the version. And now let's also try a version for okay, The results are ready. Let's start with the default version 5.1. And we can see that all the models look very futuristic. And they do a little bit like the EQ silver arrows concept model. But because we also specify a special year, and we did say as a concept model, which should alter the original model. All this produces this look which I find a little bit similar if I compare all those, but also photorealistic and also the shape and other details from the background. They all add up and together it looks very good. And now let's compare it with the version four. So what you can immediately notice is that it has more styling put into it in terms of the post-processing, like a filter or color grading applied on top. Also the shapes of the car. They're a bit more extreme, if I can say that, and more brave overall. And let's briefly compare it to the version five. So version five does look different a little bit compared to 5.1 of shapes overall. And the styling is quite similar. So it really depends from it. If I would get to render something like this of futuristic car within this environment, then I would opted in for version five here, not the latest one, the 5.1. So let's Upscale and examine version for a little bit closer. Let me actually Re-rolling that. And I will also Upscale. And let's Upscale version 5.1. Yes. And the big difference between the version four and the version five at this stage, you can see that the upscaling version is being generated. So because version five doesn't have an app scalar now, it provides the version for when you click Upscale instantly because it cuts it out from the grid. Now, let's open this option and let's check the detailing. So what we're seeing, what we actually see here is a Mercedes Benz logo, quite a good quality background, very photorealistic. I don't see any artifacts though repeated patterns like the wheel in this pattern, it's quite accurate. Also on the back wheel. That's the second version of the second option of version five. We can also see the Mercedes-Benz logo. These are quite accurate details. With a closer look, you probably noticed that this is a render But it's very hard to spot. Now what's compared to the version four? And it does provide a little bit of a cooler feel, if I can say that. So it does look a little bit cooler to me over. However, if we pay attention to the little details, we will start noticing, for example, the wheel pattern. It's a little bit chaotic. Also, we don't see any Mercedes Benz logo. There are a little artifacts like this man here. This of course can happen with version five to, but it's more likely with the version four. So here we got a more accurate pattern on the wheel, but not on the front wheel. And also some artifact in this area. Let's actually open this in the browser and zoom edit. So this part, for example, looks like it's hanging in the air and etcetera. But overall, I very much like the styling of it, especially in terms of color grading. And also the shape is a bit more brave of this vehicle. Now let's check another version of version four. We can also see here a lot more diverse shapes of the car. Those are all very different. And also a lot more brave, especially this very into, for example, let's Upscale and look a bit closer. And let's check the version five again. We can see here that even the shape of the front bumper, they're all a little bit similar. The top, the roof part, and the sides also. They are of course different, but they still look a bit similar. Whether inversion for those are all very different styles. But of course less accurate. Okay, Now let's open this up in the browser and look a bit closer to the version four. And we can start noticing some artifacts like this. The car looks good. However, still the button on the wheel is not very aligned. It's still a little bit chaotic. And also I can see some flaunting article right here which is not attached to anything. Also know Mercedes. I don't think we even got one version that has a clear Mercedes Benz sign on it. So version four is not very good with some specific details, but it does give a better luck in some scenarios. Now, we will try one more different Prompt with version four. Let's take this as an example here. And the idea of this video and the usage of Versions, especially previous versions from the aesthetics that you're looking for. So the previous version doesn't necessarily mean worst version or less advanced. It's mainly, it mainly has different aesthetics. Of course, with every version. Midjourney generates more accurate images, especially in terms of anatomy. But anyway, feel free to play around. But definitely try version for, with different styles like paintings, especially watercolor or oil paintings, sometimes work very well with version for. Let's just check the result. We can also see that the result is very different to the version five, but still quite aesthetically pleasing. And also, we don't see any errors, right? So feel free to experiment with different versions. And in the next video, we will learn how to use the niji version 39. Exploring Other Versions of Midjourney: niji and more: We continue to explore different Midjourney models. And this time we try the niji version. Before we practice it. Let's check the documentation once again. And we'll go to the User Guide Version tab. Scroll down a little bit. And we see the niji model. And what Midjourney states is that the niji model was tuned to produce anime and illustrative styles and with anime aesthetics. So mostly it was trained on the action Shot and character focus compositions. But of course, niji can't produce a very diverse and different results in different moods. Now, to use the latest niji version, we would have to specify niji 5.1 additional thing that we can also optionally specify a specific style or theme image by submitting the perimeter Style. And then either the word cute, scenic, or expressive. Let's see how that would look. So that would be the default niji version without specifying any Style. It does have a little bit of anime vibe to it. But basically it's a painting, painted image. Now, that will be the cute variant. That is the expressive. We see some more eyes and some emotions here. Also the colors and the Composition are different, more detailed and the scenic Style. So it has a bit more dramatic scenes, also, both colors and also quite detailed. Let's go to Discord now. For this purpose, let's choose the styles channel. And let's start with the First Prompt. Which would be an anime girl with purple hair and Japanese traditional dress running through the castle with ghost chasing all around. Let's trigger the product. And let's now examine the result. So I liked the quality and we really did receive an authentic niji style illustration. Let's Upscale one. And we see that the details are quite accurate. We don't see any artifacts and the overall aesthetics are on a very high level. Now, let's try to change the mood of this image a little bit by supplying a Style. So we don't describe this within our Prompt, but rather use the Style parameter. Now let's try the scenic Style. And let's also check out the expressive style. Then let's see the difference. Okay, So this is the scenic Style. So it has More Actions. And the scenery itself is a bit more dramatic. And we can actually see movement, more action in here. Let's check the expressive. And here we can definitely see a lot more emotions in all of the options. And what I've noticed is that they actually look quite different in terms of the aesthetics and styling. And the reason is, why wouldn't we just specify that in our prompt by specifying that we want to see that in this particular style. It's because those styles were actually trained separately on the model. So given the underlying data, it's different. So depending on what you're searching for, if you want a more dramatic action or action in the scene overall, you like these kinds of aesthetics? Then there is a dedicated Style for this, which was trained separately on these kinds of images. The same goes for the expressive and the cute. You can try that separately. Now, let's try one more prompt. So that's also a niji five. But we don't specify that we want an anime style. We only specify wilderness scenery, giant flower's petals, pink and white color theme. And let's add the word Painting. So here we don't want to see any anime or Cartoon Style, but we do want to use a niji engine. And I've been the aesthetics of it to be precise, to generate the painting. We will then compare how would a version five, the plain version, would generate if we asked for a niji Style painting. Let's check it out. So we didn't specify any particular style. We just specified quite particular scene. And it gave us this result, which looks quite specific and the aesthetic. But that aesthetic differs quite a lot from if we would render that with just version five. And let's try that. So here we don't specify a version. So this means that the version 5.1 will be used, but we do use a Mulltiprompt and we specify that we want a niji Style painting. The rest is the same. Now let's wait for the result. So this is why we've got, and I find the results also very beautiful, but that is not 100% authentic niji Style, I would say, if we compare it that you see that this resembles the style we're looking for much, much closer. Feel free to experiment with different scenes. Niji model works especially well with any anime style or Cartoon Styles, scenes or paintings in an Asian aesthetic style. Now, I will show one more use case for Adrian. So we want ballroom chandelier is in watercolor style. But if you see here, we actually want to get a Tile. And we also increase the stylization a little bit. We're still using niji. But let's say we want to get a seamless pattern out of this in a niji Style. Let's see what the result will be. They be, if we take a closer look, I can already imagine a very high-quality wallpaper on the walls with this texture and that would look fantastic. So that's just one of the many examples how you can use this style. We'll get back to the Style later using the niji model. For now, just feel free to experiment with this and check if you can actually find any use case for yourself. Now before we finish with different versions, I want to briefly show you the legacy models. We're not going to explore them in this course, but feel free to experiment. First of all, what version four, you can also provide a style. And style can be different variants all over the different versions and sell for a foresee or for B, they all produce a slightly different result. So that would be for the Fourier, for example. And the same prompt for B. As you can see, it's the aesthetics and the styling is different. So try to experiment with that. And the very early versions, especially the third one, it could be useful if you want to create some Painting, especially if dad is some kind of surrealistic or minimalistic painting style image. You can try that for this course. It's out of the scope, but it's a good idea to try all of those versions just to see how Midjourney has advanced in the recent year. Actually in the recent two years. Okay, So that's it for this topic. And I will see you in the next one. 40. Image Reference and Blend: Now we will learn the basics of how to use image as a reference for Midjourney and how to blend two or more images together to create something really unique. Let's go to discord. And for this purpose, let's create a separate channel and let's call it image Prompts. Before we start, let's briefly check the documentation and in the user guide in the advanced Prompts section and Image Prompts, you can find the full explanation of how this technique works. And we already know that links to images, they always go in front of the prompt and then Parameters follow at the end. It's important to know that only those five or four months are accepted. So that's a PNG. And JPEG. I will show you how to get the link for the image. Before that, I just want to show you quickly how that works in theory. So for Midjourney version five and actually the version four will be the same. You're basically providing an image, for example, you can provide to images. In this example we're providing the statue and then flowers. And this would be our result. So those two scenes are blended together. Also, the style has changed a little bit. So to styles are also blended together. And instead of providing to images, what You can also do is you can provide an image and a prompt instead of a second Image. And those would also be blended together. Now let's go to the image Prompts chat. The first way to use this option is to upload images into your channel, into the chat. So you click this button and you select Upload File. Now, let's upload those two images. When you select hit Enter. Now let's try to blend this image with this one. We can see that both images have approximately the same style. And all you have to do is also trigger the imagined command. But instead of adding a text prompt, we'll provide links. Just right-click the image and select Copy link. Now just paste that, add a space and then do the same with the second image. So we'll copy a link, will add that, and then we hit Enter. Now let's wait for the result and let's examine it. Let's Upscale actually one of those. So as you can see, we have the aesthetics and objects from both images that we've provided. This object here looks like a ship, but it's very surreal. Which is the mix of those two images. So the one, the first image basically is black and white with gray color dominant. And in this image we also see that it has a lot of black areas, especially on the bottom and on the top, and also bright orange yellow circles. So that's that's what got here. But if you notice there are black areas on the top and on the bottom. So try to experiment and get the best. It's recommended that you use images with the same Medium and the same style. So if that's a photo, tried to mix two or three photos together, or if it's a Painting, try to mix the same style. Alright, now let's check the second way of how to provide the links or References. So if you find an image online that you want to reference to. For this purpose, we will get the image from the actual Midjourney Gallery. You can try to get links from different websites too. This doesn't always work, but depending on the website where you get the link. So let's check the Gallery Let's take, let's take for example, those two. So we'll take a burger just up new image, right-click and select copy image address. Now let's go to discord. And now let's provide the second link. For example, this one with the Image Address, paste it and trigger it. As you see, we do not actually specify anything. So those are just a few links. No Prompt, no text, no parameters. So this is a plane Blend command. You can actually blend more than two images. Version five supports up to five references. Okay, we've got our result. Right. So those are quite unusual looking burgers, though, we can definitely see that those two scenes are being merged together. So the forest scene and the burger in quite a funny way. Let's scale and look a little bit closer. And why you see is that one to image. The second option is the only option that was generated without any black areas. And this option and the other options too, they have those black areas here. Now, the way to minimize this effect always try to select images with the same dimension and the same aspect ratio and the quality. It also affects the final image. Now, there is another way of how you can blend Images. There's a shortcut or a shortcut command called Blend. So if you trigger that, you can choose and upload images here in those windows, you can actually upload via just by selecting image 34 or five. This is just a nice interface for Blending images together. There's also a handy shortcut called dimension. Basically this just adds and Aspect Ratio at the end of the prompt. Let's try that and let's select Landscape. And while it's rendering, forgot to mention that in order to minimize those black areas, it is important to select images that are similar in terms of the Aspect Ratio and dimensions. But apart from that, It's also important to request the Generation with a similar Aspect Ratio. That way you can also minimize this effect of black areas. So this is our Blend command. As you see, this just converted two links, the same as here, where we triggered this manually. And also an aspect ratio parameter appeared at the end. I can already see that there are some black areas still present. But most probably because the aspect ratio of those images initially were not more of a three-to-two, but 16 to nine. So feel free to experiment with that. So this was a very quick introduction of how to use image prompts and how to blend two images together. In the next chapter, we will explore this topic deeper. We will mix the actual Prompt together with an image. And we also learn how to weight the image against the text prompt with a dedicated parameter 41. Remix and Describe: Here we explore a couple of other handy tools like Describe command and the remix mode. Let's go to discord and to the test channel. Let's start with the Describe command. But before we practice it, Let's go to documentation into the command list page, and then we choose Describe. So as you can read here, Describe command basically works as a reverse engineering. So instead of providing the text and receiving the image out of that text, it works the other way around. So you provide an image and mid journey reads that image and it tries to provide a suitable prompt that would generate a similar looking image. Now, you can read a little bit deeper in the FAQ page on the mid journey Discord server. For now, let's learn, Let's practice how this actually works. And let's find some simple image. Let's try photorealistic. For this purpose, let's choose a cat. That's quite simple. So a ginger cat, we see a winter and a forest. So that's not too complex. So we save the image. Now, let's go to the test channel and type forward slash Describe. And then you will be prompted to select a file. Let's choose the one that we just downloaded and hit Enter. So Midjourney takes some time to think, and then it provides you four different prompts that could potentially generate a similar looking image. Let's read the first one. So it's a cat in the forest with snow. In the style of digital manipulation. We have a light Setting, we have a camera model, and some other keywords that help to describe this image. Of course, this will not generate an identical image. But let's try a couple of Prompts. From here. We don't specify any parameters. Let's try the first one and the last one. Okay, Now, let's examine the results. So as you can see, the results are very similar to the existing image. But we didn't provide any reference image. Instead, we explained in a text format by getting a description from mid journey for that image. So the second version is a bit different, but we also kind of get the same Mood and the same scene. So that's also a forest of snow and a ginger cat. So this can get very handy when you have some picture or photo and you want to generate a similar looking scene, then you can just simply try to describe that image. We can also add a few extra keywords that you think can help mid journey to generate the needed Style and change the aspect ratio and etcetera. And another trick would be to mimic the style and the image is that you can use a prompt that you've received from the Describe command. But also what You can do is you can use the image prompts, so you can add additionally a reference image. This way, you're actually providing to Midjourney, not just the explanation, but also a reference. And that together usually gives close looking results. Let's try that. Let's get the link. So we add an image link And then the prompt. Alright, now let's Upscale some version to look and compare a bit closer. Now let's take our reference image. So that's the initial scene that we provided. And now let's compare with what we have. So this is very similar looking. Of course, it's not identical, but it's really, really close. And I find it a little bit more of a closer look. Then when we prompted just what the text description without any reference image. Because those options did look a little bit different. So here in all four options, we actually get a very close looking image. So that's how you can use Describe command to generate a text prompt out of an image. And also tried to provide a reference image alongside the prompt that you got from the Describe command. That way you will achieve the closest luck. Now, before we move on, I want to show another handy feature of Midjourney. And that is a remixed mode. But to trigger that remix, actually it's called prefer remix and hit Enter. And now we have the confirmation that the remix mode is on. Now let's quickly read about it. In the advanced Prompts section, you'll find the remix page, and here's what it says. So remix mode can change Prompt parameters, model Versions, or aspect ratios between the variance. So the way it works is that with the remix node, mid journey takes the initial prompt, like a reference when you trigger any versioning. And during that versioning, you can actually affect and change the Prompt or lower the Aspect Ratio or something else. So here's a good example. So if we initially prompted for line Art stack of pumpkins, and when we trigger a versioning, it will prompt us if we want to make any changes. And if we change that completely, we'll say it's a pile of Cartoon owls. That's what we'll get. And we can see clearly that the initial image has an influence on the final image. Or we can just simply change the style. For example, if we had line Art initially, and then we can change the style or even a model. Now, you can read a little bit deeper on this page because there are certain combinations of parameters that you can or cannot use with this mode. For now, let's go back to our test channel. And now let's trigger a variation. So we can trigger just that. That is an Upscale, or we can just trigger here, which is the same versioning the Re-rolling. This is also a V rolling of that option, our first option from here. Let's press that. And as you can see, we're being prompted. If we want to add any changes for the test, Let's make a small change. So we remove the Camera Model, which gives the Photorealistic look. We don't have any other photo ray tracing probably. We can also remove and we already have storybook illustration. And let's just add here one word like vector illustration. And let's see how that would work. Now let's check the result. And as you can see, we've got the same scene. However, this doesn't look photorealistic anymore. It looks more like a drawing, like an illustration, which is exactly what we've prompted for. This comes in very handy when you just want to change a small detail and you want to perfect your result. However, I personally prefer to leave the remix Modes switched off because if I want to change the Medium or something, I just trigger a new generation. Basically I just changed the prompt and try it again. But sometimes when you don't want to change, for example, you've got a perfect seen here, and then you realize that some object is missing. Or you want to change the Medium or enhance something, but you want to keep this very senior. So the cat looking from left to right with some forest in the background and stuff like that, then remixed mode could be very useful. So in times like that, just trigger the remix mode. Apparently, I will just switch it off. Then try to generate an alternative version. Then you can switch off the remix mode. Okay, Now let's move on to the next video. 42. Extra Options and Commands: In this video, we will check out the remaining Commands and functionalities of Midjourney that are not directly related to Image Generation. Now, let's start by going to the documentation and command based page. So I advise to go through every single command to understand all the functionalities that bit journey offers. Also, if you see that there is a dedicated page to this command, please go through the explanation in detail because you might find a very handy command or even a combination of those that can help you establish a very efficient and production workflow with Midjourney. Now let's check all the commands briefly. So ask command is very simple. It just works as a chat bot and tries to find a relative answer to a question. Very simple. For example, you forgot how to specify and Aspect ratio. Then you can ask how to specify Aspect Ratio. And you can get some answers like that. It can also sometimes even give you an example. And the link to the documentation, the Blend Mode we covered in previous videos, the daily theme. It's just a command to subscribe to a challenge. You can find the daily Theme channel in the mid journey Discord server to get an idea of why it's about ducts works as a similar ask or FAQ command. To get some help with Commands. Describe we've covered Fast or Relax Modes. Those modes are used to switch whatever mode you're working with. Now, Info we've already checked, but let's briefly check that again. So this basically gives you a full information about your account when it will be renewed, what mode you're working at, how many hours or Fast hours left on your subscription. Also, if you're generating a lot of images, you can check the cues. Sometimes images can get stuck in the queue. You can see it here. If you see some big number or more than five, and it doesn't change for a long time. Then you can ping the moderators in the member support channel on the Discord server. Now those two commands are not available to. They say Can standard plan subscribers. But those basically change if you want to hide all your images generations from the Gallery. So there'll become a hidden from the public. If you have a Pro plan or higher than you can trigger that command. And if you want to switch it back to public mode, then you can trigger this command. Now, the Settings command, that is quite important. Let's, let's, let's try that. So this is basically a shortcut, different modes that you want to work with. So by default, we have the latest version selected, Midjourney, five-point one, but you can select any other version that you would want to work at by default. So you will not need to specify any parameter at the end. Also, you can specify the styling or stylization mode. At just the Fast or remix mode. You can also reset all the settings back to default. And you can also see if there's any suffix attached by default. Now let's check what that is. With prefers suffix or prefer option list. So in the Settings and presets page, you can read a little bit more about this. Right now, let's just check how you can specify some suffix that will go at the end of each prompt automatically, or how you can even create an option list and then select that as your private parameter. So let's try that. We want to trigger this command prefer suffix. So we'll choose this one and we choose the option new value. Now here we can provide a set of parameters that we want to be added automatically to every prompt that we do. For example, we only worked in an Aspect Ratio of 16 to nine. Now, if we trigger a Prompt What You can see is that the Aspect ratio has been added automatically. So this kind of creates a default Aspect Ratio for us. Mid journey has its square, but we prefer to have it Landscape by default. Now, in case you want to remove or change in Aspect ratio for a specific image or change, for example, if you have any other parameter and you want to override that, or for example, this particular prompt. And what You can do, You can specify the Aspect Ratio manually. For example, you want a tall one, so two to three. And what you'll notice here is that you will receive actually to parameters of Aspect ratio. However, Midjourney will only pick up the first one. So that's the example. So you still have the 16 to nine aspect ratio by default, we override that manually and we got a tall image. The first generation seems to be stuck a little bit. Now, if we trigger that again buffer, we can then basically override that. So we're going to change, for example, the list of the parameters that are attached automatically. This is our new suffix. So that's how you override that. If you want to completely remove that, then you can just simply in the Settings menu, you can just reset that. Okay, As you can see, the suffix has been cleared. Now let's try a little bit more of an advanced way of using this. We can also create a custom option for, let's say, a custom parameter. The way it works is that you trigger prefer Options set. They provide the name, then you provide the values that you want to add. Those values will also be added at the end. One important note here is that you can only add Parameters. Those options also in the suffix. Now let's try to create a custom set of parameters. So we select prefer option set. We'll type in the name here. Let's call it PR, which will be short for portrait. Now, if you click here a little bit to the right, then you can also prevent the value. And what we want to create is we want to create a set of parameters that we usually use. For example, in this case, degenerate portrait, photorealistic portrait image. So we would want to specify and Aspect Ratio by default, which would be four to seven. Then if we were prompting in, if we were working with model five, then we can also provide the Style raw, which works very good with photorealistic images. And then let's say we also want to exclude by default, for example, a green color because we work in a certain aesthetic. Now let's try that. So as you can see, we've just created a custom option PR with this set of parameters. Now let's try to trigger a portrait of Photorealistic one. Let's take as an example, one of our previous Prompts. Let's take for example this one. So this **** would look like without that set of parameters that we specified, only the Aspect Ratio here. Now, let's do the double hyphens and set PR, which is now our custom parameter. So those should be added now by default. Now, let's wait for the result. So as you can see, all of those Parameters got unfolded. And now we've received a specific style that we were looking for without specifying all the parameters manually. Now, you can always list all of those parameters. Prefer option list. If you ever forget what parameters set you have, or you can also modify those are managed through the preferred option command. For example, here the intelligence gives us per Options set and we can modify the value here. Now, before we continue to the next chapter, just a couple of tips I want to give. So what this button here, you can always favorite the image. And then in your Midjourney Gallery, you'll find that image under the favorites. You can also do that with an emoji. It's called a star. So if you star that, that would also trigger Midjourney to put that into Favorites on your Gallery. And another trick is that in case you've received some image that you don't want to appear anywhere and you want to actually delete that. It's very simple. All you have to do is also attached the emoji. This time that would be a red X, Red Cross here. So if you choose that as you see, the images instantly deleted and it should also be gone from your public Gallery page. Okay, So that's it for this chapter. And I will see you in the next one. 43. Introduction to Advanced Techniques: This is another highly significant practical chapter, where we concentrate on mastering photorealistic Content. Explore professional photography Techniques and fine tune our results to achieve the precise outcome that matches closely to our initial idea. We will learn how different Cameras produce slightly different results and aesthetics. We will try different Film Effects. We will see how different lenses create a certain type of VU angle and effect the scene overall. Then we'll also explore different shot types. We will also go through important topics such as Lighting and see how to best and where to apply cinematic volumetric or other different types of lighting. We will also explore specific environments and their impact on our scene. Then we will begin Fine Tuning are Prompts to achieve the desired image outcomes. We will also use Image References, and we'll also try to merge a few Prompts, including image Reference and single result, by using the Mulltiprompt technique will learn important keywords that will improve the quality of the result when prompting for people in Fashion, Interior Design, and many other specific topics. We'll also learn that mid journey is able to produce well-known public figures from the present and the past and also popular places. Then we'll take a look at the very advanced and complex prompts, which helps to control the resulting image with a very high precision. We'll also dive deeper into very specific topics or seem Types so we know where to apply certain keywords. So overall, this chapter aims to provide a comprehensive understanding and practical application of Photorealistic Techniques and specific content creation. Right now, let's get started. 44. Advanced Techniques: Cameras: Now we're beginning to develop our photorealistic Prompts with the goal of achieving results that are indistinguishable between real photos and generated images. In this section, we'll dive into the realm of professional photography and explore how we can apply specific photographic techniques to maximize the quality of our prompts. Let's begin by testing different camera models. For this, we will choose a photorealistic channel. Now, I have prepared a basic Prompt with the scene of happy people in nature near Wording tab. And we're giving the scene of vibe, of enjoying the serine surroundings and the rest basically we're just specifying the style in the medium that we're using for this prompt. So that's a photo. We're adding photorealistic professional photography featured on Unsplash. Now let's try without any Camera Model. So this would be standard current aesthetics for Midjourney. Let's Upscale a couple of versions. So this is the look that we get with version 5.1. For testing purposes, I will also try a Prompt with a raw Style. Because I can see here that there's some color grading and done post-processing Style happening. Right? So this is the raw style. There are definitely less filters or color manipulations happening in those. Let's Upscale one version. And now, while we also do, is now we tried to add a camera model. Let's begin with one of the most popular professional cameras, canon IOS. We do not specify the exact model. Now, let's specify a camera model. And we will try a canon, as. This is quite a popular professional camera. So we didn't specify any specific model for that IOS series because we really don't know if Midjourney knows one specific type of that series better than the other. So we will leave that for mid journey to randomize basically. But anyway, all of those, all the cannons, they still have a certain aesthetics feel to them. Also by adding an actual camera model. While we do is we help Midjourney create a much more photorealistic Content. Now, let's Upscale one image. And let's compare to this one which doesn't have any specific model. We can say that it's a very high-quality generation, or we could even say a high-quality photo. If we compare to this image, we can see a slight change in color grading, Lighting overall, and also the detailing. We can definitely see that this has affected slightly all of the options and definitely increased the overall photo realism and photo aesthetics of it. Now, let's try the same Prompt. But we switch the camera model from cannon to Sony. And we try the A7, our model. There are different versions of the A7, our model. So let's try it first without specifying it. And then let's also try to specify the exact model of that camera. So let's go with a Sony A7, R3. So the triple I stands for that. We should actually make it bigger. So capital I. We can also change the, this part. So photo taken with Adobe changed to Shot By or Shot on. We'll also try that in future Prompts But just be aware that the specifying camera might also affect the actual composition. Because for example, some Camera Models, they're used more for one thing, for example, indoor portrait photography. And some models they are more used, for example, for nature photography, etc. so by specifying a model, for example, that is more popular in the studio setting, then this could actually affect the image and the scene itself. This is just important to remember. Now let's check it out. To and not trained eye. You can't see too much of a difference. However, I can definitely notice that there is color grading that's happening in here. And that the neat shadows, there are a bit lighter, especially on those two. That's what I usually noticed on the Sony A7 R. Let's Upscale. And let's check the other version where we specified the exact nodal. And I can definitely see even more color grading happening. So it definitely has less contrast and colors overall, they do not pop so much. We can try to Upscale that, which by the way, to me, when the contrast is lower than usual and the colors are not overly saturated. This gives a little bit more of a photo realism. So I really like to add the Sony camera model to my photorealistic Prompts. I use it quite often. Of course, it depends on the scene itself. And by the way, this is green. This is another Canon EOS Prompt. And even if we compare just here, side-by-side, we can definitely see the effect and the difference. So there's a lot more contrast on the canon that goes on. And there are also, I would say, a little bit like an HDR with an HDR effect. So there is less haze in those images. We can Upscale and also compare that. So this is a good example. You're seeing. This definitely has less contrast in it and more post-processing, but looks overall very photorealistic. And this looks more fresh and a bit more color saturated. Now, let's try a couple of more models. Now. We try Nikon. That's another popular camera, and we'll also change a little bit here. So, so put on Nikon Z nine. And then we also prompt for another model, which is Fujifilm G FX. Alright, so the nikon is ready. Here we also see less contrast and a bit more of a post-processing effect similar to what somebody has. But still I find the aesthetics a little bit different. We can try to Upscale and see a bit closer. And this is the Fujifilm GI effects. And I find this something somewhere in the middle between the canon and the Sony. We don't have very bright colors here. We do have some post-processing effects happening, but not too much. But we still see a higher contrast than the Sony or nikon. Let's also Upscale and browse that. So that's the Nikon version. So this is a quite well balanced image. If that were photo, that would be a perfect photo. If we check the Nikon again, all of those are looking quite well balanced apart from this one maybe which is a little bit dark. And that's the Fujifilm. I also like it, but that would not be my favorite for sure. So my recommendation overall is to always add a camera model when you're prompting for photorealistic Content. And if you want some effect like a filter effect like a like on Instagram filters or a Lightroom effect. Then try to choose, tried to play with the Sony and the Nikon cameras. If you want more brighter colors and more contrast being used, the Canon EOS camera, of course, you can feel free to experiment with other cameras. But this would be a good start. 45. Advanced Techniques: Film Effects: In this video, we'll continue to explore Camera Effects. This time, we'll try different Film Effects and retro Cameras. Let's continue in our photorealistic channel. This time we have Fashion Photography. So we'll prompt for a tall image and this time will include the film. Actually for the first Prompt, Let's just do without to see how it looks without any Film Effects. So we have a Fashion editorial photography of an elderly woman walking down the Fashion walk in town. We'll also add Swarovski crystals and intricate textures and to glamor photography that should produce an interesting result. And while it's working, I will prompt the next option. And this time I include the film, which is the Fujifilm pro 400 H. Okay, We got our first result without any Film effect. The first and the third option, they look quite nice. Let's Upscale the first one. So we have everything we've asked for. Also, the quality is quite good. Let's open in the browser. So we see the skin wrinkles. So the effect, the quality is quite good here. Now let's check how it looks with the film effect. So to me, it does look a bit more photorealistic. We can also see the typical Hello Style grading for the film. Let's Upscale the second version, sanctum pair. So we see a little bit more blue in here, but this looks very natural. So this is typical for the Cameras of that age that we're using the film. So if we compare, this also looks a little bit modern. So by specifying this type of film or any other film that was used in the past. We also kind of hint to mid journey that we might expect a photograph of the previous decade. That's just worth remembering. Okay, so what I will do now is I will prompt three images in a row. And then we're going to, we're going to examine all of them. Okay, so we've got our results. Now. Let's examine it one-by-one. One change I did is I removed the 35-millimeter film. Usually it's a good idea to include that. But just for the purpose of this experiment. So we only see the effect of this particular film. I have removed it. So let's start with the TRE X 400, which is a black and white Style. Let's Upscale one option. Actually, let's Upscale. Why the option for each. So that is the black and white version, also very detailed with its own aesthetics. I can see that the time wasn't changed. So the dress of people around her is quite modern day. But the style and the look overall matches this type of film exactly. Now, let's examine the next one, which is an add glove, V star 200, also one of the popular films. So we can see a bit more of a vibrant image with more sharpness. Let's check go. Actually, it's still a little bit blurry. It's typical for the film, but definitely adds more contrast into it. And also its own aesthetics. And here's another one that's colored gold Let's open it up and we can see brighter and a more warm image. So that would be a warm color up, sorry, warm white balance. By default, for this film, this film has a warmer white balance. Also the shadows, they look a little bit brighter and they're not too much contrast here. Let's zoom in a little bit. And to me, it does look a little bit sharper than the other options. Of course it could be just that option. But anyway, all the previous versions were a bit blurred. Now, let's try and unusual and a special Film which is that usually filmed on. So we'll put it. So it's shot on a special plate. Daguerreotype. So basically it was used in the 19th century. So with very first photos. And now let's see how this would affect. Okay, So we have our result. Let's check it out. And this doesn't really look like a photo from the 19th century. Despite that we had a film specified or a plate that it was shot at. Now, what we can do to fix that is we can actually specify the time or a decade when this was shot. So we'll add a photo from the 19th century. So let's see how it will work with this when we actually specify the century. So while it's being prepared, let's take some example from here. Let's take the Kodak Gold film. It was quite popular back in those days. And now, if we actually intend to shoot some, some retro style to get the real vibe of that time, then we need to specify that time in the prompt. So something like this. Of course we could elaborate a bit more. Now, let's check this. The 19th century. So we can clearly see that at least two options worked quite well. This still doesn't fully correspond to this type, but it's very close. Let's Upscale. So we definitely see that this is not a modern day. The photo does give a few of the 19th century of something around that decade. Now let's check the photo from the '80s. And as you see here, now we have a stronger wipe of the '80s in that Film Style Picture. Feel free to experiment with different decades and different Film Effects. I would like to show one more example of how you can increase the film effect. Let's check the prompt. So we keep this decade wafer. It turned the 35-millimeter film keyword and then we added shot on a Laika M6, which is the film camera. And we also specify the actual film. So now this should have a greater impact on the overall aesthetics and the style of the image. Now, let's Upscale a couple of Versions to see it a bit closer. Let's look at this closer. And now we really see how this unfolded. So now we have the film, we have the camera, we have the decade, and everything is playing very well together and producing a very strong aesthetic feel. It is related, of course, to that age. And this film, That's the second option. And we see the same here, which is a quite nice result. So that's how it works. So when you combine the film camera together with the actual film, and then you specify, for example, a decade that is not necessary, but that helps to achieve the needed effect, then you can get a really nice result. And don't forget to experiment with different scenes, not just Fashion Photography. This applies to anything. And before we complete, just as a bonus, I want to try one more thing. So I want to see how that would look with a Polaroid camera. So that's a very specific style. Now let's Upscale immediately a couple of Versions. And this produces indeed a very nice result with a very typical feel and aesthetics of that time. Alright, that's it for this video, and I'll see you in the next one. 46. Advanced Techniques: Lense Effects: We continue to explore how to achieve a stronger photo realism in our images. And this time, we'll focus on different Lense Effects. Let's continue in our photorealistic channel. And we'll begin with a portrait lens, which should create a quiet, sharp image of a person while blurring the background. So we have a Prompt of a beautiful Chinese girl smiling with short hair, standing in the street of a modern city with a coffee cup. We also specify a Camera, which is a 5D this time mark three. And we also specify a lens which is a cannon at five millimeter portrait lens. Okay, the Aspect Ratio, I forgot to add the double hyphens. Let's see how that will work anyway. If you see, we didn't specify a photo of a beautiful Chinese girl here. But we did emphasize on the back of the prompt that we're wanting, that we're looking for professional photography. Let's Upscale one of the options. Yes. And we do see what we expected to see, though the background is very blurred. And we see a nice and sharp photo of a Chinese girl in a portrait style. So this part with the Lense specification did affect the image. You can play around with different use cases for the portrait lens. For now, let's continue and try the next prompt. So this time we're looking for a macro shot, basically because we're using a macro lens. So macro lens shot of the green grass just after the rain on a sunny day. Then we specify a camera, the Sony A7 R. And we also specify the lens, which is the 50 millimeter macro lens. Let's Upscale some option. This is a perfect example of the macro ones Shot. It's very good, for example, when you want to prompt, for example, for small items like a box, different insects, or a very close-up photo. The macro lens works perfectly for that. Let's try one more Prompt with the macro. I've changed the Aspect ratio here. And now we're prompting for a tiny lady bug on a green leaf during sunrise with a summer breeze, a hazy morning, and this stays the same. So that's the Sony A7 are Camera with macro lens. Okay? So that is also a good example and use case for the macro shots. When you want to showcase something tiny, it works very well with nature. But feel free to experiment with any other tiny objects. Let's go the next prompt now. So here we're going to try and get the zoomed photo. So we're prompting for a distant wildlife photo with a 600 mm zoom telephoto lens. And the scene is an African lion laying in the shadow, and it's a professional wildlife photography. And then here we'll also specify that it's being taken with a Canon EOS telephoto lens. So basically, we're emphasizing this quite a lot. I guess it's not that easy to get the most needed effect. So that's why we'll also have the word macro twice here. So a macro shot in a macro lens. This helps Midjourney to understand what we want, especially if we have a longer prompt. Now, let's try that and check the result Let's open it up. So we've got a very good result here, because that's exactly how objects look with a shot on a zoom lens. And the 600 millimeter, it's quite a big Lense that gives a very special effect. We can also try and Upscale and look and see the details a little bit closer. If we open it up and zoom in. So you see the edges or the objects behind, sorry, the background behind the main object and items in front of it, they're blurred quite heavily, while the object itself is in a very sharp focus. So that kind of effect is what you can expect when you're shooting with zoom that in real life. Let's check the prompt again. But we did say here multiple times that it's a distant wildlife photo. We said here in the beginning of the prompt that it's a 600 millimeter zoom telephoto lens. And then we repeated that zoom telephoto lens at the end of the prompt, basically pushing this idea further to Midjourney. Now, let's try another Camera, I'm sorry, another Lense in a quiet, interesting Setting. So we had a, we have a food photography here. So what we have here is the Spanish dish potatoes, Bravo's with sharp tomatoes, butter beans, and sliced thereto sausage. And it's photography, and it's Shot on a Fujifilm GI effects Camera with this lens, which is also Fujifilm Lense. So this is quite a popular Lense when it comes to food photography or something at this angle. Now let's zoom, sorry, let's Upscale. And as you see, we also have the background blurred. We have the main object in the focus, and we have a special angle that is typical for this type of whens. And once again, it's always a good idea to match the Use Case for the Lense with the scene that you're generating. So a portrait lenses would suit best for any photos of people. Macro lenses are perfectly suitable for any tiny objects pictured in a close-up. Zoom lenses are often used in wild photography, like bird watching or even sports like football, or basically any other sport. You can research that topic a little bit more of which lenses are used for what? And then you can match the stop picturing, for example, a wildlife. And then you can try a couple of different lenses and then Describe if it's a distant photo or a wide lens photo and etcetera. Now, before we continue to the next topic, I just have one simple prompt to share. So here we just asked for a photo. Let's ask for German German town district from the roof with a tilt shift lens. So the tilt shift is a very special lens. I believe those are one of the most expensive lenses, but they produce a very interesting effect. You can research that if you're interested. But they work very well with scenes like sound, aerial photos, photos from above. And this lens makes everything look like a toy. So this looks like a tiny toy town. Let's Upscale the Options. So that's the effect. While we're looking for. This doesn't look like a photo, more like a toy. But again, we had a very, very simple prompt and we can indicate that more. This actually looks more like a photo, but still we can make it even more photorealistic by adding special keywords that are relevant to the photorealistic Content. Okay, that's it for this video. 47. Advanced Techniques: View Angles: In this video, we explore the Use Cases of different Shot Angles, suitable lenses, and other keyword combinations that can help to achieve the rate View angle. I've found a good article on this topic with an overview of different shots. Different angle shots are starting with a long shot, which is the standard Type. Standard angle, the Medium which is the closer one, and the close-up and extreme close-up macro. Hi angle. Low angle. So basically a high angle is from above and a low angle is from below. And if you other ones like point of view over the shoulder, Ariel. So let's start exploring. And let's begin with this prompt. So it's an eye level angle shot of people enjoying food and drinks on a sunny day. It's a photorealistic shot on a Sony A7 are Camera with an 85 millimeter lens. So that's a good Lense for these types of angles. And we do specify in the front that we want an eye level ankle Shot. Now let's see what this gives us. Let's Upscale a couple of versions. So this is the perfect example of an eye level Camera Shot. So it could be slightly above or below. But overall, the scene or the object will be in the center. And there'll be Shot directly from a straight angle, from the side, a straight angle. We can see the same picture here. Now, let's use the same Prompt, but we're going to change the angle now. And we're going to try a bird view angle. Sometimes also referred to as an aerial shot. Okay, let's try this. Okay. So yeah, we did get a bird view in all four options. So it's not too high, but also overlooking the whole scene. Now let's try with the same prompt about with the more closer look. And we'll prompt for a low-level angle shot of people enjoying food, basically will keep the same Prompt. But we just changed the angle. So we've got our result. So this produced a different and new kind of result. This lock the version number for looks more or less like a low angle shot because it's a little bit upwards. However, on those, on the first three options that are very remotely related to the low angle shot. It looks more like an eye level. Maybe the number three is a little bit similar. Of course, it also depends from the scene. So maybe it's not really typical to have a low angle shot of multiple people sitting in the restaurant. That will be strange. So you would have to picture people from below the table watching in the camera, would need to watch upwards. So maybe that's the issue. But let's see how we can try it in a different way. I have a little bit of a different prompt here. And instead of specifying a low angle shot, I have an upward View Angles Shot here. And also in the prompt, we have basically, this is a Fashion Photography. We have a man dressed in modern fashion. So we also specify a Lense, a camera, and a few other keywords to describe the scene. But we do not have the low angle shot. Instead we have the upward View angle. Now let's wait for the result. So as you can see here, this prompt. I mean this upward View Angles worked very well for all four options. We have it with the needed point of view. You can try the same thing with a downward view instead of a high angle if that doesn't work. But of course, this depends on the scene. Now, let's try a different type. We're going to try a typical food photography angle, which is a flat lay photo. And we're prompting for a wrap with falafel bowls and a fresh salad. And we'll also specify a different camera because I can in Gibbs more brighter colors which is suitable for food photography. Let's see the result. So this is a very good example of a flat lay. Let's Upscale one of the Versions just to see it a little bit closer. So it's a view from the top. And in a flat lay style, flat lay also applies to different lifestyle scenes or other, let's say Instagram Style pictures. Now let's test another food photography style. This time we have a close-up shot of a group with the burger. Let's trigger that will also specify a photo Photography keyword. We'll change the wording a little bit. We'll say that it was taken with a Nikon Z7. And we'll also specify a lens which is, whoops, which is 135 millimeter lens, often suitable for the macro shots. So this is also a good example of a close-up photo. And in this type, sometimes even the object doesn't fit fully to the screen, but it's not the same as a macro shot. Feel free to experiment with that. This is the next prompt. So this is also food photography. So we're trying to picture a paella at a Spanish restaurant, which is a traditional dish in Spain. And apart from specifying that it's a food photography, we specify an angle inside the prompt and we actually specify a specific degree, which is a 45-degree angle photo shooting. With that, we also specify the Nikon and the Lense, but this is the main difference. 45-degree angle. This is quite common in food photography. So if you specify something unusual, for example, an angle that isn't usually used for this type of photography. So this might not work, but because this is quite typical, we can expect a certain type of view angle. Okay. Let's check it out. Yep. And definitely at least in those two, the third and the fourth Versions, we can definitely see a 45-degree angle shot. So Midjourney understands this quite well. And just one last prompt that I want to share is an extreme close-up angle, which can become handy in situations like this when we actually want to picture a portrait. In this case, we're looking for a blue haired Fashion model. So that's a Fashion Photography. But if we only specify a portrait, most likely will get a half body portrait or some other kind of look. But with an extreme close-up, most likely we are getting the face very close. And also an effect that our object might not fit entirely in the scene, which is common for the class app. Okay, Let's open it up and we can see clearly that this is a very close up photo. Made journey also understands that very well. So this is how you specify an angle and basically the viewpoint that you're looking for. Feel free to experiment with different angles, different degrees. Also try the combination of different lenses and see what works best for you. 48. Advanced Techniques: Shot Type: In this section, we'll briefly go through a few portrait shot Types, like a full body or a half body portrait shot. Now, let's check a couple of examples. So we'll pick the Fashion Photography for this. So we're requesting, so we're requesting a high-fashion model as a carrot, specifying the Camera Model, a few other keywords. The F18 stands for aperture. We will get to that a little bit later. But the main point here is that we're asking for a full body shot of a Fashion model. So full body shots or half body shots, they're mostly useful on or for when you're picturing people. However, sometimes you can also try to apply that to an object. It could be a car or house or even some other object. If for some reason that didn't fit into the scene, you can try and add a full body shot. This just tells mid journey to focus the main object in the center and that attribute visible from head, from the toe to the head. Now let's check the result. So we can definitely see everywhere that this is a full body shot of a model in some carrot colors. And we can also see and 80s vibe and all the photos. Now, let's use the same Prompt, but we'll change the full body to a half body Shot. As you also may notice, that this Prompt doesn't sound entirely grammatically correct. So in real life, you will and speak like that. But Over half body shot of a high-fashion model, etcetera. But this is also quite important to remove any unnecessary words from the prompt because those are also serving as noise. So if we have any extra punctuations or extra words that don't actually mean something. So this is an entirely okay or even preferred, preferable way of doing your prompt. So the difference between the good and the bad prompt is between how short or how compact you can make that prompt while still getting the same result you would have gotten from the lager product. Let's check the result. And we can clearly see that this is a perfect example of a half body shot portrait. Now, let's try one more thing. And this time will only request a portrait. So instead of a photo will say, and we'll let it do the rest. So it's just a portrait of a high-fashion model as a carrot. Oh, okay, let's try that. Okay, We have the result. As you see when you play around with Wording, this can actually affect the whole scene. We definitely, you can see more influence of a carrot. But what we can also see is that this is very similar two-and-a-half body Shot. And what this tells me is that instead of specifying a half body Shot, which is three words, we can just say portrait, which makes the prompt more compact and clear. Now let's explore our last example. This time we have a Medium full shot of a high-fashion model. The reason that I'm showing this is because sometimes when you try to picture an object and not a person, than things like a half body Shot might produce unexpected results because you have the word body in here. And for example, if you're trying to prompt for a car or a house that doesn't have people in it, then this type code work much better. Okay, We have the result. And what we can also see here is that there is much more influence of a carrot. It all options. Compare it to a half body Shot option. Because here we also have a little bit of a shorter Prompt. And we actually increased the relation between the medium and full word by the dash. So it all depends on the Use Case and the type of scene that you're prompting for 49. Advanced Techniques: Lighting Part 1: In the next video, we will explore various Lighting Techniques, exploring both natural and artificial lights. Let's start by examining natural lighting and it's ideal use cases. For this purpose, I suggest to create a dedicated channel. Okay, for this exercise, let's use an outdoor scene. So we'll start with an Australian countryside, vineyard, rolling hills, and the breathtaking sight. So we don't have any light settings in here. So let's see the basic result without any light Settings. And then we'll start adding and trying out different lights. Okay, So this is the basic view without any Lighting keywords. It does look natural and quite photorealistic. Now, let's try to add a simple key word of natural lighting. And let's see if it has any effect. So I just specified these two keywords at the end of the prompt, which are natural lighting. Right now, let's examine the result. So what we can notice here is a bit more photorealistic Content. And here also it's a bit brighter and has more green color. But overall, they look pretty similar. Now, let's try specific Lighting. So we want sunset light. But in order to help mid journey understand what we're looking for, we will add to our Prompt this part, Sun raiders stream through the gaps between vineyard bushes. And at the end, we also add sunset lighting. So we don't say that it's a sunset, but it should understand this and create a sunset Mood. Alright, let's open it up and we can see a very strong effect. All of the images now are much warmer. There is more contrast and shadows and we also see more sun rays. So if you ever wanted to receive a more natural scene, try to figure out what the best to Lighting is for that scene. Is it a sunset light or is it just a natural lighting? It could also be dawn or dusk. And help mid journey by putting in extra keywords. For example, for an overcast you could use a rainy food or something like that. Let's try one more option. So here we're going to say that it's an overcast lighting. We don't specify any extra keywords. And then let's check the result Over cash should give them more. Come. Cooler. White balance result over. Oh, okay, let's check it out. So as you can see, we have a little bit more of a darker scene and a bit more contrast and more cooler tones overall. So this is just a simple example of if you're prompting for some outdoor scenes than it makes sense to always specify the type of Lighting. And in this case, the natural type of lights would be the most suitable. So let's try one more prompt and will experiments now with local Lighting or light directions. This would be our Prompt photo of a man on a city street illuminated by natural light. Natural lighting. We will start with the basic prompt. Then we'll just add a few more keywords to Stylize the scene. So that's a captivating Atmosphere in urban street photography. Now let's try night Lighting and side Lighting So first we'll try night illuminated by night Lighting. But then we can try to change the angle. And we'll also change the prompt a little bit. So instead of illuminating, a man will illuminate his face. So his face and its surroundings dramatically illuminated by side Lighting. Let's try that. And while it's generating, Let's examine the initial results. So this would be natural lighting Examples. We can also see the effects of urban street style photography right here. But we don't see too much natural lighting. I mean, most of the lights in those scenes would come from buildings and cars. So really natural lighting works way better with more of an Outdoor while more of a nature type of Outdoor. You have some natural things in your scene. Now this is the night Lighting. Now we can see that most of the lights are coming out of the streetlights and the buildings around. And now let's check the angle. And now we can see that the light has changed the angle a little bit on this picture, actually on all pictures. So we can also see the contrast between the light areas and the shadow areas. Now, let's try to remove the extra words and see if Midjourney snow understands that we want a side lighting. So we only say, actually, let's even put it at the back of the Prompt. Here we only have the photo of a man on a city street. We have the styling and then we have the side Lighting at the end. Now, in this example, we do also see light coming from the side, the shadows on the coat, also on the face. So we do have a bit more of a darker and a brighter side. However, we do definitely see a stronger impact when we do specify certain key words. So here we have, First of all, we've chosen that we want to see the effect on the face mostly. And we also had our surroundings dramatically illuminated. All those little things. They help to achieve the desired lighting setup. Now let's take the simpler Prompt and we changed the side Lighting to a silhouette biting. So in theory, this should give us a Lighting from behind the object. And wallets generate a, let's actually try one more option. In here we specify not a silhouette lighting, but a light from behind. Okay, the silhouette lighting is ready. So here we can see quite a strong effect, the object that man is in the shadow. But we have a light coming from behind because we have the urban setting. Lights are mostly illuminated from the street lamps. Maybe apart from the first option, where it could be a sunset or sunrise. Now let's check this version, light from behind. So because we didn't specify that it's a silhouette, there is a lot less shadow on the object, but also the effect is a bit weaker. So again, depending on the effect that you're using, try it both ways if you want to highlight the object. Now another way to specify a light would be to use the relative keywords like golden hour. So that would be related to a sunset Mood. Or you could try the blue, or you could try blue. Our. The blue our refers to a time when the sun is down, but it's not dark as night yet. And we should also see, of course, more blue or cooler tones overall. Right? So this is what the golden hour looks like. So we can definitely see a sunset. And this is the blue hour. So it's not exactly night yet. Looks more like an overcast after the just after the sunset or before the sunrise. It all depends, of course, on the Setting. Like sidelining will work best on portraits or different objects. When you want to actually put a light behind or on the side of the object. So we use that with a scene like this, like a side Lighting, that wouldn't work very well because there's no specific object. We're just overlooking large scenery and also the other way around. So natural lighting didn't really help us here. Or when we specify that we want a night hour or a side Lighting or the golden hour. Some others specific type of light that works best in the city environments or with objects in the center. That combination produces the best results. Okay, So we'll continue with Lighting experiments in the next video. 50. Advanced Techniques: Lighting Part 2: In this video, we'll try artificial and specialized Lighting techniques, such as volumetric and cinematic. And we'll also try a few special effects. Let's continue in our Lighting channel. And we'll choose a new prompt for that. But without any letting Setting, just to compare. So the first would be a portrait of a medieval princess inside a majestic cathedral. Intricate details, historic and beautiful architecture and wallets being prepared. Let's also try to also Prompt and add cinematic lighting. And then we'll examine and compare. Okay, now let's check the first scene without any Lighting. We can see mostly soft lights, not too much contrast or color grading. Now let's open the second one. And we can definitely see a more stronger influence. We see more shadows. And this version, in this version, the white balance is a bit warmer. And there are also a few visual Style effects going on. We did get more like a cinematic feel. We didn't specify that this was a movie. We just specified as cinematic lighting. We can also try to add a still Movie keyword. So let's try that. Okay, so that's how movies fill would look like. And we can see now a lot more post-processing going on in here. Let's try one more technique. So when we got the needed Lighting that we see too much going on, we can try to add a style raw, which basically asks Midjourney to not overdo the overall aesthetics. And the mostly stick to our prompt that we're asking. Let's check it out. And what I can see here is that it's definitely more photorealistic than the previous one. And we still have all the atmospheric and Movie Style lights. Let's compare that. So, yeah, in the previous one, especially those two, they look a bit less photorealistic. And to my eye, a bit over processed. So in case you see a bit too much styling, then just try to add a style raw and see if that helps. Now, let's try one more type. So here we specify that we want a volumetric lighting. And we'll also add the word hazing will remove the Movie still, because that's not for volumetric lighting. You don't really need a Movie still. Volumetric is more related to 3D renders, 3D cartoons or PC games. And let's see the result. So if you see the second and the fourth version, we can see that there is a volume to the light. And the word hazy helps a little bit with that. But I still find it a little bit over processed, not so natural. Now let's improve this prompt a little bit. And we'll add a photo in the beginning. That's it. So it's a photo portrait. And we'll also try the raw Style. So to get the more photorealistic and natural look while maintaining the scene, which is quite specific and actually works very well with all kinds of non unusual Lighting. But it's used in games and movies. Okay, Let's check it out. And we also see volume, volumetric light here. Here. Very good. And on the second option, also, the final, the fourth version, we also see it here. But it's more of a sun rays Style I think that the Style raw helped us to end the keyword photo, helped us to achieve more photo realism in those. Now, let's try and experiment with sun rays and patches of light. So we leave photo keyword now. But we've changed the lighting to crepuscular rays Lighting. And we also add patches of light and shade. Now the reason we add the word shade is to increase the contrast a little bit to see the difference between the light and dark areas. And some extra keywords like hazy or foggy, could also help to highlight to the actual arrays. Let's see how that would look. Alright. So in the first and third options, we can clearly see the sun rays and also a bit more contrast between the lit areas and the shadows. Now, let's go ahead and try of view special effects. This would be our Prompt. Here we have a dual white and violet Lighting, and we'll also add split Lighting. So this is another way for using light or the light setting to affect the overall mood in the scene. You could have, instead of the white, you could have tried two different colors, including green and blue, or some other colors. Let's open it up. And we can definitely see the influence of the violet color. Also we see more white and all of the four pictures. So try to experiment with this, especially the more artificial colors like green, I'm sorry, not green but maybe like a blue or red, which we don't see in everyday life. Mostly it's white or yellow, like warm colors coming from the sun. And the artificial Type works very well with urban and night Settings. And as we saw before, the rays, patches of light and volumetric lights, they work very well with cinematic types of images. Now, let's try another nonstandard use for the Lighting. And basically here we will try to emit the light from the natural scene or object. So we have fluorescent light emitting and glowing. This could give us a very interesting result. Now let's prompt that. Okay, Now let's check it out. And to be honest, we did get quite a nice effect here. We don't see an emitter or light being emitted from something. But the summaries in the first option and in the second, with only a few tiny floating particles. This gives a very interesting effect, though free to experiment with that. What if we want to actually emit the light from an object? So the first part we leave as a basic one. We don't specify any light Setting. And then using the double column, we're basically making this a Mulltiprompt. We're asking to add a glowing dress emitting this light. Bioluminescence light is basically emitted from the underwater fish or some insects. It's a light emitted from living creatures. Now, let's see if we can make the dress glow. Okay, now let's Upscale. So we did get a glow effect in a couple of variants, of course, by improving the first part of the Prompt and adding more glowing keywords. Then playing with different wording. On the second part, we can try to reach the desired results. But overall, this, this worked and we start to see the glow being emitted from the princess. Now, try to experiment with other light settings, like dim lighting or epic lighting. So the dual or split light works very nicely in some scenes. And for this topic, we're done. And I will see you in the next video. 51. Advanced Techniques: Mood, Atmosphere and Lighting: Now we're going to learn how to combine mood and atmosphere and use Lighting to extract the most out of our scene. Let's continue to use our Lighting channel. I would rename that and add an atmosphere. So we have a Lighting and Atmosphere channel. And let's start with a morning Mood with soft lights and foggy atmosphere. And this is our Prompt. We're picturing white cliffs on a foggy morning in the British countryside. We add a soft morning glow. And we also specify this camera lens and the style of National Geographic. Now the style of National Geographic, It's similar to the award-winning photography or like Unsplash Style. It just helps to produce more aesthetically pleasing pictures. So let's wait for the render and then let's check it out. Alright, now let's Upscale one of the options to look a bit closer. And we see here how the foggy morning and the soft morning glow blend together to create a certain atmosphere in this image. But what if we wanted to make this Mood and this Atmosphere even stronger? So here we changed the lighting a little bit. We added a dark moody atmosphere, and we change the camera to the Sony A7 R, which usually produces more cooler tones. Let's also Upscale the first version. And we can definitely see that it's now a more dramatic scene. It's also a bit cooler and a lot more foggy. Even though we didn't add an extra keyword that will control the fog, the dark, moody atmosphere, and the fuse light working together to improve the scene. So here we have and Italian coastline with palm trees during sunrise. In a hot summer after a heavy rain. We also use a golden hour Lighting and it dreaming Mood. Now let's Upscale. And we can see how the golden hour and the sunrise affects the mood and atmosphere. As array. We also have warm colors and we also see how the after heavy rain keyword changes the clouds and makes the ground wet. Now let's try to change the mood. And now we're asking for a hot summer during heavy rain. We'll also add a mystical Mood and dramatic clouds. And we'll also change the lighting to artificial. This usually provide some more surreal and dramatic mood, or even like a cinematic style. Let's wait for the result. Also, we tried to experiment and we use a little bit different camera models. So we have the A7, our fourth model of Sony, and here we have the A7 three while keeping the same Lense. Now let's Upscale some version so we can see how the mood changes. And this time we get a more gloomy and atmospheric seen, more dramatic clouds. We don't see a heavy rain on those versions. But of course we could emphasize a bit more or even make a Mulltiprompt out of it and just add more keywords for the rain. Okay, now let's try a new prompt. So here we're expecting a more colder atmosphere. So we're asking for northern lights at night captured in Iceland. We also add cozy things like camping bonfire lights. And we're also keeping the National Geographic Style. So let's see what result we get from this Here we're actually mixing the cozy warm lights, like camping bonfire lights. But because these are the northern lights and we have Iceland, this should produce a very nice mix of cold and warm colors. Okay, now let's Upscale. So indeed here we have a very nice Atmosphere, cozy, warm, light, red leucine is located in Iceland. And we also have nice northern whites here. So altogether this all adds up and creates a cozy, but at the same time, a colder Mood. Now, let's try a more different technique. So here we use a close-up photo of a bowl of ice, basically inaction because it's being dropped into water on a cold winter day. And we'll also add ripples, splashes, reflections on the water surface. And we specify Lighting as backlight, softbox Lighting. And we also shoot this on nikon. Now, let's see what this gives us and then we'll examine those out. Let's Upscale a couple of Versions. And we can clearly see here a cold winter day Atmosphere. We also see the backlight reflecting within the inside of the ice ball. Softbox Lighting actually works more in studio photo shooting setups. But it gives a nice touch to the overall atmosphere and makes the picture more like a studio shot picture. Feel free to experiment with those with the backlight and the softbox Lighting together. The backlight works good when you have a specific object. And a softbox is good for studio type Settings. Okay, now let's try our final prompt for this video, where we have a very specific Atmosphere. So here we have a wide angle shot of people dancing in a crowded club. Let's call it a nightclub. So this is an indoor setting. And we specify that this scene is illuminated by disco lights and also an off-camera flash. We use the Canon EOS for this purpose. It should also allow us to extract more color out of the image, just as we have here, disco lights and camera flash. Now, let's wait for the results and examine it. Alright, now let's Upscale. And we've got a very atmospheric and very moody image. We see bright colors, a lot of contrast and shadows. So we have a crowded club at midnight as the scene. But I'd also gives a special mood or an atmosphere to it. So it's very atmospheric. And you see how lights actually change the perception of the whole scene. So basically, it's definitely worth experimenting to combining different lights, atmospheric scenes for best results, tried to match that. So if we're talking about a dark mood, then things like a foggy morning or diffused lighting will improve the overall style. Or when you have scenes like this, try to use more artificial lights. You can also experiment with neon Lighting or LED lights. And this will, and this will help you to get the most perfect and atmospheric picture 52. Advanced Techniques: Camera Settings: In this section, we cover a few extra details about the Camera Settings that can impact the scene and help to achieve the needed results. For this purpose, let's go to the Photorealistic channel. But before we start Prompting, let's start to test the theory behind the settings that we're going to use. So the first one is the shutter speed. On professional cameras, you can usually set a very fast shutter speed, for example, 1,000, which means that the metrics that actually capture the image is being opened or activated only 1000th of a second. And during that period, all of the light that touches the metrics is then being transformed into an image. Or it can be very slow, like 1 s. So the shutter opens for a second and then all the light that enters is being accumulated on the metrics and then transformed into the image. The shutter speed can be even slower than that, or even manually controlled. And we will examine a few prompts where the Setting a low shutter can be very useful. The next one is an aperture. This is mainly useful when you want to blur the background and create mainly a depth of field effect in mid journey, a narrow aperture or a high F numbers, they don't really work as well. At least at the moment. We can experiment with that too. And then the ISO, this is the sensitivity. So a high value of sensitivity would be 1,000. 2000 are suitable for a low light condition but can result in a noisy image in the camera. Mid journey is struggling to produce noise, so it's very hard to produce a grainy or noisy image by default, with Midjourney, even though we're specifying a high iso number. But with low iso, we can get quite high-quality Effects and manipulate the final results a little bit. Okay, so this would be our First Prompt. We have an extreme close-up photo of fresh raspberries. We have a top-down view, diffused natural light. We have a macro lens and we have depth of field specified twice. We also add a small ISO number. This usually helps to get a more quality picture and the slow shutter. I also recommend to add overall style. Whenever you're creating very photorealistic Content. And if you specify multiple settings for the camera, it makes sense to add a style raw, but feel free to try that with and without the raw Style. Okay, Now let's check the result. So we see here a very strong depth of field effect. We see the main object in focus being very sharp. So this is a nice technique when you want to highlight the main object or the center of your scene, you just need to add a depth of field and sometimes a macro lens also helps. So we do have a slow shutter speed here. But let me show you one example of how this keyword can actually paint us a very different seem. So in this scene, in this prompt, we specify that we want light Painting, long exposure photography. The scene is a man on the beach at night light Painting. Scene illuminated by moving light sources. So the long exposure is the same as the slow shutter speed. Now let's try this prompt. Now, let's examine the result. So this is the effect that you would expect with long exposure or very slow shutter speed. We even got a little time-lapse effect on one of the guys, I'm one of the images. This is also possible with a slow shutter. And the way it works in real life is that when the shutter of the camera opens up, all the light is just being accumulated. And if you, for example, turn on the torch and just walk in front of the camera from one side to another that would draw a line with this light. So that's kind of effect of what we see here. So another cool way of using this feature is to prompt for city night views The roads at night. In this prompt, we have a bird's eye view captured on a Sony camera. We have a little liked long exposure, slow shutter night photography, a very small ISO number. And we start with light trails. And yes, this scene is the busy city world. Let's trigger that. Alright, let's Upscale one of the options to see it a bit closer. So in this picture we also see the effect from a long exposure, slow shutter speed, that busy world. And because of the small ISO number and the night city scene overall, we're also getting a nice skyline Overview. By the way, the effect from street lamps and the way that they cast shadows on the fries. I can say that this also, this effect also is related to the long exposure photography. Now let's try a couple of more scenes. So this is a very simple prompt, just a photo to sports cars on the racing track. One very important thing here is driving very fast. This gives the motion at action, and it also indicates slow shutter photography. So this should produce an image to us of two cars in focus. But the background should be blurred because of the action and the slow shutter. Alright, let's check it out. And we see a very strong effect from those words, especially on the third option here, in case you get too much blur, you can try removing some keywords, trying to play with the depth of field, because that will usually sharpen the main object or adding extra keywords like in-focus, sharp, etc. but I'm pretty happy with the results because that's kind of results that I was expecting from this prompt. Now let's try a final one. So we want to prompt for a selfie photo and we'll specify a low quality smartphones selfie photo with a young man at home. We specify low contrast. Pictured a, let's find that low contrast and pictured on a smartphone camera and will also indicate a high ISO number. Also, we have a tall Aspect Ratio, which is typical for selfies. And as far as I know, mid journey, especially the latest versions, it's really hard to make them create some ugly or bad quality picture. But things like a low-quality smartphone photo, low contrast, a smartphone camera photo, things like this can help to achieve. If we want to get, for example, a more natural and life-like look. Alright, let's Upscale. Let's take the second version. It looks more natural and let's compare with the first one. So this image looks like quite good quality, but still casual and without any post-processing photo selfie. So I would expect something like this, but maybe with a bit more lower-quality. In this picture, we see more of a post-processing thing like, like filters. We can see that the black here, the black color, I mean, the dark shadow, is lightened up a little bit. Also, this looks more like a shot on a professional camera than a selfie, but still unhappy with at least one result. What we can also tries to include a style raw parameter. Alright, so we did get very similar looking results. I'll say this option, the second option is the most natural looking. So feel free to experiment also, in this case, try adding more lower resolution and some other keywords that can help indicate. Again, because the underlying data set emit journey has been filtered out with mostly good quality pictures with higher resolution. So it's really hard to get that. Tried to combine and play around with depth of field, shutter speed and aperture to see what's really possible with Midjourney 53. Combination of Photorealistic Techniques: Now we're going to combine all of the photorealistic techniques that we learned so far. And creating truly photorealistic and high-quality content using special Techniques and keywords applied in professional photography. Now let's go to our photorealistic channel. Let's begin Prompting. So our first prompt will be a portrait with a special mode. So we have a portrait of a sailor at C. We have studio backlighting. We have a stormy Mood, a classic film effect, cold white balance. And we also have a Laika M6 model with Kodak Gold Film. So all this together creates a special vibe or a special Atmosphere. So we have a stormy Mood. We also have a Laika and six, which is a retro Camera with a film effect. And a couple of others like rugged sailor at C, weather beaten and faiths. All this adds up to create this special effect. Okay, Let's check it out. Now we can see here how all of those keywords are playing together to create this kind of romantic, stormy atmosphere. We also can see a lot of detailing, like in HDR Style, Light contrast. The bottom two objects we even have in special effect, like black and white and more of a bluish if I Upscale one. So you can also see the depth of field here. Also the Laika and Film that we've chosen for the camera and the film effect. Those are all influencing the image, producing that kind of vibe that we were looking for. Now let's try the next Prompt with an absolutely different mood and atmosphere. So here we have Michael Jackson on stage at a concert taking a selfie and we have a concert Mood, dynamic pose, which should make him not just standing still, but in an interesting pose. We're adding vibrant stage lights that are reflecting on his sunglasses. Let's see if that's going to work. We have the cannon model, the Lense depth of field. The white balance is specify it here in the temperature. And this should be like a wide daylight white balance. Then we also say here, a stage lighting, dramatic lighting, a concert light, basically emphasizing very much that this is a very brightly lit scene with very dramatic and very diverse Lighting overall. Alright, now let's Upscale couple of versions. I can see a good fit. And we can definitely see a dramatic high contrast Lighting with the concert Lighting from both sides, shiny reflections from the suit and the dynamic pose and charismatic expressions we can see are playing out in the posture and overall action in the photo. Let's check out this version. We also see here how the dramatic lighting and the concert Lighting play out. It's very bright. At the same time we have a very dark part of the ceiling or the dark part of the picture because it's a concert. And what I can also notice here is that the white balance, which is the white daylight Style, is giving this a more natural look to the image. Together with the Canon EOS style photos. And interestingly enough, the aviator sunglasses. Okay, Let's check. Yeah, we can actually see on all four options, which was only a tiny keyword specified in the middle of the prompt. Okay. Now let's move on to another Prompt. Here, we're prompting for Movie still. But instead of specifying cinematic lighting, will specify here a volumetric foggy Lighting. Because we have a very special seen here, which is N. It's basically dimly lit corridors with an agent in guarded secret facility. We want a mystical Mood Creating an atmosphere of tension and stealth. Not sure how this is going to play out, but let's see. And then we also add a split light of red and blue, which together with the scene Mood, the low-light Cameras Setting, and also the shutter speed, should produce a very interesting and cinematic result. Will also specify an Aspect Ratio of 16 to nine, which is quite a standard for cinema. So basically the idea here is that you need to try and combine the lighting, the camera, the lens, and the Film Effects and everything together. So all of this acts like an amplifier to your scene. And basically your target here is to increase or highlight all of the scene aspects or as many seen aspects as you can. Alright, let's Upscale. I believe that we managed to hit the scene pretty well. We have here a dimly lit corridor with volumetric lights. We have a red and blue split color and a little like a mystical touch to it. If you've noticed here, we specified that it's a Movie still, while also specifying that it's a Canon EOS, which usually doesn't come together, but this gives a more natural feel to the image. And then Movie still is basically responsible for giving it a Movie feel, which work together pretty well. In this version, we're getting similar results, which corresponds to our Prompt pretty well. And also, as I can see, the volumetric lighting usually works pretty well when you also add keywords like hazy or 5D a little bit because then the light has something to light up, so it becomes volumetric. So those two things coming together pretty well. Alright, let's move on to the next prompt. And here we have quite an unusual amount. We have in Egyptian warrior, in traditional Pharaoh dress in a battlefield, and we're having a fisheye lens Style effect. We also specify moody cinematic lighting, creating shadows and contrast. A super detailed historical moody photo shot on a Sony A7 or Camera. But we do specify in here quite interesting film which produces black and white images. However, the camera is a modern day Camera. And we'll also try to combine that with a fisheye lens, which altogether should give quite an exotic look. And let's try it out. And we also didn't specify an Aspect ratio here. Because of the fisheye lens Style. I believe that this will work better on the square image. Okay, Let's check the result. And we can definitely see that all pictures are colored pictures. So there's no black and white option. Despite that, we did specify the film, but everything else works pretty well. However, let's try to play with Wording a little bit. Increase the chances of getting black and white picture. Let's change the beginning of the prompt to mono chrome photo. So instead of saying black and white photo, we're specifying the monochrome, which gives more interesting results. Usually, let's try to remove the camera and the fisheye lens in the back because we did specify it in the front. And now let's try it again. So we've changed a little bit, not too much. We've just changed the camera and we've added a monochrome photo in the beginning. The rest is pretty much the same. Now let's wait for the result. Alright? And you can definitely see here now that we have a much stronger fisheye effect. And all of the pictures are monochrome, not exactly in black and white, but a monochrome, which is exactly what we wanted. Okay? So this is how you can combine all the Camera Settings and Effects to amplify a certain scene. From now on, you should be able to achieve very photorealistic Content. However, in the upcoming videos, I will teach you how to further increase that and explore the best combinations for specific topics 54. Advanced Techniques: Prompts Fine Tuning: In this video, we'll complete the Fine Tuning of Photorealistic prompts. And we'll go through a few keywords that can help with a more realistic color grading and increase the quality. Let's continue prompting and our photorealistic channel. Now, before we begin with our prompts, I just want to share one prompt that I found on the Internet and explain why we shouldn't prompt in this way. So quite often, in the Midjourney Gallery and other resources online, you'll find similar looking Prompts. We're, you'll have the scene description in the beginning. So basically, usually it says what mediums. So is it a photo or illustration? And here we have an Indian village woman at a gathering in the forest of Hamas you Pradesh. And everything else that you see here is not really related to a scene. But we have a cinematic, a photoshoot, than we have all kinds of different key words that are associated with the image quality or a 3D rendering technique. And a lot of other different keywords that by themselves don't really described the scene, but are more describing the quality. However, people usually using this kind of Prompt mostly are only copying and pasting the Prompt without thinking or reading what's written here. And very often this actually affects the seed in which you are not at the scene without real intention. And I often also will find key words like scan lines or dramatic abbreviation and things like tilt, blur and other stuff that are that can't usually work together or are intended to produce a very specific effect. So I strongly advise you that if you ever encounter a similar prompt that before you actually trigger the image generation, you read that carefully. And if you don't know what this means, it's better that you researched this online. However, in this video, I want to show the actual use case of some of that, of some of those keywords that can actually help you to achieve certain results. You're going to increase the quality. Let's take, Let's take this prompt as an example. So this is quite a basic prompt. We specified that this is a photo, we specify that this is a camera and all the other details. We also add a sunset sky, and we wanted in it Aspect ratio of three to two. Now let's wait for the result and then we compare with an enhanced version of the same Prompt. Ok, so we've got pretty photorealistic Content on all four options. Let's Upscale one of it. Now. It is a good-looking scene, but it's not as dramatic as I want it to be. So and also the quality. So we've got, we've got a quite detailed. But let's see what happens if we actually add a color grading. So here we increase the red, so we say red color grading. We also add an apocalyptic red sky, which should increase the dramatic faith, which should increase the dramatic clouds. But if you notice here we have a few extra keywords like scattering, ray tracing, octane render, optics, ambient Ambient Occlusion, motion blur, lens flares, color grading, and a couple that you've seen before, which is the depth of field and postprocessing. So all this together with the change from a photo to a Movie, still should produce a very different result with More drama happening in it while keeping the photo realism in it. Alright, now let's examine and compare. I can already see a lot more tiny details, large dramatic sky here and post-processing happening on a few options. Let's Upscale one of it. Let's take the second one. It also has some water So this is the enhanced version and we can definitely see more drama and contrast are put into it. So we see how all those keywords are starting to affect it. But one thing you'll notice is that I think we lost a bit of photorealism if we compare to the first version. Now, let's test one more thing. So we keep all those extra keywords like ambient, like octane, Render, ray tracing and the color grading. But in sped. But instead of Movie still, we specify photo and we add a style raw at the end. And this is usually my favorite of prompting for photorealistic and dramatic content. Because this combination of the photo and the Camera Settings, together with things like octane render and ray tracing, which are only used, which these are only technologies used when you render. Together with the raw Style on the five-point one version, gives the perfect balance between the quality and the photo realism. But of course this always depends on the scene and multiple other factors like the Mood overall and etcetera, of course. So you'll have to test and find the combination that works best for you. Okay, now let's Upscale. So this gives, so here we have quite nice result. It's very photorealistic. It's not as dramatic. But I can definitely see that all those rocks and the water, they're all produced in a very high-quality and good sharpness. And the scene itself is quite balanced. Well, of course, we don't see the red sky here. We don't mention that in the prompt, but because the Style raw removes a lot of the post-processing aesthetics from it. If we would still bought the sky to be red, we would just need to emphasize on this a bit more and add extra keywords in our prompt. Now, let's experiment with a couple more Prompts. So our next prompt is a shock captured in a bird's-eye view with a young boy and his dog walking through a deserted forest on a winter night. And then a few Camera Settings and award winning photography. And the second variant work be identical apart from the back of the Prompt where we specified all the extra key words. We also add a vignette effect, cool color grading, the Mood Setting, light setting. We also add the octane reader, ray tracing and etc. including the color grading. So here, even though we didn't specify that this is a movie or something, this should give a more interesting scene overall. Alright, so this is our version without any special keywords, which are quite interesting. We can definitely see here how the slow shutter speed is playing out. The bird's eye view and also the scene itself. Okay, now let's compare to this variant with extra keywords. And I can already see that scenes are a bit more interesting. Some of them lost a bit of the photo realism, but this can be recovered a bit with extra keywords that specify this or the Style raw. But it's definitely more cinematic, I would say, and definitely more interesting and detailed. So all of those things, they affect the result by a small margin. I cannot say that this is a very big effect. So of course, depending on the scene, you should try and add some of those keywords. But of course, for any photorealistic Content, don't forget to always include the camera. Some of its settings are important, also the move, Mood, the Lighting, and of course, the scene itself. Now the last Prompt, Let's see how that works. So we have an underwater photo of a bear catching a fish. We have the Camera Settings here, also the National Geographic Style and a couple of other Camera related keywords. And now let's try with color grading. So we'll specify the aqua color graded. Maldives see colors and also those extra keywords like ray tracing. So here we have ray tracing for the reflections. Because of the underwater scene, this suits quite well in my opinion. And the tone mapping, post-production, and everything else you can see here. And also the Style raw, which should make the content a bit more photorealistic while keeping all the aesthetics that we've asked for. Okay, so that's our first option. Let's Upscale one. And let's open this up in the browser. Now, let's check the second version. And then we will compare it, right? And we'll actually Upscale a couple of versions. Okay, Let's open it up in the browser. And if we compare this with the first version, once again, we can notice that there's less details here. The reflection from the water is a bit more intense. Also the influence on the scene. We can see the aqua color also the influence on the scene overall. Quite strong. Let's check again the key words. This is by the way, the second Upscale. And we can also see more intense details, more intense reflections, the aqua color grading and things like this. However, don't forget that we use the raw Style, which gives this a little extra photo realism. If you want to experiment and try even more dramatic and emotional scenes, then just remove the style raw and then you should be getting more post-processed and finished Content. Like after going through the post-production process in light room or something similar. So this is how you can fine tune your prompt to add photo realism and manipulate the colors a little bit further and create more dynamic and interesting scenes 55. Advanced Techniques: Image References and Image Weights Part 1: In this section, we'll delve deeper into the topics of Image References and Image Weights. Let's go to our Discord server and let's switch the channel to Image prompts. So for this exercise, I've prepared a photo of a black cat, and this was generated with Midjourney. It looks very photorealistic. You can use any other cat as a reference. But it's quite important that this would be a high-quality image with a minimum of 1,000 pixels in width and height. So let's take the link. And now let's do it as usual. So we Type imagine. Then in the beginning of the prompt, we paste the link. But we don't just want to replicate that black cat. We want to take him and we want to put him into another scene. For example, outdoors in the warm sunny weather. And this time, we also added a special photographers Style. So this is our prompt. So we have outdoors, warm and sunny, sunny day. The Emma Thompson is the photographer and we, we are prompting in this style. We also have some other scenes Settings here. But the main difference is that we have the link in front. Now let's trigger that. And we also note here that we want to black cat. So this is a good idea to specify in your text prompt an object and a scene that you're actually looking for. But if you want to modify it, we can continue that explanation in your text prompt further. All right, let's open it up. And this was our image. We can definitely see some similarities here. We can see the same eye color also the For is definitely similar on the second and the fourth option. Let's Upscale. Let's take the second version. Okay? So if you see we did get some artifact here in the form of a watermark. In a future chapter, we will learn how to remove those using Photoshop. But of course, you can always try the No keyword, sorry, the no parameter or just a negative way. And then you just specify like things like text and Watermark, and then you put the minimum weight to it. But I don't recommend going below zero -0.5. Okay, anyway, let's come back to the cat. So we've got the initial cat or a very similar looking cat in the scene that we specified with our text prompt. Now let's try one thing. And what we'll do is we'll use a special parameter. How old is W? Which is, which stands for Image weight. And what we're doing here is we're increasing the weight from standard one to two, which, which basically is twice higher. We keep the text prompt intact and then let's examine the difference. Alright, so we have our result. Let's actually open the initial Prompt here. Is that right? Okay. We can immediately notice the obvious difference from the previous result is that now our cat in the image looks a lot more similar to the initial prompt, to the initial image Reference that we've provided. And not just in terms of the overall walk, like for, for example, but also the posture. You can see. Now what's a lot more similar? Because in here, we've got a cat actually in a different posture. And this option is a lot more similar because the cat's walking and also watching downwards We still get like a summer vibe. But we can also see that our environment that the cat was in, in our image Reference also influence the outcome. So the cat is not anywhere outdoors, it looks more indoors, but also on a sunny summer day. This also, you can also learn how to play with the weight by specifying it at one-and-a-half or even lowering it and see how you can achieve the need of result. So this would be our next prompt. So we have a ginger OK, let's type that. So we have a photo of a ginger boy, and we specify that the boy is eating ice cream. As you can see on the image Reference. We only have a boy who's smiling sitting inside. Now, let's see how this would affect the result. Alright, let's open it up. So if we compare, the boy looks quite similar. It's not identical, of course. But we can also see that Midjourney took the description about the scene and ordered the background also from the image Reference that we provided and of course, the image of the eating ice cream action. We can see everywhere. And in a similar fashion, you can specify a new background or a new scene. But don't forget to indicate the actual object that you want to, that you want Midjourney to extract basically from the image Reference and then add it to the scene that you're looking for. Of course, this works only if you specify an action like eating ice cream. But the results will not be consistent. And it's just a general good practice to help MED journey and indicate the key aspects that you want to pay attention to in the image Reference. Alright, so for the next exercise, we're going to try and blend two images together. So we've tried already in one of our previous exercises. This time, we're trying with a person was more of an abstract Blend. Now we have two different artworks. Both were generated in mid journey. All you need to do is just paste the two links separated by space and hit Enter. Of course, you can also specify and Aspect Ratio, or even just add some texts into it. But if you just want to combine two artworks, but make sure they are in the similar, are the same aspect ratio. It's not always needed, but it is recommended because otherwise you could get black borders. Okay? And we can see the merge of two completely different aesthetics in one image. The second version actually looks pretty interesting. Let's open it up and let's compare. So this was the second image Reference, and this was the first one. And this is the result. So feel free to experimenting and emerging two different styles. And also don't forget to check the Aspect Ratio. So when you do that, now for the final prompt, we have a dog. And let's say that we want to now take this dog and put it near to the lake on a summer day. So this is how our prompt would look. So we have the image as a reference. Then we specify the scene. A cute dog near the lake on a sunny summer day. Then we have the special aesthetics. So a photo taken with Sarah Thompson, we specify the camera portrait lens. And if you are the key words that can just beautified the image. However, what we'll also add here is we add an image weight. And let's start actually with 1.6. So that would be one way. And then, which is basically a bit increased. And then mobile also do is we'll try a smaller weight at 0.7. And then let's compare how those work. Alright, so this is our First Prompt with increased Image weight. So this is the reference. A dog sitting outdoors on a terrorist image weight is 1.6. And we did get quite a similar scene. Again, we can see maybe here, but then it's not obvious. We have a lake somewhere. So this dog almost looks identical to the one we provided. This is the smaller weight. Let's check it out. And we can now see the effect of our text prompt. So everywhere we can see the lake, of course, those two dogs are not similar to our image Reference. But those two, the second and the third, they are quite similar. Let's check. Yeah. Let's Upscale. Let's compare a little bit. So it's not identical. Of course, you need to virile a bit more on the Options, on the options that look similar to the target seen, the reference. But in case you notice that you can't get the scene into the result or an object or the scene from your reference image. It's not reflected well enough in your results. Then just try and play with the ways by increasing it or decreasing it. That's it for this video. And I will see you in the next one. 56. Advanced Techniques: Image References and Image Weights Part 2: In this section, we will continue to explore the use cases for Prompting Image References. Additionally, we will take a look at how to combine objects from one image with the scene of another image. Let's continue in our image Prompts channel. Now let's take a look at our Image References. So for the first exercise, we're going to try to take the which from this image, from the first Reference. And then try to combine it with this scene. Basically I placing the which at the road in a foggy and dark environment. First of all, let's try and combine the Prompts without any texts. And let's see how mid journey interprets this. Don't forget to specify the Aspect ratio. What's the Aspect Ratio doesn't need to be perfect. But the closer you are to your Image References, then the better the results will be. Alright, so this is a result. Let's Upscale. But it does look like a 50% mixture from one image to the other one. But we don't quite see the witch on this streaked. Let's open it up. So it is a more 50, 50 per cent MC mixture. But now let's help mid journey to understand this a bit better while we are looking for here. So we're supplying to Image References the same links as before. But we also exploring in here that we were looking for a Movie still with a which pictured from behind walking at night down the foggy road. And we also add a mystical Mood. Now let's see how this affects the result. Right? We have the result. So this is a perfect merge with a little assistance to Midjourney of how to do this merge. If we take this example, we see that image is a bit brighter and we also see that check the reference. More influence from this reference of the scene, the surroundings to our prompt, to the result where we didn't specify any texts prompt. As you can see here. But in this, it's a lot more closer to the scene overall. To our second image Reference. It's much darker, and also it has this foggy lamps effect. Let's Upscale. Now, this looks a lot more closer than what we intended to receive. Of course, we can enhance the text Part a bit more. But this is a perfect example of how you can combine to Image References and just assist mid journey to understand that a bit better. I mean your idea. Now let's take a look on our next example. Here we have a photo of an Asian looking man. Here's an urban street in New York. And let's say that we want to combine and we want to place a similar looking man on the street. Now, let's try to prompt and will also supply here two links. And we'll explain a little bit that we want a portrait photo of an Asian man with a blue coat and hat posing on a New York Street in-between New York buildings during bright day. So we do specify a bit more that we want special type of buildings. And we'll describe the man a little bit more so that we don't just say a man. Because mid journey would identify this as a man with a blue coat and hat. In general. That's how by describing objects from one image Reference to another, items from the second image Reference like seen, for example. This way you can actually help mid journey to link those together or linked specific attributes from those images together. Alright, now let's check it out and let's Upscale one version So as you see, this is a perfect blend with a seam from one image Reference and an object from another. Because if we take a look at our image Reference with a man, we do see here a large heel on the background. We see a folk, or we don't see anything of that here. Instead we see a New York background, which is exactly what we asked for and we did provide a reference for that. Let's check it out again. So that's another good example of how you can combine a scene and an object with a low description. Of course, don't forget to choose similar aspect ratios. Of course, single Medium. So if you're trying to merge photos, those should be photos. Merging and Cartoon tuple photo or a photorealistic image would not produce very good results. Now, let's check out the last example. And here we have a landscape photo of a retro car, more of a Cuban Style. And then the second photo, as you can see, we don't have a road here. We have a lake instead. And it looks more like a German type of old town or at all churn type. Now let's try to merge those. And first we just try without any description. And then we add in the second prompt, we add a small description. And let me just say a retro Cuban car parked near the lake in German Old Town. So basically we're describing the scene and the object, and let's check out the difference. Okay, So we have our first result. And we see that the object, the car, is a very close match to our reference. Even though we only have the front of the car here and the side, but all four variants here, there are pictures from height. I can also notice balance on some variants. But overall, Midjourney understood quite well that we're trying to integrate an object from one image and the scene from another, because the scene itself was quite ND. Let's take a look again. I mean, the center of the steam. And here it's the other way around or more going on in the center and not so much around it. Now let's check out our second example where we actually specified that it's a car parked in a German Old Town Road. And what you see here that we did get something from our texts Prompt. But to me and more looks like that the car is parked on some Cubans street, not a German street. Also, the car looks different, so we didn't get a close match. This is quite close, but the color is different. So that's why it's important to Describe your prompts and as much detail as possible without any excessive words, of course. So we could try adding the red retro Cuban car. It's grabbing the scene in the second sentence, for example, by even repeating some words. And as the last resort, we could even try to increase the weight for images by specifying the parameter is W and then 1.2 or 1.5, or even a higher value. Of course, it depends on the scene and you should experiment and see what works best for you. Now let's try the last prompt. And while we do here, is we will combine three images. So we will combine the New York Street, the Asian man, and we'll also combine this mystical foggy road. Where was it? And this image. So together, this should create mystical photoshoot in the New York, but more of a movie style or something like that. And we also add an extra keyword here, like mystical evening and horror movie Mood. So this is the example, or you could specify three or even more, up to five Image References in fact. And try to combine all of those in one scene. So we're combining the object here, we're combining the location from the second Reference, and we're combining the Mood from the third one, which is the mystical foggy road. Alright. So this is the result. Let's Upscale. We can clearly see that the object and the surroundings and the mood is just a combination of those three Image References plus our description for it. So this was the last example of using Image prompts for this video. And now we're ready to move on to our next topics. 57. Photorealistic Content: Fashion: Starting from this video, we will begin practicing different photorealistic topics. We will try to utilize every skill that we've learned so far. And we'll apply the most suitable Techniques for each topic. Let's start with the Fashion and the magazine tab of content. For this purpose, you can either use of Photorealistic channel or I suggest to create a separate category. And we call that photorealistic. Let's drag this channel into the newly created category. Yes, we want to sink permission. That's and you can also reorganize the other channels and also drag them in here. Because the Lighting mostly contains photorealistic Content and some other channels to. But because we're going to practice multiple photorealistic Styles, I suggest creating dedicated channels for each topic. So then you can always come back to the channel and remind yourself, what are the suitable Techniques for this particular topic? So let's create a channel called Fashion. Let's begin Prompting. So in the First Prompt, we have a model woman during a Fashion Week event, and we have a backdrop with a space themed ambiance. We do specify the Lighting here, the camera. We'll also add a couple of keywords like ray tracing and post-processing. But the most important thing here, when we're prompting for the fashion magazine type of content, we also want to include that this is a Fashion photoshoot or a magazine photoshoot, or something related to fashion. This is very important and it can also be used, as a matter of fact, in regular kind of portraits Content. For example, when you want to prompt based on some image Reference and some person, if you want to create some really nice and outstanding Style, then Fashion usually works really well for portraits. Anyway, let's see. Let's wait for the results and then examine it. Okay, Let's Upscale. We did get quite interesting results here. I can see the influence of a space Theme and the buck headlights. And of course we have the Fashion postures. And I'm quite happy with the overall quality. Let's check the second verse and let's check it a bit closer. It's quite nice. We can even see the texture on the close here. We also don't see any artifacts or extra fingers or anything like that. So this is interesting. We didn't specify any kind of lighting. I mean, the color that mid journey chose blue and violet as a main. I also think that the intensity has been influenced by this broad daylight thing. You can cycle and make a few extra Options. I think this can produce a very interesting result. Now, Let's try the next one. So here we have a captivating portrait photo with a young Vietnamese actress. Wolf specify a few extra keywords like long hair, H. And of course, the technicalities here, we chose the cinematic lighting. But the most important thing here is that we had an inspired by keyword. And if you can notice that we didn't specify anywhere that this is a fashion photo. But because Peter Lindbergh is a fashion photographer, this should influence our Style quite a lot. I did specify style raw. Let's first try it without it And then we'll see how it works with the style rock keyword in it. But the idea here is that you can also achieve the Fashion look by adding these types of keywords like inspired by or Photoshop by or in the style of. And then specifying the artist that is related to the topic. And of course, the artist or photographer should be very well-known or trained by mid journey so that it understands the Style and the aesthetics of it. Okay, So this was the option without the raw Style parameter. And we can see how words like captivating, radiant beauty and other technical keywords here are beautifying the image. So here we have the raw Style parameter. Now, we did get a more natural look here, but it's not really that obvious if we compare. And to be honest, I prefer when we're talking about the Fashion Photography to keep this style raw away from the prompt. Because that's Midjourney adds its own aesthetic vibe to it. Okay, let's now use the next prompt. Now, this is a bit more complex Prompt Structure. So we have a Fashion photoshoot of a handsome white Nordic woman at midnight with dark, starry night. Then we split that prompt, so double column and then we specify the second prompt, which is the fashion shoot of Nordic woman at a fashion event in Tokyo. So basically here we're mixing the same object but different scenes. If we can say that we have an event in Tokyo. And then here we only specify midnight, dark, sorry night. Now the third part here contains only the characteristics or Effects. So we have photo realism and surrealism. Here we have the camera and everything related to that, including things like ray tracing and post-production, also the K. So let's trigger that. Oh, and we also have a Landscape Aspect Ratio. All right, we've got our results. Now, let's Upscale a couple of Versions. And we can actually see on all of them that they are very diverse and quite surrealistic. Now we can definitely see how the white Nordic woman model plays out here. I can say that the photo realism is a little bit lost in a couple of versions. Those two here, but then you just have to cycle to get alternative versions and find that perfect photorealistic version. But overall, this is quite an impressive result. And as soon as he starts splitting the Prompt, don't forget to include the main theme in every part. Like here we have a fashion shoot on the first two parts. And then the third part. We didn't include that intentionally. To have a bit more of a diverse seen, we will focus a bit on surrealism and minimalism. Okay, try to experiment with that and don't forget to V role when you have very diverse Options, then, right? Roles and V roles usually produce more different results when they are, of course, combined with the Mulltiprompt. Alright, so this is our next Prompt. Here we have a street Style magazine Fashion photo. Actually, let's delete the Fashion and keep only the magazine Photo. And we have a Mexican man model, a crowded street. And of course some other important keywords specifying the camera and a few other things like depth of field. We have a portrait aspect ratio. Alright, now let's check it out and we can see a very polished and professional photo here. Let's Upscale. And as you can see, we didn't really specify the Fashion, but the word magazine photo, as you know, we didn't specify the Fashion keyword. A combination of a magazine Photo and sturdy Lighting, and a few other keywords actually work together to create this very professional look and vibe for the image. Now, feel free to try other street Style magazine photos. You can include a Fashion photoshoot combo, keyword if you want a more fashionable look. Now let's check out our last prompt. So here we have a photo of a tall Fashion model on a Fashion walk. We also specify a Fashion Runway, and we also add here a surreal dress. And it always makes sense to play around with very creative dresses, especially for Fashion type content. We also specify here I Laika Camera, which is kind of a retro Camera. And we have here a film effect, 35-millimeter Kodak though. Another important keyword here is the Tom Ford Fashion Style. It's related to a Fashion House and New York. Okay, So we have our resides. I can already see the quite unusual Versions that we have. Let's Upscale a couple. You can see that the Fashion walk or Fashion Runway is a perfect match for any Fashion Photography. And as noted before, please feel free to experiment with any different kind of outfit. Even the very unusual type Midjourney will definitely surprise you on that. Okay, So that was for the Fashion Style. Now let's move on to the next topic. 58. Photorealistic Content: Products : In this video, we will explore product photography and the few creative approaches to showcase the object. And then Outstanding way. Now let's create a dedicated channel for this. And then let's take a simple prompt to begin with. So we have a commercial food photography of a red wine bottle. We specify here a sunny summer day depth of field, and a few other keywords that should help us to increase the quality. The main idea here is to receive an eye level photo of a red wine bottle. Let's trigger this. And as you've noticed, we didn't supply any image as a reference. As with most objects, it usually helps when you provide a reference or if it's a well-known object, or for example, a wine bottle Type. You can also specify this in the description and then you will be able to get a much more closer look to the original. Of course, this will not produce the identical look of the product, but will definitely help. Alright, let's actually Upscale. Let's take the second and the third. So this produced quite photorealistic Content, also very aesthetic with nice shadows from vines and the sunny summer day vibe, which is exactly what we prompted for, of course, when it comes to the real and existing product, then most probably this image would need to be Photoshopped to attach the real label here, for example. But at least what You can do with enough Re-rolling, you can actually try and catch the closest looking bottle or an object through the actual product if you have one. This is the second version with corks here and a different bottle design, but also very photorealistic with a lot of detailing and a very nice shaped bottle. Alright, now let's try our second prompt where we have a top-down view with a sport ethic. We do specify this a couple of times. We do have the second time here, a full set of football outfit laying down also the completely empty white background and minimalist. Those two key words are also very important for product photography. And we have a bit more keywords that explain and indicate the overall type and quality of the scene. With things like vibrant colors and professional color grading, we indicate that we want something more colorful, whether the background should stay as minimal as possible, basically just white. So this is what we have. Let's Upscale. So this is the type of flat lay that you can get with products by specifying a white background. And that top-down view. Of course, if you get something funny going on on the image, you can always Re-rolling or virile that particular version. Or you can even Photoshop because manipulating an image with a white background is way easier than with highly detailed and complex background. Now, let's go ahead and try our next prompt, which is a bit more creative. And here we have a powerful explosion effect of red dust. And the object is, the product is the red lipstick. We could specify minimum, just white background. But let's experiment and try the natural and see what it gives us. We also have particles floating and fine details. This should help us to produce a bit more detailed and interesting results So this looks quite interesting. We can clearly see an explosion here and a lot of tiny particles around floating. The background is also quite detailed. Let's check the other option. Save here. And let's say that we wanted a little bit more different packaging. We can always make the variations, which is the role. Or we can also provide a product Reference. But just bear in mind, that depends a lot on the background of what the product is. So if you just have a product on a plain white background, than it will make sense to also specify here white background. But you could still include all the effects and different keywords that helped set the tone and the CME for the image. So as you can see with variations or V role, we've got very similar looking images. But for packaging and in other details, they are slightly different. So it always makes sense to try to virile, see the alternatives and then choose the best looking one. Now, let's take the next example. So here we have a pair of hipster Styles, square sunglasses on a white background. We also specify a minimalistic and modern and slick design keywords and a few other ones that help to achieve a better image quality. Let's try that. Alright, let's Upscale a couple of versions here. And we can see here a squared type of sunglasses. Also some gradient and reflection similar here. Of course the sunglasses, they look different if we compare them. They are all in a similar square Style. So if you're searching for a particular color, for example, or some particular gradient on the Lense, you could specify that in the description. You can also try Re-rolling that and repeating this until you hit that perfect variant. Sometimes this could take up to 20 or 30, or even more roles and variations until you managed to hit that exact style. But even then, if you still cannot, then you can still supply the image Reference to help Midjourney. This is the other Variation. Also all square sunglasses with a very minimalistic background and some gradient within the glasses themselves. Let's move on to the next prompt. So here we have pink and purple bath bombs on a white background. We also identify this as minimalistic and pastel color palette. The gradient packaging can sometimes help if you want to add the colors, especially if you've already specified them somewhere in the beginning. And this is also a close-up. Commercial chat. Close-ups actually work quite well. Follow the products. The flat lay and close-up are my favorite whenever it comes to some commercial products shooting in mid journey, of course. You can also sometimes specify keywords like macro lens and depth of field. It can also increase the sharpness of the main product and will help it to stand out. Right? So here's our result. They all look quite photorealistic. Let's Upscale a couple of versions. Okay, this is one of the Versions which looks like a typical and good quality photo of a product. In our case, it's a bath bomb. Same here. As you already notice that the white background and minimalistic gives quite professional and high-quality products images. Now, let's try one more round. And this time we have an iPhone 12s pro phone case with the titanium and glossy texture. And we also specify here Tech Studio, minimalistic and luxury environment, which is not a white or empty background. This actually should produce a quite detailed background. And luxury environment to usually produces a more darker and higher contrast environment. We also specify a Lense here, and we've added a 35-millimeter film without specifying which Film. This is exactly what should add a filter effect to the image. Right? Now those look quite interesting and also quite accurate. This was impossible with previous versions, but version 5.1 definitely solves the issue with inaccuracy in special famous Products. Let's zoom in this a little bit. And yes, we can confirm that the accuracy is very good. Also, we can clearly see that the iPhone has a case, and this is the glossy titanium texture Style of the case. So here are some keywords, high contrast and HDR. And also the luxury environment helped to achieve this kind of quality. And darker images. They're more suitable for a type of luxury product, or even some high-end gadgets or jewelry. Okay, So that's it for this topic. Please experiment with various backgrounds, such as minimalistic whites or even with some specific color. And then you should be able to find the perfect match to showcase the product 59. Photorealistic Content: Interior Design: We continue to explore a specific topics in the Photorealistic Style. And this time we'll prompt for different Interior Design styles. For this purpose, let's also create a dedicated channel if you can call it Interior. And this would be our First Prompt. And this time will last for a beau ** Interior Design Style. This particular Interior Design style is very colorful, vibrant, also very cozy. And if you check the prompt further, you can see that we're actually also specifying the color palette. And we added quite a lot of keywords. And we add quite a lot of keywords that specified this particular style. Things like traditional Moroccan patterns, natural textures. You can find all of those little pieces in beau ** Style. So whenever you're prompting for a particular style in interior or exterior, it always makes sense to describe those key points of those particular styles. Here we also have a photo taken by and the name of the artist that is more suitable for our case. And the most important thing for interior photos is the wide angle lens and also the natural light to get a good quality photo of Interior Design than those two keywords are very important. And if you noticed, we didn't specify the exact room here. But it is a good practice to actually dedicate what room you're prompting for. Is it a bedroom or living room? Oh, okay. Now let's check the result. So that's how the bow house style with all the extra keywords, what luck? And I find it a very close match to the actual bot hostile. I'll images, they do look like very high-quality 3D renders. So you'd want to receive a more photorealistic Style like it's an actual photo, than just tried to emphasize this like we did with our photorealistic Content, where our Prompt, our scene was only taking ten to 20% of the whole prompt, and the rest was only specifying the actual Style and the quality. And here, if you see it's not much about the description of the actual photo realism and the Style rock can also help to achieve that. And let's actually try that for next prompt. So here we specify a bit more keywords that are related to an actual photograph. We specify the Lense, the aperture, the shutter speed will also say award-winning photography. And let's add this, the raw Style to a two. And we'll also start the Prompt with a photo of an industrial inspired bathroom. You can read the prompt. Will we? I'll just trigger it now. And just as in the previous prompt, we specify with multiple extra keywords for what Style we're looking for. So we want an industrial inspired bathroom in a rooftop loft. But we also specify extra keywords like, for example, materials. We also specify the palette of grays. Here, again, the recycled materials. Okay, let's take a look. So this looks more natural. Let's Upscale. And I can already see that it fits perfectly the type of Interior Design that we asked for. We can definitely see now that this looks more like a photo rather than a 3D render. And of course, the industrial inspired style works very well here Feel free to add things and use keywords like luxury or expensive furniture, especially if you're prompting for a bedroom or living room in combination with this style, this usually produces very nice results too. Even though this does still look like a higher-quality photo. On a closer zoom, you can see some artifacts and things floating in the air. If you take a closer look. If you want to see the perfect photo realism, then just try to make variations. And then you can also try to Photoshop this style or the photo that matches the idea that you prompted for. Now let's take a third version, a third Prompt. Here we have an interior design inspired by axle reward. You can research the artist's name. Here we want to have some aesthetics from that artist. But if you notice here, we also see captured by a lineup Morales, who is another artist. So what we're doing here basically is merging different artists into our scene of Interior Design. And let's just add that photo and extra things like Fine Art inspired and minimalistic. They all adapted to a specific Interior vibe that we want to receive. So things like this can serve you as a source of inspiration for all of your Interior ideas. You can even try to experiment and merge. For example, a beau ** together with industrial inspired and etcetera. Not always, but sometimes this produces something so unique that in real life will be very difficult to imagine. Okay, So this is a very unusual looking Interior Design. Very cozy. Let's look at this a bit closer so we can see how all those keywords and explanations plays out. We could specify our raw Style to increase the photo realism of this and extra keywords that represent the photography. But I'm quite happy with the result. And also we've got quite accurate shapes and also we've got and also we've got quite accurate shapes in the room itself. Now let's try a non minimalistic approach. And this time there will prompt for a modern luxury bedroom inside a modern luxury villa. We do emphasize a lot here that this is a photograph. Have side angle professional photography with shutter speed, the Lense Settings and etcetera. We don't have too much specification of the Interior Design style here. Apart from that, this is a modern luxury villa Style. Let's see what it gives us while it's generating. Let's try our next prompt. And here we go a little bit further and we specify are basically inspiration. So inside our luxury villa, penthouse. So this is the interior Style we're looking for. We also specify an artist inspired by. And the rest goes for the photo realism and the quality overall. Let's trigger that. And now let's examine the previous Prompt. Let's Upscale a couple of Versions. And we can feel the modern vibe to it. Also the luxury touch that we specified in the prompt. Also, not too many artifacts or not needed objects. This option is even more accurate and it also has the smarter and look to it. Now let's check the Burj Khalifa inspired Interior. And we've got a totally different vibe here. Let's Upscale and look a bit closer. Now we can see the influence of the Burj Khalifa Style here, or Arabian influence. Not just in furniture, but even the colors. Because we didn't really specify the sun. But because of the extreme heat in the United Arab Emirates, this is quite relevant scene which gives a really strong warmth to it. And let's check this version. Here. We can also see the mix of modern type of furniture and Eastern like Interior Design vibes like here, and some others, tiny objects that gives this feeling in general. So feel free to experiment. You can also try to search for famous architects or Interior Designers to get some inspiration. You can also to your Prompt. If you want to receive a very photorealistic Content, then don't forget to emphasize this. And it's better to start here with the photo or a wide angle photo. And then of course, the actual style you're looking for. That's it for this video. And I will see you in the next one. 60. Photorealistic Content: Outdoor and Landscape: In this video, we'll explore the best combinations of Cameras, Settings, Lighting, and other key words for Outdoor and Landscape photography. Let's create a separate channel for this purpose. So this is our First Prompt. Let's take a look a bit closer. So here we're shooting the Nevada desert with a wide angle photo or a wide-angle lens. We have a Salt Lake Landscape. After rain during golden hour. We specify the camera position, the lights, which is natural sunlight, warm tones and long shadows. The salt works very well with golden hour, which is a sunset time. And we add a couple of interesting extra keywords here like boost, contrast and saturation, and vignette effect. Because we're shooting Landscape, the Aspect Ratio is very wide, two to one. And it also, this Aspect Ratio also works very well with any wide angle shot. So let's try this. Now. Let's check the result. Now, let's Upscale one version. And you see here how the wide angle and the desert look very good when combined. The golden hour after the rain creates these very interesting clouds. So we're getting a split color off of yellow and blue, which is a very nice effect. Now, let's just try a slight alternative of that. Because we did specify as Salt Lake here. For now, let's choose this location, which is a bit more specific also, and also a desert in Nevada. And they're called salt flats. And we keep the golden hour. But what we do now is changed the tone to white. So just try to highlight the white effect. Alright, this time we got a little bit of a different result, but the one we were expecting. So stoped flats. And you also can see here how the Cameras positioned below. Camera Position also works very well with something really vast and something really flat, like a desert in this example. And we also see everything else that we've prompted. The golden hour and the wet ground after rain. And also a bit more of the off-white tell them is. Okay, now let's try the next example. So we've got a macro shot in nature with the natural background, natural lighting. And we'll also specify a few key words here that should, that should boost the quality, like octane render, which should work very well with flowers and nature, and also volumetric lighting. But here we have the tall Aspect ratio. Now let's try that. Now let's Upscale. Let's take a closer look. And we can definitely see the different lighting techniques like moody, cinematic, volumetric, everything affecting and creating very nicely post-processed macro photo, which also looks like it has the nice light room filter on it. I think the octane render also gives this a little bit more of a polished vibe to it. Alright, now let's try another wide-angle Prompt. Here we specify that the wide angle photo is being taken from the apartment window of the awe-inspiring beauty of the Swiss Alps. We also indicate here a soft, cool light of the morning, dramatic and dynamic than again, soft and natural light. And we also mentioned the photographer, Jimmy Chen, who is also a famous mountain athlete. Let's see how that looks. Right now. Let's Upscale a few versions. This version is the closest match for what we've tried to achieve with this prompt. So if you want to keep a more casual or everyday vibe with a photo, but at the same time, make it dramatic or on inspiring. Make it with something really beautiful. It makes sense to add Elements of window view when we're talking about the landscapes or Outdoor views. The same works with the Fashion or the astronomy categories, where you specify and indicates some everyday elements in the photo. It can even, it can even increase the effect of the main scene that you're trying to expose. This is also a very nice and high-quality image, but it is standing out a little bit by being a little bit over processed or not so natural looking. Of course, we can always try to include a style raw or even try a fifth version, not says 5.1 version, which usually gives more natural and less post-processed images. Now let's take a similar looking version and zoom in a little bit and see if that helped. And we can see here a lot more shadows and contrasts. So it does look a bit more natural, but still could be improved with extra keywords or by manipulating the different lighting techniques. Let's move on to our next Prompt. Here we have an urban street Theme. We're using a new Sony Alpha A9 Camera. And we also specify the raindrops a few times. And together with ambient lights and enhance contrast and saturation, this should produce a nice result. The Aspect Ratio is also tall. Okay, let's Upscale. Here. We also have a very good example of the tall Aspect Ratio when you're picturing in urban street. So we can fit tall skyscrapers and the long streets. We can also see a little bit of depth of field and a higher contrast. The high contrast works really well on the city streets, especially in rainy or night rain Conditions. And now let's try the final prompt. So here we have another example of a wide angle photo of a picnic table in the Croatian countryside near the beach. And here we have in autumn, foggy, autumn morning, as we do specify a few times a misty atmosphere, soft glow, Lighting, sharp focus, and the natural color palette. Alright, now let's take a closer look. And it's really amazing to see how well the newest mid journey version understands the whole topic and the scene composition. Because the countryside near the beach, the Croatian BYOB and the picnic table, and also the foggy morning. All of those key, all of those key words creating the exact scene and the mood that we've prompted for. And the words like autumn, Misty, and soft glow. It's also a very nice combination. You can try it. You can try and apply this to different locations and new Landscape scenes. And try also summer vibes and also populating the scene with people. Family image would suit. A family picnic would work very well with the summer vibe here. Okay, That's it for this video, and I'll see you in the next one. 61. Photorealistic Content: Foods and Drinks: In this video, we will explore food and drinks, photography, and learn how to showcase different types of food. Let's create a dedicated channel for this. And we'll call it gastronomy. And let's start with drink photography. And we'll use a bit of a creative approach here. So commercial photography, as you know, it's quite important to showcase a product. In shear, we'd have a vegan drink, what splashes, and a powerful explosion of green color dust. So that sets the tone for our scene will also indicate a few more important keywords like studio lighting, then vibrant colors, and also professional color grading and award winning photography. Let's trigger that. So as you noticed, we didn't specify that we wanted a minimalistic background or a wide background with some types of dishes. This works very well, but usually drinks and food look much better when there's something behind them. Or the table where it stands is not empty. Alright, now let's Upscale a couple of versions. So we can see a lot of vibrant colors playing out here. We can see the explosion and the green dust. So the splashes and the explosions, they work quite good when you want to showcase some drink in the creative style. Now, let's try, oh, actually let's check this option two. So we have more greenery going on here. Sunflowers. Overall. It really showcases a vegan drink in a quiet, creative, and interesting way. Now, let's take a look at this photo. Let's use this as a reference. So we want to create a similar looking photo with a hand holding a cocktail. So all we need to do is add the image Reference link in the beginning of the prompt and then Describe what we're looking for. So apart from describing the actual scene here in the beginning, we also add things like glowing drink, cocktail ads, and also food photography and cocktail photography. What I've noticed is that when you include food photography, even your scene only holds a drink. It always creates a more interesting and vibrant seen. The rest here just describes the quality and the style of the image. A Camera, some post-processing, ray tracing, natural lighting, and etcetera of what we do have an image Reference. Always pay attention to the Aspect Ratios when you supply an image Reference so they match. Alright, so now let's Upscale. And we can see quite a close match to our image Reference and to our prompt. Now, if we would want to increase the likeness of the cocktail to our image Reference, then just modify the image. Wait, and try a 1.5 or 1.8 values. But I'm happy with the result here. We've got a good quality drink and food photography. We can see a good depth of field, high contrast, and even some influence of the Sony A7 are Camera type. Now for the next prompt, let's try a macro lips. So macro photography. So it's a creative macro food photography of multiple coffee beans on a white, minimalistic background. We specify the studio lighting, we add shadows. There's a modern vibe to it, and the rest is the color and the camera So apart from the macro type, we also added here 35-millimeter film, which should add a little bit of a filter type of fat. So whenever you need to picture some small object like greens or something similar, it makes sense to highlight them with empty background or just sort then and multiple little dishes. Or if the grain is in addition to some main dish, which also should be, of course, in the scene, looks Upscale this. So the quality is quite good, very photorealistic, even on the Zoom. And it's a very good example of macro photography in a plane minimalistic, white background. You can also experiment with darker backgrounds for coffee beans like gray or something similar. Usually they also work great. But don't forget to specify a well-lit Atmosphere or well-lit scene. Now, this would be our next prompt. So here we have a photo of a dish at a high dining restaurant. We've specified the type of dish. We add the colorful key word, Creative and luxury food photography. The scene or the background should be bright. We add a couple of effects and also a specific artist's style. Now let's see how this will Book. Now, let's Upscale. So in all our Versions, we see a high dining vibe. We see our dish. And what's important is that it's very well lit and it's a bright scene overall. Let's zoom in. And the photo realism is also very good. Mid journey, starting from version four is actually very good with organic stuff, like greenery is food and wildlife. In this case, we can see a very high-quality render of a certain dish. We even have some Creative splashes of gravy. So this type of render can also be used to get some inspiration for serving even the whole dishes. Now, let's go and try a couple of more Prompts. Here we will use a flat lay minimalistic Setup with the breakfast, poke a bowl with avocado, sliced carrot, and grilled tofu. We also specify the food photography and the artists and also the vibrant Lighting. So nothing too specific here. But a flat lay usually produces very nice setups. Also, the HDR helps to get some very nice balance between the shadows and the white areas. And yes, we can definitely see that HDR in the artist's style. He has a very good influence. Apart from that, we are getting more options with a 45-degree angle rather than flat lay. Or the third version is a perfect match to our Prompt. Let's Upscale this and see a bit closer. So we've got quite good quality of the scene. Also very high photo realism. And these little articles with the seeds and the grilled top also looks quite good. So feel free to experiment with flat lay. It works very well, especially for breakfast setups. Now let's try the final prompt. And here we have a bit more of a complex setup. So we have a 45-degree. Let's even add a dash here. Food photography of a mouth-watering burger with golden texture, fresh lettuce. And we also specify Many other things that we want to see in the picture, like cheese, etcetera. And then we'll also add the photo taken by the artists with the camera. And then all the other little pieces that should adapt to the overall quality. And it does make sense starting from the fifth version to specify all this. Of course, don't just stack up all the keywords. Choose the ones that are actually relevant and can help to produce the needed quality. So if you're chasing for more contrast, then use more studio and HDR Lighting. Things like post-production and depth of field can also work, and vice versa. If you're looking for more minimalism, that it makes sense to specify white background, even white Lighting. Remove any T words like highly detailed or very detailed. So it really depends on the scene. Okay, let's Upscale. And we can see a perfect burger that will match any advertisement. It's very juicy. A lot of cheese here and everything else that we specified. Let's check the version. Also quite similar looking. And what's more important, very photorealistic. Now, only the final tweak. What we can do here is we can increase the Stylize to, let's say 800. This can help to produce even more texture and more details. Let's add highly detailed also. And we can also try to diversify the styles. Okay, We've got quite an interesting reside with a very huge burger here. And I would say a bit more diverse, while still all of those, they were in our style that we specified here. So this is quite interesting. And we can also see here the effect of Chaos value. And if we zoom in a little bit, we can also see a little bit of extra stylizing soul. A lot of Stylize and going on here, but still within the Photorealistic range. So this is just a little tip to increase the detailing and to also try and get a bit more interesting and unusual results if you need those. And this basically applies not just to the food, but the stylization works better with scenes where there are a lot of little details or highly detailed seems. Okay, that's it for this video, and I'll see you in the next one. 62. Photorealistic Content: Public Figures: This is a special topic of this chapter where we will try to create images featuring well-known public figures from different decades. We will also aim to combine them with iconic and well-known destinations. If you think you might prompt a lot in this specific Style in the future, then it's best to create a separate channel for that. For now, we'll just continue in our photorealistic channel, which is a general channel for all our photorealistic Content. And then let's try with this white well-known type of photo with Albert Einstein with a crazy smile. Now, we don't specify any Image References here, but we do specify a full name. And what's interesting about this is that we tried to give it an effect of the '90s VHS tape effect. We don't specify any camera here, so only the VHS could give it a special look. And we also place Albert Einstein on the London Eye, which was built much later. So often you can create very interesting combinations of things that could not exist at the same time, like we have here. And I just had an idea while it's being rendered, let's actually try and add Chaos to it a little bit. So Chaos of 25, because when you do have something crazy or PFK-1 going on, without specifying too many keywords, that an extra Chaos parameter can allow Midjourney to go a little bit more funding crazy. Now, let's Upscale a few versions without the Chaos parameter. This looks quite natural. It also has this VHS or retro effect to it. Let's check this example. Also like a photo, a big, crazy and a very strong smile. But this one looks the best so far. So we're getting a bit of a retro vibe here. This looks like the London Eye. Of course, we can Re-rolling until we actually find the perfect match where we do see the London Eye or any other scene that we've prompted for. Let's actually try that. Let's see the Versions with increased Chaos. And we can definitely see that this is over the edge. It is extremely creative and strange. But this version here, I think it's spot on what the London Eye and the one didn't type buildings in the background. And a nicely smiling Einstein here. And here, we also have a nice version. The third one. And actually the first one also, the London Eye looks a bit different than the one, the background. But overall divide is perfect. And on the third option, we actually see a more closer match to our scene. Now, let's take the next example. And here we have Nikola Tesla, who is a famous inventor from the past, standing near his invention, the Tesla coil. But we placed him in the future in a very sci-fi like office. While we still want to maintain a bit more of a retro Mood by specifying a bit more of a retro photo. And the Laika, which is a retro Camera, also the film effect. Now let's see how this will work out. Alright? So I can see the top to the first and the second variant, or the closest match, where we have some elements of sci-fi and also the retro. While the Versions from below, they More match the actual time and the scene also from that decade. So this looks quite nice. And this is the more sci-fi version of it. Of course, we can make a few variations to make Tesla or closer to how we actually looks. But for now, let's try and take this prompt a little bit further. And we'll try to add a bit more of a dramatic mood or Movie effect to it by specifying a Movie still by Christopher Nolan as a separate prompt and making this a Mulltiprompt. Now, let's check what this gives us and how it affects the actual scene. Now that we can see that the effect is quite strong, and this type of Prompting with a double column, in this case, it is even amplified the effect. So there's a lot more drama and more interesting surroundings now. And this variant is quite nice here. So try to play with that. If you want to add more drama or make the scene more vast than split it by a double column, then specify a Movie still, buy a relevant filmmaker or producer. And Christopher Nolan is here because phi mostly uses Film to direct his movies. Anyway, let's move on to the next prompt, which brings us back to our time. Here we have Brad Pitt at Venice Beach and Los Angeles. So we have two things to combine here. So well-known person, Brad Pitt and the quite famous Venice Beach and Los Angeles. Okay, let's Upscale one version. As you can see, this is an extremely close match, even though we didn't provide any reference or anything else specific. This is how you can try to put Public Figures or your favorite actors or musicians to a specific location. Like here, we take musician Kanye West and we're mixing it with the Toronto tower location. We're also giving a film effect here. And we specify large floor to ceiling windows, which makes sense when you want to showcase the view behind those windows. And we've got quite natural looking photos here. The first two variants, they look much closer to Kanye West. Variant 3.4 could be V rolled it, or you could Re-rolling the whole scene to get closer to the person that you mentioned. And now our final prompt. Here, we'll try Bill Gates at Walt Disney World Resort and will provide, let's say, nonstandard surroundings and funding magic Mood, which you don't usually find in combination with this person because this is an entrepreneur. Let's trigger this and see what it produces. And we can definitely see the likeness almost a perfect match with the real person. And also the scene has the fund mode and a bit of a Walt Disney vibe to it. So try to experiment with that. Feel free to try different emotions. Smiling, angry, or sad and crying. Just add those keywords in front of the name. For example, smiling or laughing. Bill Gates. For the purpose of this course, I decided not to add anything to say abstract, and not Showcase any specific person in exact emotions. But feel free to experiment. This works really well with the latest Midjourney. Okay. Now we're finished with this video and I will see you in the next one. 63. Vector Style: Logos: Starting from this video, we'll be exploring new various cases of Midjourney. We will focus on producing new vector Style graphics and different illustrations, mainly for commercial purposes. And we'll start by exploring different Logo and icon styles. If you plan to produce Logos and Icons in the future, then I suggest you create a dedicated channel for this or even a group. In this course, we'll use the vector channel. Now there are quite a lot of different styles of vector images and logos that you can work at will only explore the basic ones and the most popular so that you know and understand how to use them. So practice when it comes to vector Style illustrations or drawings. And we'll mostly use a square or default Aspect Ratio. Now, this is our first prompt. So here we're asking to produce clean and elegant to D, or two-dimensional flat Logo. And a couple of extra keywords to give it like a steel, which is the nature insisting ability. Now, what's trigger that? As you've noticed, this is a very short prompt. And it's also a good practice to keep it quite short because only specifying the Medium and what you're looking for within that illustration. Or what batter. Alright, now let's Upscale the Versions. And you've ten. You can see that we've got a perfect, very clean and very flat Logo with a tree. And it also gives the sense of nature. We can also manipulate the shapes. So here we didn't specify any specific shape, but you can specify like a round or square shape Logo. Now let's try the next one, which is more about the technology. We'll also specify here a minimalistic style and asleep smartphone as a main object. So we have a futuristic touch and symbolizing technology innovation and connectivity. With these type of illustrations, the word vector has quite a big impact. Or things like 2D flat keywords in our previous option. Okay, now let's Upscale. So this variant looks quite nice. Of course you could specify the colors or any additional lines or drawings you want to see on the background. Those are the very simple and basic examples. So you get an idea. Now, let's try a more enhanced prompt where we specify that we want a black and white logo with a stylized globe or by interconnecting wines. But will also increase our stylization parameter here to create a logo which is a bit more detailed. Now, let's check a few examples. And we'll see here the lines that are interconnected creating the globe. Here's a bit more of a minimalistic version, but with quite actual representation of an actual land. You could also try and play with the stylization parameter. Where if you would want to, for example, decrease the lines or increase them, then at the same way you could try increasing or decreasing the stylization value. Now, let's try one more example where we also specify the color. And we're specifying an abstract shape and Creative logo. And we don't specify an exact item or object. It's just a fluid wave or mark that resembles a fluid wave. Let's see what this gives us. Alright. You can also see how the word abstract affected the shape Keywords like fluid created the actual object. And also we see the red that we had the than dynamic red lines. And they will still work pretty well. So try other shapes. You could leave the abstract shape but change the same fluid wave. You could swap for the water wind, or something more specific, like some object, and then see what this produces. Let's move on now to the next type. Here we specify that we want a squared mobile app type Logo with round edges, also has 3D elements. And our main object is a majestic lamp with intricate details, capturing the strength, dominance, and legality. You can also try to increase the stylization when you're prompting for intricate details are highly detailed Logos. The stylization can actually help you to increase the details. Okay? I think the fourth option here is the closest match. So we do have the squared Logo with round edges. We also have the tech vibe to the background and we have the line. Well, basically, it's everything we've prompted for, but the Other Versions, they're looking, alright, but not so close to our initial prompt. And the reason is because we've mixed here quite a lot of different keywords. So we want a 3D Elements. Also. They're asking for a line weld, mixing this with a nap Tech and a specific shape. Now let's experiment and let's try to remove a few keywords. Let's only leave strength and dominance. Let's remove the extra third key word that describes the Mood. Now let's, instead of the mobile app, tech Logo, let's just say a game Logo. Let's remove the 3D. And now we're left with just a square with the round edges Game Logo with a majestic lamp line, intricate details. And what the little bit of a Mood description you can leave. And also you can try the, instead of the game logo, you can try the app Logo, or even just a logo without any specification. But in this case, a majestic line capturing the strength and dominance, I think will work, but well with the gaming keyword. Let's check it out. And we can definitely see a much cooler line illustration. And it's a nice-looking logo. I think the three variants, they definitely captured the strength and dominance. However, we've lost the Logo shape completely. And that's because the square with rounded edges Logo usually is more common for an application Logo or mobile. So let's try to change that and see how this effect said. Oh, let's actually just try the app Logo. So compare to this, we only changed one word. Instead of the game logo. We specified the app Logo. All right, and as we can see, this still hasn't produced the needed shape of the Logo. Now, how we can try to fix that, we can actually create a Mulltiprompt. And now, and we'll also add a special and extra keyword mobile app. And then on the second prompt will specify again that this is a square with rounded edges, mobile app Logo. So this time we're emphasizing this quite a lot, that even the target is now more to receive that exact shape. But we still leave everything else off of the majestic lion and the associated Mood. And now you can see that we're getting much closer. We've got to perfect shapes. The second and the third variant. The first one is wrong, but the third one is also quite close and could be used. Now you can try Re-rolling and you should be able to guess at least one or even two, or even three variants with the needed shape Now, let's move on to the next prompt. So here we have the clean round logo of a coffee shop. Will spot that. The next prompt will be in artificially designed to coffee cup with or late vintage background, which basically describes the object in the scene. And only the third option describes the actual Mood and styling a little bit this vintage inspiration. Now, let's try that. So as you can see, we only have a Logo in one of the prompts. In the first one. And the other two prompts are only describing the scene and the mood. And we've got quite nice results here. We see the ornate background at least on three options. Also, a bit of a vintage inspiration. And in one option we actually got even a, what I believe is the coffee shop in the background. So this is an example of how you can combine using the Mulltiprompt and specifying the type, for example, of the, this is a logo of a specific shape in Blend prompt. And then use the next prompt to specify the scene or the object. And then you can add a third prompt to further describe some characteristics of the logo. Okay, now let's explore a different version of the Logo, which is called a mascot logo, which usually represents or has an animal in it, or a person. And it's very often related to sports and e-sports, in our case, is an electronic sports like PC, gaming tournament, and etc. so it makes sense to use that with some animal or do a character if you have one and have some dynamic poses or actions that can also have some vibrant colors in it. Let's prompt that. Here we have a perfect example of a mascot logo and a fierce animal illustration. What in it? You can try different animals, like alligators and lions and tigers. Some wildlife hunters. And those work quite well. But let's move on right now to our last prompt. And here we have the pencil hand-drawn and botanical Logo with other details describing the actual sinking. But the main point is that this is a pencil drawing botanical Logo. And let's see what this produces. And I can already see this produced nice results, especially the third and the fourth. Let's look at it a little bit closer. This is a very nice example of handwriting or pencil hand drawing. And because we specified this as a Logo, this object does not occupy the whole picture. So it's more centered and it's quiet. For example, in a round shape, then it could be easily used as a Logo. Also Logos or Icons. They're usually more finished than the image. But I mean the prompt that it doesn't have any exact specification. Okay, so we're finished with Logos. Now, we're going to try in the next video and prompt for different icons and icon packs 64. Vector Style: Icons: In this video, we explore how to create nikon, a sticker and icon in sticker packs. Let's continue in our vector channel. And now let's try to create an icon of it in a road sign Style. So we want a clean icon in a road sign Style. And our topic is wild life ahead. We also specify that we want minimalist detailing, so we don't want to have too many details in it. And if you extra keywords like traffic control and safety. Now let's see what this produces. Okay, and the first 3D variants, they match our Prompt very nicely. Especially the first and the third variant. We can even see the arrow that the thing symbolizes the arrowhead keyword. This is definitely a road sign Style. So this is quite basic example. Let's move on to the next one. Here we have a stylish, rounded and modern 2D flat icon, and we have a laptop as the main object to specify the minimalistic design, the coloring, and also the technology in the and the productivity keywords. So those should produce the needed design. And the main idea is to get the flat icon, also the technological styling overall. Now, I would say all of them, all of the variance, there are a perfect match to our Prompt. We have a stylish, rounded, and modern flat icon. And you can also notice that there is also a difference in how Midjourney draws icons and logos. So try to experiment and see what works best for your design, either a logo or icon. They produce similar but still distant aesthetics. Now let's move on to the next icon. So here we have an isometric icon with a cityscape. Also we say that it's immersive perspective. And then with other keywords, we specify the scene and the mood. That is being the main keyword here, isometric icon. Now let's see. And what we can see here is that those images are not really Icons. Those are more of isometric illustrations. Now let's try to fix this. So what we do here is we add an extra keyword and change the wording a little bit. Okay, so we add that this is around icon. And instead of featuring will say with vibrance cityscape with towering buildings inside. So this scene should be inside a round icon. So this should be more logical to produce an actual icon with scene that we've specified inside it. Okay, now let's check it out. And now we can see that our fixed worked perfectly. So in case you ever encountered this issue that you're getting more of an illustration than an icon. Then try to add more specification to what kind of icon this is and insert the scene inside. So this thing, this is an inside of eigon are cityscape. This also worked. And also, if that doesn't work, then instead of the icon, try the word Logo. Sometimes it also helps. Now let's move on to the next prompt, where we have the rounded icon showing adorable little pandas. Actually, let's change it to a panda. Then again, we specify that it's rounded contours and then characteristics of the scene and the mood and etcetera. So let's try that All right, and this is the perfect result. So our Prompt has worked. But now, what if we want to create multiple similar looking prompts? So let's say we want pandas, but all of them are slightly different and all within the same style. And what we can try to do is just add a word pack. And here we have let me, yeah, so we have an icon pack here and we'll also change the Panda, two pandas. The rest is absolutely the same. And now let's see if Midjourney actually understands this. Now let's check it out. And this worked quite good, at least in the two Versions we see. And those are different icons in quite a similar style. This may not always be the case, then you just have to Re-rolling or emphasize on the icon pack a bit more, basically specifying this a few times in different wording in the Prompt, or even creating a Mulltiprompt out of this. Let's take a bit of a closer look. And we do see here quite different variants. But what's most important is that all of them are in the same style. So this is the Re-rolling. And again, we're quite close and we even get better results here. So we have three options with pandas. And of course, what you can do to further diversify and receive more different versions, you can just create a variation. So let's Upscale, for example, the first version. This is what we have. Then what you can do, You can make variations from that. And now, look, now let's look a bit closer. Maybe not so close. And what we can see here is that we now have a lot more different options to choose from. And we can see that for example, if we take the first column here, that we are getting a little bit similar, at least in terms of the coloring. But the style is quite different, especially on some pandas. And then what do you can do? You can, for example, pick the version that looks the most different from the initial prompt, and then make variations of that again. And then keep doing that until you keep getting an extremely diverse result. So basically you can repeat that process without limits and then you can collect a virtually unlimited amount of Icons on that topic. Now, let's come back and let's try to prompt in a bit of a different way. Also for an icon pack. Here we specify high-tech and minimalistic design. So that's our main topic. So to D flat, modern high-tech Icons, we specify a bright gradient, symbolizing technology and productivity. And then we split it. And in the second prompt, we add that it's just a to D flat icon back, basically forcing to produce images in this exact Medium. Now, let's try that. And what You can also do, instead of specifying a to D flat icon pack. For example, you can always go back and choose, for example, and adorable keywords. If this is related to some Game Icons or games for children, or something similar. Or you can also specify things like pencil hand-drawn, or anything. Midjourney understands any Medium. Oh TE, let's look a bit closer. So if we look at this variant Here we can see that not everything is related to a high-tech or technology, but at least half of those icons are. And then again, what You Can Do, you could just Upscale this and then keep making variations with collecting only the ones that are suitable. And in this way, you can build a whole collection of icons in similar style without re-prompt doing it by just Re-rolling this. Now, let's try a final prompt. Here. We're actually using something a little bit different. So we're using a sticker instead of an icon or a logo. And our mood and seen is a beaming summer sun with the vibrant colors, dynamic lines. And we also add keywords like joy, positivity, and the print designs, which is a quite important keyword. When do you specify something like a sticker? Because stickers can actually then be printed out on a special paper and used as physical stickers. All right. And as you can see, we have got a perfect match to our Prompt. Also the sticker, it has an edge around. Sometimes it's a different color edge, but that is one of the differences between an icon and a Logo. Now, let's imagine that we want to get a pack of those stickers. So if we keep the same prompt level only at the word pack, let's actually change it to stickers pack. And we also add a white background. And let's see how this will help us. And you could see here that mid journey understood our Prompt and created a whole pack of similar looking sons and even some other objects we didn't ask for. But what's important is that everything is in the same style and aesthetic. Then what you can do, of course, is Upscale the version. Take a closer look. And if that is the style that you're looking for, then just keep making variations to create even more stickers within the same style. Alright, let's take a bit of a closer look. And if we take this example, the first sticker in the row, they all are quite different. Others, they are also different. So we even start to see some new objects, but everything is within the same styling. Okay, So that's it for this video. And I will see you in the next one. 65. UI Elements: Now, we'll learn how to create visually captivating UI elements for applications and websites using Midjourney. Additionally, we will explore how to create mock-ups for both digital and physical products. Those Mock-ups can be then showcased to use later in screenshots of an application or website. Or for example, to represent a physical Canvas Painting AI-generated environment. So let's start by creating a dedicated channel for this. And let's call it UI and Mock-ups. And let's begin with a simple example of creating Elements for a website Application or even for a book. There were a lot of used cases when you needed specific similar looking Elements to be added in a digital product or even in a physically printed product. So here we have a vintage inspired website design. For French winery. We have ornate, delicate illustration is vineyards wine bottles, and we specify the tone and the color. So this should help Midjourney to understand the vibe that we're looking for. And because we specified that this is a website design with multiple elements and it should understand and create a few elements. For this case, a small note, everything we generate here, and also everything we generated in our vector channel in previous videos. Those are all flat images. So this has just been xi or JPEG image. And you would need to use an external imaging editing tool to actually finalize it, removed the background, for example, and prepare it for printing. We will cover the basics of how to do that in future videos. But for now, let's focus on the actual generation of those elements for different use cases. Now, let's look at this a bit closer. And what you can see here is that we have those elements. So we could cut out the top part here and use it to be the phi, the header of a website, for example, or Logo, than the image of a winery, which we asked for in the prompt. And we have different bottles. And if you noticed, the text is just a random generation. So this would have had to be edited and if Photoshop. But once regarding the design elements, this is a perfect use case when you can actually prompt, specify the style, and then use third-party tools to finalize the design. So here we had quite a basic Prompt. Now let's be more specific and specify that we want UI elements. For example, for an application. So this is the style that we're looking for. We specify UX and UI, which stands for user experience and user interface. Then the design kit Elements and a couple of extra keywords that can help to shape the styling of those elements. Now, let's try this. Okay, Let's take one example, Upscale this, and look a bit closer. Here we can see different types of elements, but all of them more or less in the same style. Then what you can do to receive more diverse and different elements. You could then just to make a variation of this image. And by doing this, expand the amount of elements that you can actually get in the same Style. Now let's try a little bit more alternative prompt So here instead of the sleek, we've added asleep to D flat design kit. And now we have the red lines keyword. But apart from that, we also specified a little bit higher Chaos value of 15. This should produce more different styles. And the stylizing is also increased. So those two parameters of increased Chaos and stylizing works and applies really well for any UI elements or any other elements, including the Icons, for example. Now let's look at this a bit closer. Let's take the first version and we can see that it's a little bit blurred. So let's check the second one. This one is a lot more sharper. Now we also see the red colors influencing all the elements. And we can already see some useful elements like buttons, circles, lines, and overall, this acts like a very good Mood Board, which can be used further by designers to use this as a source and create some very specific shapes. If Midjourney can't generate that. Let's check this variant. So same here, plenty of elements. And in a future chapter, we'll look into how do Upscale that. We can actually increase the size of the element, which should also impact the quality in a positive way. And now let's take the same prompt and modify it a little bit further. So here we add a white background. And we also specify without shadows, but not through the no parameter, but actually within the prompt itself. While you're not very strict with removing certain objects or elements, you can actually specify some simple keywords. And it will also increase the correlation between those two words. I a dash between them instead of a space. So mid journey doesn't interpret this separately, like without any shadows, which could have a diverse effect, will keep the Chaos and the stylization. But we'll also try to prompt that in the niji model. The fifth version, which is the latest currently, this in theory should produce more of painted style Elements. Alright, now let's Upscale. I think that the first two versions are very close. Let's open it in a browser. And again, here we see that there is a big variety of different elements. We see the white background. We don't see any shadows here. And this is the model in luxury, the fifth version. So try to experiment. If you don't, if you're not getting the immediate results, you can also, you can always try the niji engine. So now let's take another example where we also described that we want interface elements for software or website in a futuristic theme while leaving supply things like UI buttons, Icons, and UI elements. This makes sense when you want to see certain elements in the image. Then we continue specifying our Style. We also add colors which are purple and blue and also the gradient. But apart from this, we actually specify, Let's check this image. So we have a reference of the elements that we want. But because we've specified a lot of keywords here that describe the actual element kit. This reference should only Guide mid journey and help to understand that we are looking for the different type of lines and separate objects in this kind of style. Alright, now let's look at this a bit close here. We can definitely see the influence of our image Reference and also all the keywords from our product manifesting themselves in the tiny details here. So this can then be up-scaled further and cut away from the image using Photoshop and then used for a website or an application that you're building. Now, let's take a step further with stylization. Let's say that we want even more of a futuristic look. Then we can add keywords like cutting edge will keep the UI design elements here. We add the word kit, which should help to separate those elements. Then again, we'll say the futuristic Elements. We add things like glowing buttons, holographic displays, and also the sci-fi settings. We also add a parameter here with the stylization of 750. And let's see what this can bring us. Now. Let's see, let's zoom in this. And we definitely got a very sci-fi and futuristic design. All those elements, then it can be cut and used in the application, a game, or website. If you do get a blurry option like here, then you can also try to make a variant or just use another version. So all of those elements and panels can be nicely implemented in some futuristic project. Now, let's try our final prompt for this video. And here we have a hero section for a military website. But we also wanted in a UI design style which should produce something similar. But we felt want in Elements, we want more of a mix in the artwork and the UI design. We also want to specify a wide aspect ratio because this goes on the top side of the website. Now, let's Upscale some version that looks the most pleasing and take a bit of a closer look. So this then can be used on a website or some other applications. The text here and the numbers would have to be replaced in Photoshop or similar software. But this is just another example of how you can combine and use mid journey to populate your application or website with your content. Now let's move on to the second part of the video, where we create different types of Mock-ups 66. Mock-ups: Now let's proceed with the Prompts that generate Mock-ups for apps, websites, books, and paintings. We will continue using our dedicated channel. And let's say that we wanted to create a mockup for E buck. So we would need to use an e-reader or an iPad for example. And in our First Prompt, we have a top-down close to view angle of a product mockup, product mock-up showing of an epoch. And we have an iPad laying on an office table during the summer day. So we don't specify too many keywords about the photo realism of this picture. If you do need to generate a very real photorealistic of this content in the mockup. Then of course, you need to specify all those keywords, including the camera and the Lense, and also the Lighting and etcetera. So this is a very basic example. Okay, So we have the result. Let's look at this a bit closer. Let's take the second option. And here we have a nice setting which gives this summer vibe a very well-lit room. So this would be ideal for any book or a story that gives the same kind of emotions and feelings to the reader. Or you have to do here is just cut in the cover image of the book. Now, let's take, let's check another example. So here we have a more digital story, which is a medical app UI design. So we focus in the prompt on the actual design elements. We don't specify too much for what the style, the surroundings are, etcetera. So this should produce quite simple and straightforward mockup. Now let's Upscale a couple of Versions. And once again, here you can simply place the screenshots of your application or even a website. And because we specified that this is a medical topic, we've got quite a suitable color palette to it. And another way of using this could also be concept design or concepting. Where you don't have an application yet, but you have an idea and you're looking for a feel and a vibe for the app. This can be used quite easily as an inspiration for a concept for the future application. Now, let's take a look on another example. So here we also have an application. We have a top-down view with a minimalistic and luxury background. And we also add here, that'd be hands and Figma and dribble product design. So this is quite important and very relevant to anything where you prompt for UI, for an element, or even a mock-up where digital product is presented. And this here is a photo. Let's Upscale a couple of versions here, two. And we can see that this is a photo with the luxury vibe and black or dark colors, which could suit very nicely on some financial app or a similar commercial application or website. So here we have the two. And once again, those Mock-ups can be used to highlight any web or app-based application. Now let's check the next prompt. So here we have a high fidelity 3D rendering mock-up, presenting an application or website on screens of a smartphone or a Mac. So we have our realistic lighting and showcasing professional digital products. That's one of the most important keywords. Let's see how that looks. But basically the idea here is to represent the same application or a series of applications on different gadgets and screens. Okay, Let's check it out. So the first couple of options look very interesting. And they match the idea and the theme of our Prompt quite close. Let's Upscale. So with the mock-up like that, you can represent a digital product or even an e-book or a movie, or anything that is accessible through those gadgets. All you have to do is to cut those images from the actual screens and just insert the screenshots of your product here. And that is the second version, which only has two screens. But I find it more aesthetically pleasing, especially for any high-tech or digital product. Now let's take a look on the next prompt, which should produce not a photorealistic, but more of a digital mockup, which also works quite well. Oops, what happened for applications and digital services, or even UI design and concepting. So here we have a flat digital mock-up that elegantly showcases various apps and websites. And we have a clean and minimalistic aesthetic. The well-organized screens Showcase gadgets in an organized way. We indicate things like shadows here and the highlight user interface. And as you see, this could be very suitable for different pages of an application Showcase, for example, or even the product Mood Board or concept. And now let's take a look to the next one. So this should also be a digital mock-up. We don't specify that this is a photo, and we use Mulltiprompt here to highlight the desktop and mobile and tablet interface. So this should produce multiple gadgets. We also specify here coloring and that the interface should be friendly. Now let's see how this would look. Alright, let's look a bit closer. And we can see here that we have multiple gadgets here, including the iPad, smartphone, and the computer screen. So here we have another example of showcasing any application or digital service. We can also try to play around and change the background and the colors of the background to make them closer to the product that you're showcasing. Now, let's look at the next prompt. And here again, we're prompting for photorealistic Content. But instead of showcasing a digital product, this is a mockup for a painting. So this is an actual physical painting. Let's try that. And let's quickly go through the keywords. So we have a boss Style, Office table setting, self-development Theme, Fashion, volumetric lighting. It is a photorealistic and professional photography style of content. We also added raw photography featured on Unsplash and product design photography, which I believe is one of the most important. And here we also start to specify the Aspect Ratio because I think in the previous one, those were just square. Okay, now let's take a bit of a closer look. And the second option looks like a very close match. Now we do have a slight frame here. Of course, you could make variations of that picture and try to find the variant without the frame because it usually doesn't have any. Here's another nice setting that we have with dark background. And basically this works the same way as with the application Showcase. So you're just inserting your image on top of the existing one with a third-party tool. And then blending it together by lowering the contrast and brightness and things like this. Okay, now let's check another example. Where instead of the photo we have a 3D render or framed poster mockup, standing on the floor in a well-lit, luxurious, modern Interior. And we also increase the stylization a little bit here. So this should produce a bit more interesting result. And we'll also have a tall Aspect Ratio. Okay, Now let's take a closer look. And you can see that we didn't get too many options with the actual painting standing on the floors. Of course, we could recycle that until we get one. But we also have quite nice Versions from this. And all the paintings are foramen, which was the main point of our prompt. Alright, and now let's just try our last Prompt. And this time we have a Book Cover mockup. And we will try to add in the background of beach with the fluorescent and luminous waves, which is a nice phenomena that you can spot on mag depths, for example. We'll also add a style raw this time to give it a more natural look. Actually, let's add that this is a photo of a in this. And this should produce interesting results. Now let's try another version. But this is quite interesting, especially the first and the third one. This could suit some magical story or a romantic story. The third, the last option here is not really what we were looking for. And here we also have a couple of interesting Versions, like the second and the third one. And here also, and here instead of cutting away the existing cover of the book, the material cover matches your initial idea. What You could also do is if you have the title and the design of the cover on the transparent background, you could just lay it down on top of the existing cover here. And then this could create a very nice effect even for the digital books. This could also work as your main mockup that then could go to the stores and etcetera. Okay, we're finished with this topic and we're moving onto the next one. 67. Illustrations: Comics, Cartoon and Anime Part 1: Starting from this section, we will delve into new and exciting mediums, such as character illustrations, Fantasy Style paintings, and various other tabs of illustrations that mid journey is capable of producing. Now let's begin with creating a dedicated channel for this. Oh, and one more thing. Let's move the lighting and atmosphere for two more photorealistic, I think it more suits here. Let's just quickly check that. Yeah, so try to keep your channels organized. Okay, now let's switch to illustrations and let's begin Prompting. Now, this is our first prompt, which is a retro Cartoon Style prompt. From now on, I will not be reading full prompts. You can pause the video to read it fully. But now you should be able to understand the impact of most of the keywords that are in the prompt. And the main idea here is the cartoon style in a classic and retro collage. Let's open it up and we can see how the vibrant color and the action, the energy and the excitement, all the keywords create a very energetic scene. We can examine that further. For now, let's continue to the next Prompt. Here. We will be trying a Disney inspired Cartoon Style. And we'll also try charming characters here with captivating expressions. So this is a fairy tale style illustration. Let's see if MID journey understands that. And as you can see, mid journey is very well aware of Disney style and fairy tale illustrations. So here as everywhere else amid journey, you're only limited by your imagination. So feel free to run for specific scenes or actions in this style. I would add that specific description here or before this phrase. So somewhere in the middle. But don't make the sentences too long. Now, let's go ahead and try our next prompt. And in here, we actually have a collection in anime style. We describe what kind of aesthetics we want. So let's trigger that. And also let's take the same prompt. And now we will try that in the niji v5 model. And then we can compare. So this is the version 5.1 and it's a bit distant from the actual anime aesthetics we prompted for an anime celebrity. Let's now check out the niji version and we can definitely see a much closer match here. Now, let's try to improve this a little bit. So we'll change this incorporating expressive eyes will change too big eyes. And also, we'll try a special style that is available with niji, which is CMYK. There are two more styles available, the cute and expressive. Feel free to try both. Let's check it out. We can also see a much closer match with the anime aesthetics. But I think the previous one was a closer match to the actual celebrity feel of an anime celebrity. So try to experiment with those. And we're going to the next prompt. And we'll try a caricature. Where are we Blend humor and artistic expression. So hand-drawn caricature of a celebrity. This could give a very funny and interesting result. I didn't specify any exact celebrity here, but feel free to indicate and play around with certain names. Alright, so this is the type of result that you can expect from this kind of Prompt. Try to explore with different Tan drawing styles like pencil Or even a quick sketch style that could also be interesting. Let's move on to the next one. Here we have a 3D Pixar inspired Cartoon. And we have animals here as the main object. Lifelike textures and vibrant colors usually match quite well with this style. And once again, you can be as specific here in specifying exactly what animal or animals that you'd like to present. Otherwise, mid journey would just randomize that for you. Okay? So we definitely see here the influence of Pixar, which is a famous studio for creating 3D cartoons. And of course, the animals and the overall mood match quite closely with our prompt. And I think the detail in gear is quite strong and also very accurate. Okay. And let's move on to the next one. Okay. Here we have cheaply Style character with an oversized head and cuteness. So the GB Style is quite cute by itself and can be used to produce cute characters. Or you can even specify some well-known people in this style. And this usually produces quite entertaining results. By the way, let's actually try that. Let's first examine quickly. So this is the tab of illustration you can expect from a GB Style. Now, let's just quickly try one thing. So here we boot a public figure who is Bill Gates in our example, feel free to boot any other person. And let's just see what this would do to the character. Alright, so feel free to experiment. Also. If you want to match it closer, you want to find a variant that matches it than just Re-rolling that or V roll the closest looking image and find the perfect one. Alright, so that's it for this video. And let's move on to the next one. 68. Illustrations: Comics, Cartoon and Anime Part 2: In this video, we'll try a few other famous comics and illustration styles in both niji and version 5.1. Let's continue in our illustrations channel. And First, let's try a manga comic style. So here we have superheroes in dynamic poses. And we do specify manga a couple of times. Okay, Let's open it up. So this is another illustration type that you can use in your Prompts. Experiment with more specific scenes and objects. But it works best with themes like superheroes and some actions teams. Now, let's try the same Prompt. But here we specify on niji, the fifth version. And as you can see here, we've got pretty different results. And also I could say that they are quite clean results, so we don't see anything else being generated or randomized. It would have compared to the fifth version. So we only have the superheroes, which is what our prompt is about, only superheroes. So no scene or a specific background. In the default model, we see those things being added or randomized. Okay, now let's try another Comics. So here we have retro sci-fi comics, and we have spaceships as the main scene and object. But the main difference here is that this is a hand-drawn retro sci-fi comic. And this also has very nice aesthetics. Feel free to experiment with everyday scenes or casual objects. There'll be drawn in quite an interesting way using the style. Right? Now. Let's check it out. As you can see, it's very detailed. We could even lower the stylization a little bit here. But we definitely got the vibe that we prompted for. And of course, the sci-fi and the spaceship Theme. Now let's try the same width, the niji lateral. Whoops. Okay, Let's check it out. And to my eye, this looks much more aesthetically pleasing and also a lot more closer to the actual Cartoon Style. So that's why it always makes sense to try it in both engines. Wearing your prompting for something that's related to cartoons or Comics. So try the default version of Midjourney and always try the latest niji version. So in our next prompt, we have Pop Art and quite a simple scene. So it's just an office space in vibrant colors. Let's see how that looks on 5.1. Let's look a bit closer. And I find it quite nice. A nice mixture of pop Art and then Office Theme in fiber and colors that can be used as a poster in an office or some other room. Now let's go ahead and try the niji version. And as I can see here, our initial prompt, the model five-point one, in terms of the coloring and aesthetics, is a bit closer matched the Pop Art. So that's why it's best to experiment and always try to models. Because for some Styles, niji does much better. And in some styles, the default train model understands and does better. But here we see some fusion of those styles and our prompt, which produced another very interesting result that of course, could also be used as an illustration or even a poster. Now, this is our last prompt for this video. And let's try that in the standard model and also in the niji. Here we are of the black and white pencil drawing style and the scene is safety road signs. Let's see if Midjourney understands that. Alright. So the default version, understood it quite well. This is a perfect example of pencil drawing. Now let's check the niji. And we can also see that in this case, also the default model produces a much closer result to the actual style that we asked for. Also the black and white Theme. And in the niji, we got a few colored elements and the pictures look more digitally drawn. Again, as always, feel free to experiment and use those styles with different scenes, objects, or even people by providing the links to the images as a reference. That's it for this video. And I will see you in the next one. 69. Illustrations: Game 3D Characters: In this video, we'll explore the process of generating game character concept designs. Additionally, we'll learn how to create Game Assets. Now, let's launch Discord, and let's make slight arrangements in our Server. Let's rename the general channel to chat because we're not going to Prompt Here basis just for chatting with your friends. And let's take our test channel and let's rename it to general. So we can Prompt anything in here that doesn't fall in any specific category that we already have. Now let's start with our First Prompt. Here we have a concept design sheet with a 3D Style video game character. We'll also try and get an isometric view from this keyword. It doesn't have to necessarily be an isometric Style. But this helps to set a certain perspective for all the elements. And other important keywords are 3D, miniature and modeling, and the cute character style. So those key words just set the style that we're looking for. And the idea behind a design sheet is to get multiple variations of the character in one version. So this way, all the versions will be in a similar style. Alright, let's Upscale, for example, the third version here. So here we have a nice concept design of a Viking. This thing can be further used by the game developers and designers to create a character out of this. Before you choose your result, I would recommend that you recycle at least two or three times to just give the possibility for mid journey to express itself. This is crucially important when you have a Chaos value higher than 30, right? So that's another version. And here we have the same character. So the first, second, and the last one, they're very similar, but they have to be viewed from different angles. Okay, Now let's try our second Prompt. And here we're prompting for a famous fictional character. We want a full body model. And what's important here is that we specify at, for 3D. The Maya 3D is just another variant of 3D rendering and modeling software. We have Art station, Behance and DreamWorks. They all have their own aesthetics. When you mix those, it gives a nice combination and certain aesthetics. Okay, now let's check those height. And I can already see that the second version is quite nice. So we have a view from the front and from the back. And we also see the influence for this specified character, which is a blonde girl on our concept character. And feel free to experiment with different elements. You can specify a red dress when you want to get specific type of element on a character. And we'll move on to our next Prompt. Here we have a cybernetic and futuristic type of character with an exoskeleton. And what's important here is that we specify a quite well-known artist, you Jie Xing kawa. You can Google to see what kind of aesthetic C has. And here we specify an additional stylization perimeter. A little bit of increase in the Chaos. Alright, let's look at this a bit closer. And I personally like all the four Versions. And we can see that all four Versions here match our Prompt and styling. Maybe too much stylization going out with the first one. But the second one is perfect. And also the fourth option where we can see the actual exoskeleton applied really nice to the character Now let's move on and we'll try to generate Game Assets with this prompt. So here we have an enchanted amulet. You can tell what styling it is. And a couple of special keywords like game icon and Game sheet. Alongside the style of Hearthstone, will give us quite typical and good-looking assets. Alright, so let's take the second and the fourth version for here. And if this is the style that we're looking for, then basically those are ready to use. You will only need to cut it away from this background using a third-party tools like Photoshop. And those will be ready to use assets that designers and Game Developers can implement this directly in a game without any overhead. Let's now try the next prompt. And here it's a bit different. So we're prompting for Game graphical user interface mostly will want to menu assets. We do specify extra keywords like widgets and buttons. And in this example will only Describe the theme a little bit by just saying that this is a Gothic still Theme. Of course you can take this further and Describe the textures and colors. Also, a white background or light background like here, helps sometimes when you're cutting the actual elements out from the image. Now, let's take a closer look. So all four Versions here have quite good quality. Maybe only the fourth one is a bit blurry. But then those elements can be used as a background to the menu or as a button or some widget. You could also remix that. And if you like the Style, Let's say you found that this is the perfect match, but you want to change the color or something, then you can use the remix mode. And of course, any letters and words you find what we have to be Photoshop to and removed. And we will cover this topic in our next chapter, how to use the basic tools in Photoshop to do that. For now, let's trigger our last Prompt. Here. We're asking for different Prompts, different game Prompts. Those are just different elements. We don't specify which ones, we just say different. And we emphasize multiple times that these are game assets. That this is an asset sheet. And online games Style. I think the gold painted will influence the actual prompts that will get. But always specify the Chaos a little bit higher, just to get more variety and look what this can bring us. Okay, now let's Upscale. I think this version, the fourth one, is quite good. And let's take a closer look. And we see here that we have plenty of different elements. But all of them in the same Style. Look at also, go back and just make variations of that. And with this, you can expand further the amount of usable Elements. And then you can just repeat that until you get the needed amount of Elements. Let's zoom in. And we see that the aesthetics are very much the same. So everything is very alike. But we do have different props and Elements on all Options. So with version Nin, this converted to this here we even have a totally different set of elements than this corner here. So those were quite basic examples. Just to give you an idea of how to get the elements, concepts, Prompts, even character designs for dames, and they cut those away and actually use it directly in the game or an application. Okay, that's it for this video. And I'll see you in the next one. 70. Illustrations: Other Types: Now let's dive deeper into various types of illustrations that can be utilized for different purposes. For example, to print out posters, create canvas paintings, or use digitally for applications such as online advertising campaigns, e-book illustrations, and websites. For this, Let's go to our illustrations channel. And let's begin Prompting. So this is quite a basic way for prompting for an illustration. We don't specify too many styling keywords here. Apart from this being a digital illustration, we're more focused on describing the environment and the actual scene where prompting for. So let's see how that looks. Okay, now let's look a bit closer. So this is how a basic illustration looks on the version 5.1. And we'll also see the perfect match to description of our scene. Now, let's be a bit more specific. With our next Prompt, we're starting to describe the style a bit more. And we'll try and Art Deco like a vintage retro style. We'll also describe the background and surroundings here, which are the ornate patterns. And if you extra keywords like charming and nostalgic, Artistic illustration should help to achieve the needed Style. And then we got a quite nice mix of those styles. And for example, this illustration is almost perfect. With just a little bit of Photoshopping. We can polish it and then it's ready. Same here. And if you're not happy and you see some strange faces or artifacts, then just keep making variations until you find the perfect one. And we're moving on to the next prompt. So here we have a tribal tattoo design. We describe it a little bit. We emphasize on the tribal Art. Let's see how this will Book. Now let's take a closer look. Here we have another good example of illustration that can be, for example, printed out and used in a tattoo shop. And we see a nice blend of the tribal Art tattoo with both lines and symbolism. So if you're prompting for such illustrations, then try to mix and match different styles by specifying multiple elements. Now let's try the next prompt. So here we have a playful pop-art style illustration. So we don't really specify an exact object for the scene. We are more focusing on a style over here, which is a combination of classic fairy tale and Pop Art. Also, what I think is an important keyword is igniting children's imaginations. If you've noticed already, starting from our First Prompt Here, we're having a little bit of a different approach to the Prompts. We're expressing ourselves in quite a logical way, similar to how we would explain this to a person. And if we compare to, for example, photorealistic channel, than we can clearly see the difference in our photorealistic prompts in our previous chapter, where we have a more keyword like approach. And as soon as you start polishing your Prompts, always try to remove as many unnecessary words as possible. But at the end of the day, of course, it always depends on what medium you're working from. M for these type of illustrations. A more lyrical approach with just a few key words for describing the actual style that you're looking for. Works really great. So for this example, we see white allies, Vicks between of pop-art style mixed with fairy tale, very rich compositions. And we can definitely see how keywords like dynamic Composition and expressive characters manifest here really well. Now let's try a different approach where we prompt for our botanical illustrations in those specific style of this painter and botanist from Belgium. You can research the styling a little bit and we want to see different forest mushrooms. So this should produce not one mushroom per image, like a little Guide to forest mushrooms. Okay, this second version is a perfect match. Now let's take a closer look. So these mushrooms then can be cut off from this image and used as Elements for different purposes, like books or other illustrations. And if they would make variations from that, then we would see new types of mushrooms appearing here. I don't think that absolutely every mushrooms should match the real existing one. But because there's so many varieties of mushrooms, I think it's possible to find in every version a close looking real mushroom. And we can see that this versioning, we're getting slightly different mushrooms on every version. Anyway, let's move on to the next prompt. And here we're prompting for a graphical illustration with a financial chart or relation graph. And we have keywords like representation of a complex dad and data. And we have specified that we wanted this in a modern and minimalistic illustration style. All right, What's take a closer look. And we have another good example here of graphical elements. This could be used for a website. Also, this could feed quite well into any, any financial PowerPoint presentation. Now let's try our last prompt for this video, which is quite a specific one. So we have a vintage postage stamp. We're specifying that it should be in a blue ink Style, and we'll try to get it with Abraham Lincoln, those posted stamps. Alright, let's take a closer look. And I think the first option, basically all options are giving very high-quality retro stamp designs. Of course, the words we need to be photoshopped or recycled until you're getting something meaningful. But that's quite rarely. That happens rarely with the current version. In the future of Midjourney Versions, this should be fixed. Okay, So this was just another example and use of mid journey. There are hundreds of types of illustrations that exist. Here. We're covering only the basic ones, or let's say the most popular ones. And with some exceptions to prove once again, that you're only limited by your imagination. Though, feel free to try any other type of illustrations that we didn't cover yet. Just don't forget to match your keywords with the style that you're looking for to help mid journey understand you better and to receive a more professional result. That's it for this video. And I'll see you in the next one. 71. Photorealism and Fantasy: In this video, we'll explore two intriguing Styles, photorealism and Fantasy. Additionally, we'll try to blend our Style description with aesthetics of certain famous artist. Let's continue in the illustrations channel, and let's start Prompting. So our First Prompt features space and astronauts venturing into the unknown. We also have a specific artist Ashley would. So let's see how this looks. Here. We don't Blend any specific style, so we're mostly relying on this artist is aesthetics because we didn't Describe what sort of illustration or what type of illustration this is. Okay, let's take a closer look. So those are all nice paintings of an astronaut with this very specific styling. But those can be printed out as a canvas or framed paintings or simply as a Poster. You could try different scenes with this artists, especially anything related to space or futurism. Now let's take the next example. So here we described that this is just a Fantasy Style, which is a quite broad description. Then comes the scene where we have someone fighting. We have them in the cyber punk say, which is the futuristic style. So we're kind of merging two decades or two different times than we described the scene a little bit more. We don't supply any artists here, but we do Prompt this in Nietzsche version. Okay, Let's check it out. And you can see that Blend here. Cyber punk vibe, the colors, and also some aesthetics of Asian and Samurai. The first and the second version look quite nice. And we can also see here how things like neon lights and dynamic energy boost the Cyberpunk styling. Now, let's check out the next prompt where we have a more grit and dark scene. So we have here an industrial vibe with a steam punk, green and dark theme featuring machinery in war hammer style will also supply an artist's name here. So all this together should create a very interesting result, but also have a wide aspect ratio. And also, we'll try to get that as a 3D render, which should add volume to the picture. Alright, let's Upscale some Versions. We can see in here a very strong influence of the post-apocalyptic node, green dark, and war hammer. Also the artist's style matches quite nicely all of the Versions. So this is a good example of blending styles that work quite nicely together. Let's take the next prompt, which is in a similar style, the dark and green. We have a dark, surreal world. We specified here that this is a dystopian Style. And we also have an artist that you can research a little bit, the styling of this artist. But what's really interesting here is that we don't really specify any exact location or any exact object. Most phrases just describe the mood overall. So this works quite nicely with artworks or more like Painting Styles. When you just want to let mid journey choose the scene that could give the viewer the feeling that we're describing here? Okay, Now let's check this a bit closer. If we'd look a bit longer on some specific, basically on any of the pictures. We could definitely feel the mood of all those key words that we've specified in the prompt. So you can use this technique of specifying emotions and feelings instead of the exact scene. When you're prompting for something that allows a very broad creativity. When it's a certain artist or styles that are surreal, a mystery, or Fantasy. In this case, we also specify that this is a dystopian Style. Let's also try out one thing. Let's try to prompt the same image. But in the niji engine. And let's take a closer look here. And same here, we're getting similar emotions from the description that we gave in our Prompt. I find this, those Versions in niji a bit more creative. This produces quite different aesthetics. If we compare to just the version 505.1, I mean, then let's go even further. Let's take the initial prompt. So we don't use niji this time. But what we do, because this is a very surreal Theme, Let's try to increase Chaos. Be on 50, let's put it to 60. And stylization give it 1,000. So this could produce really crazy results. Let's wait for the render. And indeed we're getting something interesting. In the fourth version. Maybe a little bit too much stylization going on here. So this is just another use case example of the Chaos and stylization Parameters. Always try experimenting with those when you're prompting for some kind of painting. Now, let's check out the last Prompt here. And we're having majestic dragons here as the main element. We have the mythological fantasy will mix medieval mystical creatures. And what's more important is we're going to mix two different artists. Now, let's see what. Now let's Upscale the new Versions. And what we've got here is very unique mix and a very rich scene. There's a lot of detailing going on here. There's a castle in the background, and of course the majestic Dragon. Let's check out this version. There is also a lot of detailing and I can see here the mix of our artists which produced the third variant of Painting Styles. Okay, now we're finished for this topic. So feel free to experiment and mix different topics, scenes and artists. And this way you can really explore and define your own aesthetics vibe 72. Pixel Art: This is a short introduction to pixel Art. And we can try a few working to Prompts that could generate this specific Style. Let's continue in our illustrations channel. And let's take our First Prompt where we have a flower shop. And we've also specified a Mood here, which is a rainy day. Now, let's trigger that. The main idea for Prompting behind pixel Art is to keep all the Prompts very short. Otherwise, it's hard to get that exact style that we're looking for because Midjourney just starts to get confused. Sometimes the niji version can help, but most of the time just a short prompt is enough. Okay, Now let's take a closer look. So we're getting the vibe of this Style. And we see the French flower shop, the rainy day in all Versions. Oh, actually, not the first one. But the second version is, I think, the closest match to the pixel Art work Style. And there's a big difference between the 8-bit, 16-bit and the 32-bit artworks. Midjourney understands this sometimes, but not always. So usually you would just have to prompt a bit more to choose not between the four variants, but for example, 16-32. Now, let's try the eighth bit and see if Midjourney understands that. So here we have a bit of a longer Prompt, eight bit pixel Art with a dragon character. But we emphasize this at the end again, that this is a retro pixel Art For Damon. Okay, now let's take a closer look. And definitely in two versions. The second and the fourth here. We're getting the perfect match to I requested style. The first and the third option. They are a bit irrelevant to the style, but those two are perfect. And we also get a dragon and an adventurer on all the options. Okay, Now let's try another technique. Here we have the sprite animation sheet, 32-bit pixel Art Fund and colorful. So let's try that. And if you noticed, we didn't specify the scene or the object. Nothing. Just the type of artwork that we wanted as basically a spreadsheet. Okay, let's look a bit closer. And once again, two versions, the third and the fourth version. Those are the perfect match to our Style. Okay, let's try the last prompt. And here we also have a 32-bit pixel Art. But we're specifying that this is a 2D isometric scene, which is a NEO Tokyo street and the rain with neon signs. A couple of key words that work very well with pixel Art. But the main difference is where Prompting this in the niji Style, but not the fifth version, the first. So the older version of niji, I found that the initial niji version works quite well with pixel Art, not the latest one. So let's see what this will render to us. And we can definitely see that the third and the fourth Versions are the closest match to the pixel Art because they are pixellated. As you can see, the 32-bit version of pixel Art has a higher fidelity or a higher detailing, but not as high as the first and the second example. We did get an isometric perspective in the first and second Versions. So once again, try versioning, creating more variance until you find the perfect match. Alright, so this was just a quick intro for how to prompt for pixel Art. Now, let's move on to the next video. 73. Cover Images: Book Cover, Movie Poster, and Thumbnailsr: In this section, we will cover the highly popular use of Midjourney and other Generative AIs to create cover Images for various purposes, such as book covers, Movie posters, and YouTube video thumbnails. Now, for this purpose, we could continue using illustrations or we can create a separate channel. For now, let's just continue and illustrations. And our first prompt is a YouTube thumbnail. And as a scene, we have a big airliner plane flying through the storm clouds. And we'll also add that this is a horror movie Mood cover image. And then we specify keywords that help to produce the Photorealistic and polished image. And for this particular scene, the HDR and things like crazy lightning and the red tracing can help to give a bit more drama to the scene. Now, let's take more of a closer look. What's I find the top two versions giving the exact Mood that we asked for. Also the airplane is quite accurate. The third option is also alright. But we definitely get this dramatic mood. And this could be a very perfect thumbnail for any video about some plane incident or investigation, or even for a movie trailer. Just don't forget at YouTube, thumbnails require a Landscape type of image. Now, let's take a look at another example. Here, we're using a Mulltiprompt and we have three prompts. So a close-up illustration. Then we have a scene of children playing during very extreme heat Conditions. And then we specify the style in our third part, which is a 2D vector cover image and tall Aspect ratio. Now let's check the result. And here I think the first version and the third version. What if feed quite well because they have plenty of space on the top four Book name, for example, or any other texts and Elements. Let's Upscale the third one. And what's interesting is that we've got almost everything that we've specified. So this is a close-up illustration. Not really close up, but we have the main object near the viewpoint. And because we wanted to include multiple children playing in an urban street, this worked quite well. And we've got the mood and also for the color, right, to give the feeling of extreme heat and in urbanized area. Alright, now let's move on to the next Prompt. Here. We're also prompting for a cover that could suit a book. It's a tall Aspect Ratio. And we have, as a scene, six-year-old blonde girl. We're specifying that we want a cheerleading and adventurous Atmosphere and also the fire watch Style. So we want to mix all this together while still having some adventures and a mystery vibe into this. Right? Let's take a closer look. And I think the first variant is quite spot on for everything we described. I can see the fire watch callers here. We can definitely see some adventurous and mystery Mood here. So this example could suit very well in a children's fantasy book as a cover. Now, let's continue to the next Prompt. Here we have a single Prompt with those plated with a double colon. And we will try to mix and give a very special look to the image by specifying things like isometric upwards viewpoint and things like futuristic and modern building magazine Cover Art. So with this tall Aspect Ratio, this also could suit not as a Book Cover, but as a magazine cover for commercial building, different projects Okay, let's take a closer look now. And as you see, this gave us very nice results. We could also increase the Chaos parameter to receive more different looks. But definitely, maybe not an upwards View Angles, but an isometric Style. And the coloring commercial building design included in the strong violet colors worked pretty well here. Alright, now let's check the next example. So as you can see, this is a very well-described to prompt. It's a single prompt. And we want to get the matrix type of vibe and then mix with cyber punk in a very photorealistic kind of scene. And here we also have a tall Aspect Ratio, which could suit a buck or a magazine or something similar. I'll write. Let's zoom in a little bit. And I think the first and the third version are nice. They're very close matched to the style we asked, especially the third of very Interior. It could perfectly suit a movie cover or a Blu-ray DVD cover, Et cetera. Though, I knew our prompting for whatever image that would suit a movie, then it's always a good idea to use the Movie still for that. Then it will produce the right kind of effect. Now let's try another example. So here we also have a movie cover or Movie still with extra terrestrials evacuating the population of Earth. I don't think that this will be the exact scene, but this should produce the right thigh. And we also specify that to View from the cosmos with the Photorealistic keywords. And then with a Mulltiprompt, we actually remove any possibility for text or words to appear. Well at least will lower the chances with a minus one. Wait. Let's also add, let's try this time wide aspect ratio. And let's also try to increase the Chaos a little bit. And now let's see what this can produce. All right, let's check it out. And we can see here some apocalyptic Mood with breathtaking views. But we don't see any View from the cosmos or any extra terrestrials. Now, let's adjust our prompt a little bit. And let's also change the aspect ratio. So this is the alternative version. So we removed some extra keywords here, like photography and photoshoot. And we also changed the wording a little bit. We added the UFO ships because that's what we wanted to see more of. And we've also removed the extra Chaos level. Now let's take a more closer look. And we can now definitely see some planetary level disaster. I mean, we could also see this in here of asteroids are planets colliding. But here we also see the extra terrestrials and ships. So try to play around with Wording. Because even if you changes like we did here, gave us quite a different result. Now, let's try our last prompt. It's quite simple. And here we're trying to get a vector Style illustration. We indicate also that there's slack and Behance. Those keywords should help to get a nice quality illustration. And the scene is just a tech startup. And what the Mulltiprompt will specify that we want it in a gradient style. Alright, let's take a closer look. And we've definitely managed to hit the styling. Also with the second and the third version here, we can see the office Setup and a bit of a startup Mood here. So this was just a short introduction of how you can use mid journey to generate different types of cover Images. Starting from children's books and ending up with dramatic movie covers. So that's it for this video, and I will see you in the next one. 74. Famous Painting Styles with Blending: Here we will explore a few famous Painting Styles and attempt to blend well-known styles from artists of the past decade with futuristic or surreal scenes. Now let's go to discord. And for this, Let's go to the styles can. And let's try our First Prompt. So here we have a painting by Leonardo Da Vinci. But we specify a techno Renaissance with a futuristic cityscape and flying cars. And if you are the keywords that are not usually associated with the paintings of Leonardo Da Vinci. So let's see what this can bring us. Okay, let's take a closer look. Now we see that there's a blend between the Leonardo Da Vinci style and the futuristic cityscape is producing a quite unusual result. And I would say very unique. This way, you can even create your own personal style. So always try to mix different themes and styles. Or even use key words that indicate that you want a painting by two different artists. Now, let's take the next Prompt. Here. We have Vincent van Gogh with a post-impressionism Style. And the main scene is dreamy, Starry Night. Let's actually change the aspect ratio to a wide. And this should produce interesting results. And we'll also try to increase the difference with the extra Chaos value just to get a bit more diverse results. Alright? And we can definitely see some interesting Options here. Even with the horse-like creature in disguise, some Dragon heat error, and extremely unusual starry skies. So that's quite a nice example here. Now let's take the next prompt. This is another famous painter who worked mostly in expressionism. But here we have urban expressionism and we have quite a surreal scene here. And basically we're trying to get a more urban or modern city life. We also increase the Stylize parameter and the Chaos parameter here. So let's try that. As we can see, we've got quite interesting results here. We can definitely see the influence of this painter. And the scenes are the City of life, which is quite an unusual but nice visual. But let's try a little bit more of an alternative prompt and replace the stylization and Chaos parameter with niji v5. And now let's see how niji can actually paint this. Because the niji model is quite suitable for any type of painting. Not always. Of course, it produces a close match to the style, but it really depends on the rod and the style that you described. Sometimes it works, sometimes it doesn't. All right? And we've definitely got quite interesting results here. This option is very nice and surreal. But I also see less influence on the actual style of the author. And all the options have a little difference in Painting Styles. Especially the third one. It doesn't really fit to our Prompt, but it's always worth trying. Okay, now let's check out the next one. So here we'll also try to use niji and we have a surreal Painting in Baroque style by this author. Then we basically Describe the same Okay, Now let's take a closer look. And we can definitely see that the Seed worked quite well. And because this author, Johannes Vermeer has famous paintings in quite a realistic style. We can definitely see his influence here. But only the fourth option is a real close match to the actual style. But we don't see too much surrealism, maybe just a little bit here. But the third version is quite far from the styling that we prompted for. Okay, for now, let's move on and try our next prompt. So here's quite an interesting prompt, which is a minimalistic fusion painting by Pablo Picasso and Cubism Style with geometric shapes and etcetera. And what's interesting here is that will actually lower the stylization of it. Now, let's see how this affects the image and if it's actually suitable for this type of Painting and Style. All right. And all four options are very close match. But I think that the third and the fourth option here are very close match. And they also give some sense of machinery or robotic Figures. Overall, all four options are quite nice and interesting. Okay, let's go try the next prompt, which is a quiet, surreal scene. And we've also chosen a right author for this. And we'll try to increase the Chaos. This usually works quite nice when you have some things surreal or when you have some kind of fusion and you fuse multiple styles and artists. Because this can allow mid journey to express itself a little bit better. And we can see that we have very strong surrealism in all Options. Some elements like clocks and the styling overall is related to Salvador Dali. You can also try this in Nietzsche. I think this can also produce a very interesting result. Now, feel free to mix some stylings of those images. Try different engines. You can even tried mid journey version four because it has its own aesthetics to this. And now you know enough styles to get yourself going. Always experiment, mix, and try to polish your own aesthetics using some persistent keywords. That's it for this video, and I'll see you in the next one. 75. Upscaling: In this chapter, we will cover the basics of finalizing our images using third-party tools. We will install an open-source Upscale air and use Photoshop to edit the Upscale Images. Let's first take a look at how to properly save your images. And then we'll continue to the up scalars. So one of the ways is saving it directly from your Discord channel. Let's find some Upscale diversion. For example, this all you have to do is just right-click and choose to save the image. You can also do it here. And then you'll be prompted with the location. Choose the one you need, and then just save it. Now let's see another way for downloading your own images. So by going to the mid journey website to your page. In here, you can check the up scales or grids. Grids are the initial images to where you have four variants. And then here in Options, you can also save it with this button here. Save Image, and then it will just download. Now, let's find it. So this is our image. Let's check the details. And we can see that this is a PNG file. So that's quite important to know and understand that Midjourney produces all of the images in EMG. So it's important to know that all the files that Midjourney generates are all in PNG format, which is good because usually this is a lossless type of format. It means that it's not compressed and has the maximum quality. Now, the important part here are the dimensions of the images. And this is exactly the reason why it's recommended to Upscale the image further using an AI Upscale where because if you would want to print the image on Canvas, for example, or you're monetizing on mid journey and selling the artworks than this resolution is too small. Future Versions of Midjourney should produce a higher resolution. But even then, for any physical printing, you would still need to upscale the image. Now let's check the first step, scalar. This is a paid variant, but it's a cloud hosted Upscale. Or you can, You can see here the result. What it can do. You can try it right here from the front page, but basically it's quite straightforward. Let's check the pricing. So you can get free credits per month without paying with a free account. But if you want to Upscale more images than you need to upgrade. But I find this pricing quite reasonable. Now, let's check the second version. So this is another Ai up scalar, which is Cloud hosted and one that I can recommend. It has similar pricing, but ten credits instead of the eight. But Baidu say that they watermark the images created using the free plan. However, the quality of generated images is quite high. And you can also often not for a monthly subscription, but just for a pay-as-you-go purchase, a certain amount of credits. But of course you can try this before you purchase and see which one up scales better for your specific image Style. Now, let's check the next one. So this is a free online service, at least for now. It's been here for awhile. But it doesn't provide a very good quality on Upscale images. But at least that's a good place to start. If you can't use an installed version or the software that we're going to Overview next. Because installed up scalars, they do have quite high requirements for a computer. And usually they need quite powerful GPU, which is a graphical processor. So if for some reason you have, You can't Upscale locally on your computer You can use one of those services and this one should be free. Now, let's try to Upscale are image. This is the most important part. You always need to choose the right model. Up scalars tend to have 34 or even more different models that suit different types of artwork, paintings, graphics, or photos. In this case, we can either choose an artwork or photo. So of course we choose Artwork. If you want higher resolution, then those would need to be upgraded. And never choose noise reduction for Midjourney because Midjourney never has any noise. Now let's try it. Okay, it's finished. Let's download and then we can compare. So here we have, on the right, we have the original from mid journey. And on the left here, this is the Upscale diversion. Now let's zoom in the original. And you can see that the face is getting really pixelated. It's losing the quality. And this is how the Upscale diversion looked. All of the particles in the background also gotten nicely up-scaled. This free up scaler works quite good with illustration Types, but not so good with a photorealistic images or photos. And it's also limited in the size of the image that it can generate. Now, let's check the dimension. And we can see that it was increased four times or four racks from the original, which is a good start. But we can do much better. And for this, we need to install and Upscale are locally. Now, let's go to this link and download this app. So you have to scroll. It's good if you read the description and all the notes that it has. This is the important technical note that you need to have a compatible GPU to Upscale images. But as it says, you can try that and see if it works. In the worst-case, it'll just not work. You can find documentation here in the wiki. There is also some troubleshooting and a big compatibility list. What's important is that you can install this on a Mac, Windows or Linux, because we have windows here. Let's continue and install the Windows version. For this, you will need this file. Now, let's quickly go through the installation process. It's super straightforward. When you're done, just launch the application. And let's now check the settings. So if you prefer the white Theme, you can change this here. Now what's important is I would recommend the images be saved as a JPEG. They will still have the high-quality, they will just have a smaller size. Everything else leave as it is. And the actual upscaling process is very simple. So the first step is choosing the image that we're choosing the model, and then we're choosing the output folder. Let's go to downloads, and let's choose the not Upscale the version. Then here, you would need to try what works best for your kind of artwork. Because this is an illustration and a digital Art Style. Let's choose this option. But we can also try to double Upscale that, which means that it would be up-scaled four times, and then that result will be Upscale to four times again. But this indeed has a note at this may not work properly with all images. It all depends on the initial size. And if we would double Upscale that, this is the dimension which is very big that we would get. So let's just take that for now and let's try it out. Alright, it's finished. We can compare the results here. So this is the second Upscale that we did with our software. We see that the type of file is JPEG and that it has a smaller size than the one we use Style mine. Let's check the result now. So it also did a great job. And because this is an open source software, you can run it for free at your own computer. All you need to have is just a compatible GPU. Now, that's it for this video. Now let's learn how to finalize our images 76. Image Editing Overview and Introduction: Adobe Photoshop: Now, let's walk through the basic image editing process to finalize the image. Will use Photoshop in this example. But you can use any other image editing tool of your choice. As most tools follow the same principles and provide similar functionalities. Before we start with Photoshop, Let's just quickly go to the Adobe website and let's check out the current offers they have. Let's see, creativity and design. And let's compare the two plans. And this would be my recommendation of the subscription plan. If you don't own Photoshop yet, which is around ten pounds a month. And you're also getting light room, which is one of the best tools for photo editing. And it could also be handy to polish your photorealistic Content. Now, if you're a student, you can also opt him for the entire Creative Cloud apps at a discounted rate, But please check the conditions first. And I want to show just one alternative to Photoshop, which is called gym. It's an open-source and free software. It has very base out very similar functionalities if we compare it to Photoshop. Also quite powerful image editing to. And it's available for Windows, Mac, and Linux. So if you don't have Photoshop yet, feel free to check this out. Now, let's briefly go through the user-interface and the core functions of Photoshop. So on the top here you have the menu bar with all the important functions. Let's start with opening and Image pictures. Let's choose some of the images that we generated with Midjourney. Let's take this example. So this is your main working area. On this tab bar, you will be able to switch between open documents. And when you're done with your edits, you can just save as. And it's very important what kind of file format you select. If you want to do an export in regular image format, then your best choices are JPEG and PNG. Png would give them more lossless type of format, meaning it won't have any compression and it will not affect the image quality. And JPEG can save you space and produce a more smaller file size. The image might lose some of its quality. I recommend leaving nine or ten for optimal balance between the quality and size. The next important menu bar is image. Here you can adjust the brightness, contrast, colors, and a lot more. We will cover this in the next video. And you can also manipulate the size. It's best to work in pixels. If you want to resize the image, then I would only recommend this if you want to downsize or reduce the size of the image. If you want to increase the size, then please use Upscale are first before manipulating the image size here. You can also crop the image by using the Canvas size. Let's say the height is too much and we want to cut the top and the bottom to reduce it. Just proceed. And now you can see the difference after we've cut the image. Now, it's quite handy to have a history tab in front of you, somewhere on this tab group. To enable that, Let's check another menu which is called window. And here, just enable all the Aldi important tabs that you want to have quick access to. So if you see, I have history enabled here. Let's try that again. And every time you enable a new window, you can always reposition that. Just drag it and put it in the location that you want. You can also close the group with this option. The Layer menu. Here you can create new layers or adjustment layers. So Photoshop and many other image Editing tools use a layer principle. So it looks like a pie where you'll have one layer on top of the other. Let's, for example, draw the image, the square. So we've drawn that on a new layer and we can enable or disable And manipulate each layer individually. Also right-click on the layer, check the Blending options where you can adjust the shadows and other effects for the layer or delete the layer altogether. Now, if you're planning on having a complex editing on your work, than I suggest to create a layer from background, from the initial image. Just click Okay, and then duplicate that layer here with this option. I'm making a copy just as a backup. You can also hide it and then work on the visible layer. And another important menu tab that I want to show is called filters. You can go to the Filter Gallery. Again, an idea of each filter and what it does to the image. Okay. Now on the left-hand side here, you have a tab with tools starting from the selection or move tool. For example, if you want to move an object or an image. Vector can also hover your mouse and it will show you a handy helped tip of what this tool does. If you long click and hold thing, you can choose an alternative version of that tool. So in this introduction, I'm not going to go through all the tools we have here. Just always remember to check on your left hand side what kind of tool is selected. Now we're coming to this location here. And this tab basically shows the options for the current selected tool. So Let's say if we select the text editing tool where we can add text to the image. Here we will have the options for this text tool where, for example, we can choose the size of a text, where we can choose the color of the text, the font, and some tools, they actually create a new layer. So if you notice here, when I created the text, we actually got a new layer so we can hide it and we can also move it with this tool. Now to zoom in or zoom out your image. You can either use this location here or you can control click Control Plus or Control minus on Windows to zoom in and zoom out the image. And all of this on the right are called panels or windows. So that was the menu option where you enable or disable those windows. The one that's quite handy that I would recommend opening as soon as you launch Photoshop for the first time is called history, where you can go back and forth in time, reverse some of your actions. Now before we move on to the actual editing of the images, Let's go to the edit panel and let's open Settings. Now I'm going to show a couple of recommended settings on new installations. So if you have the latest Photoshop, one thing I suggest doing is enabling the Content Aware of Tracing Tool. So as soon as you enable it, you will be able to create quite easily paths or to trace the paths and vector images. Let's take this just as an example. Okay, one moment. So this is the vector Style image that we generated before. And in case you want to convert this into a real vector. Of course, you can use Image path tracing tools in Adobe Illustrator or similar. But you can also try the new Photoshop future. So if you select this, Let's create a layer out of this. Now let's go to Settings. In the Edit menu. You go to Preferences and then general. And I will show you one important setting. So if we go to Performance and if you have more than 8 gb of RAM available, I strongly suggest to increase the history stakes to 300 or more. Otherwise, because the default setting is 50. And that's done enough when you're performing quite a lot of changes and edits. And in case you want to revert to a much earlier state, the default setting will not allow you to go beyond the 50 stages back or 50 actions back. Alright, so now that we understand the basics of user interface, we're ready to proceed with image editing. I'll see you in the next video. 77. Image Finalization with Adobe Photoshop: In this video, we will learn how to remove unnecessary objects from the scene. Because currently Generative AIs may not always produce a one-hundred percent accurate result. Additionally, will explore a few useful functions. Interests in brightness, saturation, and contrast. Now, let's open an image. You can practice on any image that you've saved from your Midjourney Gallery or from Discord. Just make sure to Upscale at first. So here I have prepared and up-scaled image. Now, before we begin, let's just create a backup. So we duplicate this layer, remove or hide it. And now we're going to work on a separate copy of this image. Now, let's zoom in a little bit. And we can see that this image has little bright dots all around the image. And here you see in this scene, they don't make too much sense. So let's see how we can remove those. It's always a good idea to zoom in almost to the maximum because it's going to be easier to work. And now let's check this tool here. And let's choose the Spot Healing Brush on top here and Options. Let's reduce the hardness to 50 and the size just enough to cover a little bit like this, to cover the object that we want to renew. And then just click on the object like this. And you can see that they are being filled with a similar background that is around this object may not always produce the best result. So then you would have to make a step back in history and retry it again, try to resize the brush. Or we could also try a little bit different tools, and I will show you those in a second. Okay, let's remove this dot also. Now, let's say that we want to remove a bigger object like this. Then you can use another tool. Select here, it's called an ISO. Now, press the left button on the mouse and just try to select it like this. Then right-click this area. Select Fill, and then Content Aware Fill. Click. Okay. And as you can see, a quite large area was basically just generated and removed. Now, I want to show another way of manipulating an image. It's a bit more advanced. But it can be really helpful when you want to not just move an object, but maybe changed the shape of it. For example, make the fingers shorter or longer, or manipulate some other object. For this, you will need a clone tool. Let's also, let's put the hardness a little bit more. And hardness basically means if you can see here, the brush. So when the hardness is minimum, the edges are very soft. So whatever you paint is going to be blended in. But this doesn't always work for every occasion. And when you have a hardness of 100% than the edges of this brush will be quite sharp. Let me show you what I mean. So let's say we want to remove this object here. So with the clone tool, you would need to hold and halt the area where you want to clone from. So let's say we want to take this rock here and we want to paint it over. Actually make, make sense to try and match the texture So for example, we see this black line going at an angle here. Let's say that we want to extend that. So let's try to click here. So you will need to hold out and then left-click the mouse. Let's zoom in and then just start drawing. And as you can see, the edges are quite sharp because I've chosen 100% of 100% of the hardness. Let's lower it. And then let's try again. So as you see, we've extended the line here. Now let's choose a new area. Then, for example here, let's choose it again. In this way, you can basically remove or add objects to your scene by cloning some other part. So this is an extremely helpful tool. You'll finalize the generated images. Now before we move on, let's briefly check how you can manipulate brightness and contrast. So this is the very basic brightness and contrast adjustment. So if you would want to make the picture more bite, you could increase the contrast and decrease the brightness. For example. Let's decrease that. You can uncheck this box here to compare. And now let's say we want to decrease the saturation. But instead of using the classic saturation, I advise to try vibrance first because it's not so aggressive and it produces a more natural look when you enhance or even lower the vibrance. We can compare. Okay? So now we have a bit more natural and moody look because mid journey tends to produce quite bright images by default, I'd always advise to check how the image would look with a bit more lowered vibrance. Now let's check a couple of other handy functions. One of my favorites is the photo filter, which allows you to make the image warmer or cooler. So here you increase the intensity. As you see, we have the warm filter by default with quite low intensity. But because we have cooler image, that warm filter makes sense here. So let's try that. Maybe not so warm, like this. And let's compare. So this is the way to adjust the so-called white balance. Now, let's try the same with a cool filter. That's what we need to do. This actually looks quite nice for you, increase the blue. But of course, I would advise to do that before we adjust the vibrance. Now let's go to History and compare with the initial state of this picture and the current state after our small edits. We could also just hide and renew this edited layer. So as you can see, even the small edits like this, they can help you finalize and extract the maximum out of the artwork. After you're done, don't forget to save your work. It's a good practice to have a backup in a PSD file, the Photoshop file. Now let's export it as an image. Here. Export As. Now, I choose the quality, but I don't suggest to have the maximum quality. It's better to leave it somewhere around in this area. If you want a lossless format, then the PNG would be best. And then just click Export Okay, Now let's check how you can cut away objects from an image. For example, a Logo, to make the background transparent. Or when you have a whole collection of objects, for example, UI elements or objects for a game? Well first, we always start with duplicating the layer will hide the background. Now, this is quite a basic example. And what we can try to do here, if we just want to cut the image here, is trying to cut the background. Now, in this area, in this tool, select the Magic Eraser tool, and then click on the area you want to remove. All right, let's zoom in. And we can see that this worked quite well. Now let's check another way. So this is another very handy tool. It's called the quick select tool. Let's increase the size a little bit. And then you just paint over the object you want to select. Let's say we'd want the animal here. And then if we want to expand our selection, we'll go further like this. Don't forget to select everything inside. If there was something. If something wasn't selected. Now you see that we've chosen just this orange background with our animal. And let's continue expanding our selection. Okay? And basically now you can just copy paste. You see when I pasted a new layer was created. So if I remove this, we can see that that's selected object was cut nicely away from the background. Now, what we can do, we can hold Control and then left-click the layer here. This will select all the visible pixels in this layer. And now let's say that we want to remove this unnecessary space around. Now we go to edit and we select, oh sorry, not edit the image. We go to Image and we select Crop. Now every, now the image will be cropped to only have that selected area. Now, let's save our work. Now let's export this with transparency. So for this, we need to select PNG and keep this tick box enabled. Same transparency. And then just click Export. Now, if we open it up, we can see that there is no background here and we'll only have our object. So this was a very brief introduction to image editing. So if you're very new to this and want to learn more than it makes sense to find and study course about image editing. And I would suggest sticking with Photoshop because of its popularity. And there are a lot of guides and tutorials of how to do certain things. So this chapter is now complete. And now let's move on to the next chapter. 78. The Terms and Conditions, and Community Guidelines of Midjourney: Welcome to the last chapter of this course. Here we'll go through Midjourney, Copyrights, Terms and Conditions, technical limitations, and Ethical Considerations. We will also then check out a few prompt banks, latest trends and Ai, Midjourney's roadmap. And then we'll summarize what we've learned in this course and then go through the Recommendations and possible next steps. Now, let's check on a few important takeaways from Midjourney's Terms and Conditions. Because people tend to skip the boring reading of Terms and Conditions when they agreed to those while creating an account, for example. So I would like to highlight a few important things, but I do encourage to spend ten to 15 min and go and read through the full list of the Terms and services which can be found on the Documentation page under policies Terms of Service. Now, let's begin with the summary for this topic. Service availability and quality. Server services that are provided by Midjourney may be modified or changed without guarantees of quality, stability, uptime, or reliability. Users should not create dependencies on those services are assets and the company, meaning mid journey. Here's how liable for any harm caused by such dependencies. So this is quite an important point if you're planning to utilize Midjourney for commercial purposes. Because this project is still officially in the beta phase, you should always have one or even to backup solutions. So the next one is age requirements. So I'll users accessing the services basically confirmed that they are at least 13 years old and meet the minimum age of digital consent in their country of residence. Now, the next one is one of the most important parts, copyright and trademark. So all users using Midjourney grant the company a worldwide non-exclusive license to reproduce, display, perform, sublicense, and distribute the text, meaning the prompt and the images that were produced using this service. So these are the rights that we're giving to Midjourney. And now these are the rights that Midjourney kids to us. The most important part here is that user's own the assets they create on this platform. This literally means that everything you have prompted for using your own Discord account belongs only to you and no one else. However, for non-paid members or members on trial, they are granted a license under the Creative Commons noncommercial version 4.0, attribution international license, or all the assets they've created during the trial or free period. One more note on the content. You can always completely remove everything that you have ever created. Currently, this could be done by either marking the image you want to delete with the Red Cross or red X emoji in Discord. Or for example, in some cases where you've lost access to your Discord account or something, then you can also request the deletion the email. And you would have to right to privacy at midjourney.com. And the last thing about the rights is that if you own or if you work for a large company or enterprise, which has more than $1 million gross revenue per year, then all of the employees would need to have a Pro membership as a minimum. So in this case, just check all of the Terms and Conditions throughout. I'm not going to go into details on this part. However, it's just important to remember that any content that you have linked in your prompt, for example, Image References. And if that content was used without the owner's consent, it could potentially lead to the removal of that content or the result could potentially be deleted if it was prompted with a link to current image without the needed permissions So here I also wanted to show that the governing laws for Midjourney is California, USA. For payment and billing. Basically, all users are free to cancel their plan at any time and may even request a refund. If they didn't use the service to do that, they would need to send an email to billing at mid journey.com. I mean, for the refunds. But to cancel your subscription, you can do that directly from your account at the billing page. Now, the last part, the Community Guidelines. So all users are expected to be respectful and to avoid creating or using abusive, pornographic, or violent content. Please always check the current banned topics in the mid journey Terms and Conditions and the community guideline page, we will check that page shortly. And repeated violations could potentially lead to a permanent account ban. So always take extra care when Prompting. Because sometimes the mid journey bot can identify even harmless prompts as violence or find something else that is forbidden in them. And if you keep trying to generate an image with something forbidden in them, it could have diverse consequences. So if you do see some issues like that, then it's worth checking in the support chat, in the members support chat here. If the topic that you're trying to prompt is banned or to find some other health in general. So through this link at their Creative Commons, you can read more about what kind of Copyrights you're getting when you're prompting using a free account or when you're account is on trial. Now, let's go through Community Guidelines. So this is the important part. All of the content that you generate must be PG13 rated. You can research a bit more what that names. But in two words, it just means that any contents that is not suitable for children under the age of 13 should not be generated. So that's another important one. So you shouldn't repost creations of others without their permission. You should be careful when sharing your content or links to your content outside of the mid journey Community. Because not all communities are welcoming of Ai Art. I will not go into more details on this page, but please go through the full list of banned topics and Image Types. And there are also a few notes about that, about the Public Figures and how not to prompt them. But what's important is that if at any point you see that the image that was created from your Prompt can potentially violate the guidelines or Terms and Conditions. You can remove them easily using this emoji. And if you see some images in the public Gallery or on the public Discord channel, you can report those using functions displayed here. So that's it for this video. In the next video, we'll go through Midjourney Limitations, Ethical Considerations, and we'll also go through pros and cons regarding regenerating images of AI in general 79. Midjourney's Limitations, Ethical Considerations and Copyrights: Now let's continue with Limitations, Ethical Considerations, and pros and cons for using AI to generate Art and images. So during this course, we have explored and learned how to use Midjourney and implement our creations. So now it's the right time to talk about its impact on our lives and on the lives of others. There are important ethical considerations to keep in mind. So let's have a look. Let's start with pros. Obviously in general, there are positive aspects that drive us towards progress. Air generators have made it possible for anyone to create stunning artwork or virtual photography, regardless of their artistic skills. With AIR generators, all you need to do is enter a Prompt and the AI engine will create an image based on it. The clear and more detailed your prompt is, the more interesting and precise the results are. This means that anyone can produce artworks or images to be used in commercial projects. From 3D character concept designs to surrealistic paintings and artworks. And beautiful photorealistic seems. The possibilities are literally endless. Another significant advantage is cost effectiveness. There are many AIR generators that are free to use. Mid journey is a paid service, but it's one of the most, if not the most cost-effective, considering the fact that it basically allows unlimited, high-quality generations on relaxed mode. And you no longer have to worry about hiring traditional, expensive Art or design studios. Instead, you can easily learn to create your own designs without breaking the bank or spending weeks on designs. Lastly, AI technology can classically revive well-known Artistic Styles. Mid journey has learned from famous artworks and Painting Styles and can create new Art pieces inspired by renowned artists. And it's really remarkable how those AI engines in Midjourney pay homage to the masters of the past and bring their styles back to life. But there are many more positive psychological impacts than just those pragmatic ones, such as positive effects on your mood and psychological state overall, when you are capable of creating with such an ease, a piece of artwork that can unpack your emotions and aesthetics. It can also be beneficial when creating photos of your dream locations or some past experiences. As David mentioned on the office hours conference call, this can have dramatic and positive effect on your well-being and overall condition and can even help to heal some of the psychological issues. Now, let's go to the cons. So while the AIR Generation offers exciting advantages, we must also consider the ethical downsides and also some of the copyright concerns. At the moment, due to lack of regulations. Copyright of Ai Art is somewhere in a gray zone where every country and individual interprets existing laws different way. Air generators usually learn from existing images, videos, and texts supplied by different third parties and open sources, and also by scraping data from the Internet, sometimes without having valid permission from the original artist. This could potentially lead to copyright. A fence where artists find their artwork styles replicated without their consent or proper credit. Also, I just want to show one case here. So this case started last year in September In 2022, when this comic book writer used Midjourney to create illustrations for her BAC. And when the artist applied for the copyright protection at the United States copyright office, the office had declined it because all of the images were created solely using Midjourney at the time. This, of course, doesn't mean that there was something illegal. But to apply Copyrights for your images that you've created using mid journey. Currently there are no exact regulations on that topic. So every country and every Copyright Office treats those cases differently. So in general, currently there are no specific generated Ai copyright laws existing, making all this copywriting quite complex. And lawmakers around the world are still grappling with how to protect artists rights and determining the boundaries of a AI-generated Content. However, the same as real people. Nobody can forbid you to get inspired by certain artists and to create unique artworks, but in an existing and well-known Style. Now, another concern is that increasing use of the AI generated images instead of commissioning human artists can have a significant impact on traditional artists revenue with easily accessible AIR generators, potential customers may opt more for a AI-generated artworks instead of supporting the traditional artists. This further intensifies competition within the Art industry. And this is already sparked some widespread protests in some Art communities on mine. Certain platforms have even opted to ban AIR it altogether. Therefore, before uploading your artworks to any third party online Gallery, please ensure that it is an AIR to friendly community to avoid any disputes or tension within that community. Now, I just want to show you a very simple example here. This is just one example, but Art station is one of the largest design an online Art Galleries. And you can see here with those images, traditional artists are protesting because many people using mid journey and other Ai generators, our uploading their artworks in different online galleries. However, it is worth noting that despite online protests, many well-known and large communities have opted to embrace the new technology rather than protesting against it. In fact, some of the largest online galleries have even introduced their own integrated AIR generators directly on their platforms, which was perceived very well by the artists. Now let's see another example. So here we have the same inquiry. We're searching for Ai Art. And we definitely don't see any protests here. And we see quite a lot of different Art. And I can tell that most of those, if not all, were generated using AI. So those were just two simple examples. And you can see that the ethics of using AI and Art are complicated. For now, it's really hard to say whether it's right or wrong. Each person has to decide for themselves. But what You have to know is that many well-known and recognized artists are already on board with the a AI-generated tools to help unleash their creativity even further. But on the other hand, some artists found themselves being abused. As AI gets better and better. There'll be new laws soon about how AI Art can be used. Those laws might regulate how AI creates Art. Or for example, it might stop it from copying other artists. This is just a prediction, but it's worth checking those regulations egg your country of residence. So is there anything that Midjourney does to solve those issues right now? Well, firstly, they give users a comprehensive guidance on fair use principles, but also encourage them to ask for permissions from that corporate owner if they're unsure. Midjourney is also committed to promoting creativity and Ethical Content Creation. They also make sure that they're prompts and images don't actually infringe on other artists rights. Setting a good example for the whole industry. Now, it's worth noting that mid journey currently has some limitations. We're now at the subscription plans page. And as you already know, the relaxed GPU time in eliminate about is only available starting from a standard plan. There are also fast GPU times that are limited to those amount of hours depending on your subscription. The stealth mode, which can be activated only on the Pro Plan. And it allows you to hide all your jobs from the public Midjourney Gallery. Otherwise, all of the images that you generate are publicly visible, including the Prompts. There's also a limit to the amount of concurrent jobs that you can run and jobs that can go into the queue. Apart from that, Midjourney is currently limited to operating exclusively through the Discord platform. Another note is that mid journey may not be able to generate an accurate representation of a specific person or a place depicted from an image Reference, for example, unless it pertains to a widely recognized public figure or a well-known location, but even then, it might not produce a precise image. Also, the Midjourney B2 might refuse the generation of certain scenes even if the prompts do not specifically violate the Terms and Conditions or the Guidelines. Just a simple example is a cocktail Bloody Mary, which I was not able to reproduce due to the word blood in it. And lastly, mid journey might have limited knowledge or understanding of highly specific subject or Painting Styles. This could also be where service interruptions may occur during the beta phase of Midjourney, which is the current state. Finally, there might be some uncertainty or ambiguity surrounding the regulations related to content ownership on mid journey or other Ai Generative NGOs. So try to follow the latest developments in this industry and it's worth monitoring on any upcoming regulations. In Ai news in general 80. Tips & Tricks, Recommendations and Course Summary: Now let's move on to the tips and tricks, other resources, the future of AI and the conclusion. So to get some inspiration and to learn from other Ai artists, it's worth checking and monitoring the prompt banks, where of course the main bank is Midjourney itself, the Gallery. You can use search here to browse for the prompts on the topic of your interests and see how others do it and try to learn from it. Of course, you'd have to take a closer look and check to see if the results were generated from a professional Prompt. Or it's just that Midjourney infused a very basic or small Prompt with its Creative randomness. Another resource that's worth checking is called Prompt hero. And to gain full access to the platform, you would need a register, a free account here. Now let's just briefly go through the Midjourney Gallery here. And in similar fashion, you can sort here the artworks to show the newest, first or the highest rated. You can choose a version here. And of course, EUR. You are able to see the prompts on all images here. Just also always check the version that was used to generate that. Another resource is called prob tante. As far as I know, it doesn't have the mid journey prompts here. You would also need to register to gain the full access. But in similar fashion. You can get inspiration and find the Prompts here. Even though the original prompt was used with some different AI tool. Another tip would be to use some text generators like Chad GBT for example. You would also need to register for free account. Currently because this tool is still in beta, it's pretty Use and you can use chat GBT basically as your assistant and ask it to generate some scene or idea that you can then use to generate a Midjourney to create an image. Let's try this simple text prompt. So we're asking chat GBT to describe a surreal photo with white roses blooming on the rooftop of the New York State, sorry, The New York Empire State Building using only two sentences because otherwise it would generate a long description. And then we'll also ask to only include a few important keywords that would describe the shooting angle, Lighting, background, scene, Mood, isolating aesthetics. We'll also ask chat GBT to be creative. That's just a very basic and simple example, and this is what it gave us. Now, let's go to our Discord and let's try it. I would of course, if we want a photo, I would of course include a photo of a photo with. Let's also include a few extra keywords to describe the style of the photo and also the aspect ratio. But we keep the scene that chat GPT generated unchanged. Abc, that it works quite nice. And we have a very high-quality, but a bit of a surreal scene in our photo. So feel free to play around with this tool. It can be used for many other topics, but it's out of our scope. So try it out. There's also plenty of extensions for Chrome that could allow you to interact with this tool and ask for a mid journey Prompt directly Now let's briefly go through what to expect from MIT journey and AI in the future. So at the time of the recording of this video, mid journey Versions six is expected to be released quite soon within the next one or two months, which would be July or August 2023. And with Versions seven to follow later this year, those Versions will introduce new features and confirmed upgrades include fire Image, higher image resolution, potentially increasing the resolution to 2048 by 2048 pixels from the current 1024. Improved texts understanding. So future Midjourney Versions will have a much better comprehension of descriptions, making it easier to generate scenes based on the written input. Users can utilize a more conversational language without confusing the Ai. So those are the confirmed updates that are coming. Now let's see some circulating rumors about the future Versions of Midjourney. Here, just a few of them. Most likely, there'll be much more customization options. Users may gain greater control over Variation in images created by future mid journey Versions, allowing for specific changes according to their preferences. Another one is 3D bottles. Future Midjourney Versions might enable the creation of lifelike, pretty objects and scenes. Adding two new dimensions of new artistic creativity. Then the enhanced hand drawing images. Those could be produced with much later Versions of Midjourney, which will increase realism and detailing, potentially becoming indistinguishable from naturally and manually drawing illustrations. And another one is video creation. Unlike previous versions, future Midjourney Versions may allow users to generate dynamic and life-like videos, opening up new possibilities for storytelling. Some anticipated futures and upgrades were mentioned directly by David, the CEO of Midjourney, during office hour conferences. So it will be interesting to see what new Midjourney Versions bring to the table in the future. Looking beyond Midjourney, Ai has the potential to revolutionize various industries and aspects of society in the near future. Here is just a brief overview of the possibilities and potential impact that Ai may bring it in the upcoming ten to 20 years in the related niche of Design and Art. So in graphic design, AI could start autonomously generating visually stunning creations, Logos, and marketing materials, working as a creative partner alongside the designers. If Fashion Design Ai might accurately predict Fashion trends, leading to a more personalized and sustainable fashion experience with custom clothing designs and Interior Design, AI powered virtual assistance could analyze architectural plans, understand user preferences, and offer innovative design solutions, providing an immersive experience and informed decision-making and Art. I could collaborate with artists, emulating styles and creating novel Art Forms, resulting in breath-taking an intricate installations and in 3D modelling, Ai could streamline the 3D modeling process, enabling designers, engineers and architects to create intricate models with advanced features. In real time adaptions. The future holds immense potential for Ai transformative capabilities across multiple domains and not just Design. Most probably in the near future. It will affect things like politics, economy, medicine, and science. So we're living in very exciting times and it makes sense to follow the Ai news and any trustworthy resources. So we've reached the conclusion page. Let's recap the course and check key takeaways. So through this course, we embarked on an exciting journey through the world of artificial intelligence. And it's remarkable ability to create stunning images. Now, let's summarize the learning path of this course and recap our journey. So we began with an introduction to AI and mid journey Exploring the concept of a AI-generated Content and the various applications it offers. We delved into Midjourney's features, navigated through Discord, and learned how to stop and utilize Midjourney efficiently. We then immediately focused on the Art of Prompt engineering. Learning how to structure prompts and compose captivating scenes. We discovered the impact of Prompts and Style, color, texture, and the Seed, allowing us to understand and create Visually Appealing Images. We learned advanced Prompt Techniques, such as Weights, Prompt combinations, and other model variations. Additionally, we explored Photorealistic Techniques deeper, including Camera Settings, letting Effects, and specific environments, enabling us to create lifelike virtual photos. Moving forward, we advanced into alternative image styles, delving into vector Art, Mock-ups, cartoons, illustrations, and more. We explored the various applications such as website design, character creations, Cover Art, showcasing the versatility of Midjourney. We concluded the course with learning how to finalize images using Adobe Photoshop and similar tools. Focusing on image enhancement and adjustments. We discussed Ethical Considerations, Copyrights, and important Terms and Conditions of Midjourney. We also explored the latest trends and future developments in mid journey and AI in general. Lighting the roadmap for future advancements and the possible near future with AI. Now, for the final thoughts and recommendations, first of all, I congratulate you on completing this course. You have now acquired valuable skills in utilizing mid journey as a graphical Generative tool as you continue your journey, here are some recommendations. First is to practice regularly. Keep experimenting with Midjourney and continue honing your skills. The more you practice, the better you'll become at generating incredible images. Engage with the Midjourney Discord community to explore in new mediums and possibilities. Collaborate with fellow users to exchange experience, prompts and stay updated with the latest advancements. Then expand your knowledge by exploring Other popular AI tools like Dolly, lexica and other emerging tools in the field of Generative Content. Experiment with new styles and applications. Push the boundaries of your creativity by exploring new and unexpected image styles and applications beyond what you've learned in this course, challenge yourself to create unique and diverse content and also stay informed. Follow Ai and Midjourney for the latest developments. Check the roadmap and advancements to stay at the forefront of this exciting evolution. And remember, mastering your Prompting skills is an ongoing process. And with continuous learning and exploration, you can unlock your full and even unexpected creative potential. The best of luck on your journey. And never stop learning. Goodbye