Midjourney for Beginners: Step-By-Step Guide to Generating AI Art | Arnold Trinh | Skillshare

Playback Speed

  • 0.5x
  • 1x (Normal)
  • 1.25x
  • 1.5x
  • 2x

Midjourney for Beginners: Step-By-Step Guide to Generating AI Art

teacher avatar Arnold Trinh, Marketing Director, Artist, Designer

Watch this class and thousands more

Get unlimited access to every class
Taught by industry leaders & working professionals
Topics include illustration, design, photography, and more

Watch this class and thousands more

Get unlimited access to every class
Taught by industry leaders & working professionals
Topics include illustration, design, photography, and more

Lessons in This Class

    • 1.



    • 2.

      Class Project


    • 3.

      How to Set Up Midjourney


    • 4.

      Midjourney Settings V5


    • 5.

      Formula for Prompt


    • 6.

      How to Write Prompts on Midjourney


    • 7.

      Midjourney Basic Parameters


    • 8.

      AI Art Generation Process


    • 9.

      Generating on Web Browser Interface


    • 10.

      Examples of AI Art


    • 11.

      Replicating Art Styles


    • 12.

      Expanding Images


    • 13.

      How to Upscale Images


    • 14.



  • --
  • Beginner level
  • Intermediate level
  • Advanced level
  • All levels

Community Generated

The level is determined by a majority opinion of students who have reviewed this class. The teacher's recommendation is shown until at least 5 student responses are collected.





About This Class

Want to learn a new skill that will open up a whole new world of possibilities for you? Artificial intelligence image generation is the future! Welcome to the world of Midjourney and AI art generation! Where your potential is as far as your imagination takes you.

With step-by-step instruction and hands-on exercises, you'll learn the basics of this application in no time. This class will guide you through everything you need to know to get started with Midjourney.

Designed to be more than a crash course – it will give you a solid foundation of principles that you can apply to any type of creative endeavor. These lessons are fit for artists and creatives of all levels, from those who are just getting started to those who have years of experience.

In this class, you will learn:

  • How to use Midjourney to create AI generated images

  • Tips & Tricks to navigating Midjourney 6

  • How to create the best prompts

  • The different styles of images you can create with Midjourney

Don't miss out on the opportunity to be a part of the revolutionary world of AI-generated art. With this Midjourney course, you'll learn the latest techniques and technologies in an easy-to-follow format, allowing you to quickly and effectively create stunning, unique pieces of art.

You'll be provided support and inspiration every step of the way as you discover the full potential of AI-generated art. I'm so excited to be your instructor for this course, so let's get started!

Want to learn more about advanced prompting techniques? Watch my Advanced Prompting Midjourney class. 

Want to learn about practical use cases for AI Images? Check out my AI for Creative Directors class.

Meet Your Teacher

Teacher Profile Image

Arnold Trinh

Marketing Director, Artist, Designer

Top Teacher

I'm Arnold, a lifelong creative and have been a solopreneur for the last 7 years.

My goal is to empower your creativity and craft your skillset for the AI revolution.

With over a decade of experience in creative marketing, I've had the privilege of shaping stories and working on creative direction for fortune 500 companies to small independent brands, learning every step of the way that the most powerful content comes from a place of authenticity and shared insight.

Over the last year I've spent considerable time experimenting with and pushing the boundaries of AI in professional creative work. This exploration has not only transformed how I approach projects but also opened up new avenues for creativity that were previously unimaginable.
See full profile

Related Skills

AI & Innovation AI Tools
Level: All Levels

Class Ratings

Expectations Met?
  • 0%
  • Yes
  • 0%
  • Somewhat
  • 0%
  • Not really
  • 0%

Why Join Skillshare?

Take award-winning Skillshare Original Classes

Each class has short lessons, hands-on projects

Your membership supports Skillshare teachers

Learn From Anywhere

Take classes on the go with the Skillshare app. Stream or download to watch on the plane, the subway, or wherever you learn best.


1. Introduction: Image creation using AI has opened up infinite possibilities for aspiring artists and it's actually not that hard to use Midjourney, the leading AI image generator. Hi, I'm Arnold and I've been in the creative world for over a decade now. Midjourney has made it extremely accessible for anyone to create artwork. In this class, we're going to learn how to create your favorite art styles using AI. More specifically, I'll teach you how to write a better prompt and to dive into every individual aspects of generating an image. We'll go through case studies and dissect how certain keywords affect a prompt. I'll also guide you through all the essentials so you can have a step-by-step guide to navigating the journey. In this class, we'll discuss how to write prompts for better Midjourney renders, finding inspiration for your prompts in the community and how to navigate through the Midjourney platform. This class is for all artists, whether you're just getting started on Midjourney or you've already done a few generations. For your final class project we'll be creating your very own AI generated image. If you're ready to begin, let's get started. 2. Class Project: One of the best ways to practice on Midjourney is actually to do a project. For this class, our project is to help you make a work of art. Now, most of the big struggles that come about when somebody is trying Midjourney is that they're just creating a random image. Most people that go on for the first time, just type in something and generate an image out and it's like, that's cool and just give up. But the thing is, Midjourney is so powerful because it is all based on your imagination and how you guide the prompts. In the course, we're going to learn exactly how to guide your prompt and make it specific in a way that Midjourney understands what you're trying to create. The principles you're going to learn in this class along the way, is going to help you build the foundations on how you're going to be able to create and mold a Midjourney prompt so that it comes out exactly like a work of art that you want it to come out. The keyword here is it's artwork purpose-built with intention. This is how you really push the platform to its limit. Now, make sure you post your progress along the way and submit your final project to the project gallery, because that way I can give you feedback and help you write a better prompt if you're lost or if you're trying to figure out how to get this exact thing to work out. Along the way as we go through the different steps of the class, keep in mind how you're going to use certain techniques to craft your work of art into exactly where you want it to be. The next lecture you're going to learn about how to setup and get started on Midjourney. 3. How to Set Up Midjourney: Let's start and talk about how are we going to make our art on the Midjourney platform. Now, to start off, we need to navigate to midjourney.com and on midjourney.com, we going to click on "Join the Beta". Then with Join with Beta, it's going to open up an app called Discord. Now, Discord is mainly a chat app, but Discord is where Midjourney lives. Once you get on Discord, you're going to accept what's going on there. They're going to run you through a security check and verify that you're human. Once you go through all that, you'll join the server. Then from there, there's a lot of other things that are going on, but really we just need to navigate to the newbie section and this is where you're going to be making your art. Now, newbie section is where you get to use their platform and create digital images. It's also where the non-paying members are using this platform. If you are a paying member, you do get to create your own private room so it's not spamming crazy like this, but with the newbie rooms, it's actually pretty cool because you do get to see what people are creating. As you scroll around, you get to see what people are typing in. Like this one is an Aztec woman wearing feather clothes. There's that, and then there's these other ones too. You get to watch what people are doing. Ironman. This one's really cool, a handmade drawing of Ironman. But basically, if you want to make something and you're a non paying member yet, then you'll be using the newbie room. These are newcomer rooms, newbie room right here. Now just to see what people are making. Let's jump into making our first image just so I can show you how powerful the platform is. Now, to start off with an image, you're always going to do the slash and then type imagine. Now after you've typed that press "Space", and then this prop pops up. Then from there you can type what you want to generate an image of. For example, just to show you how crazy expansive this is, I can type in a car model. Maybe Porsche 911 driving along the ocean side in Hawaii. I'll do that, press Enter and then I'll let it load. Now as it's loading, it's waiting to start because other people are also creating images at the same time. It's going to take a little bit of time, but it does give you an update generally. Here it is, zero percent, it just started just putting things together. If you watch it, it starts piecing things together slowly. Here's 31 percent. It's pieced this in. Here 62 percent. It started piecing a little bit of the other stuff in too. Then, yes, 93 percent. We're almost done. Then you can also see other people's stuff working too. But it looks like everything's already done. Here is what we got. It just finished generating all these images. Now, we can click on Open in Browser just to see what it looks like. Zoom in a little bit. Now, isn't this pretty crazy? It just created images of this Porsche car to the exact model, just through words. There are lots of ways that you can control how the image outputs as well. For example, we typed in just the Porsche car. We could do imagine again, prompt Porsche 911, driving in the countryside with snow in anime style. Then they do have this mode cod, niji mode, which is focused on anime. I do know that function. I'm going to type it in. Later on down the lecture, we're going to talk about different functions and different parameters that we're going to put into to fully customize and have a bit more control over what type of images it generates. Let's go with this. Then it's loading. Waiting to start, and give it some time, then it should start any moment. Now, it does take a bit of time when it's busy. Like right now it's really busy. It's about morning in the United States right now, so a lot of people are on. Here it is, it's just getting started. It's at zero percent, but it's going to start building it. In the meantime, you can watch what other people are doing. Here's this one, full perspective of a male Viking. Here's what this image looks like. It's pretty realistic. Some of it gets really good. Now the newbie channel is not where you get them art, but you do get to see a wide variety of what the platform is capable of. I'm going to show you in the future how to browse the main galleries to see what other people are doing. Here is what we got. We did this in an anime style and this is what it rendered out. Kind of cartoonish and it looks pretty nice, it is what you asked it to do. That's the newbie section. That's where you're going to be practicing all the things that you want to test out. As somebody that is just starting on Midjourney, this is where you're going to be living. If you don't have a paid membership yet, you can't start a private room, but this is fine. You just have to work with having all these different pop-ups coming on, which is fine. I did that too for awhile until I really liked the app so then I had a paid subscription. All you're going to do is test on here. For the rest of the lectures, as we're gone through different practices, you can do it all on this newbie section for free. This is the intro section just to get you affiliated with the app. Now, let's move on to the other parts and get more specific into Midjourney. 4. Midjourney Settings V5: So now we're going to talk about the midjourney Settings. I'm gonna go through every individual setting. I'll describe it to you so you have an understanding of what to expect as you go into the Settings, talk about the midjourney Settings. And you can access that by typing in the chatbox slash Settings. And then when you type that is Settings box will pop up. So we just have it up already. But here it is again, in this Settings box, you have access to a lot of these different things. So we're gonna go through each of these and explain what it is. So you have an idea of how to think about which one to choose when you are running Midjourney. So these, the first column, you have all the Midjourney versions. And Midjourney versions are basically for people who have been here for awhile and they've been creating throughout all the different versions. And they want to tap back into it because every different version has a different way of rendering the certain prompt. And when they're making their Art, for example, I, their style is in a certain way. They're used to making something in Midjourney version to, for example, and what version three version for the image comes out different and even versus five admire be significantly different than from Midjourney version two. Those people are going to want to still have access to Midjourney version two. So they would choose that. Now, as a new user, I would recommend having the most up-to-date Midjourney because that would give you the best high-quality renders or that would be the most capable of creating whatever it is you're trying to create because it understands us better. Now, Midjourney has also released 5.1 and 5.2. We're currently on 5.2 as I am creating this video. And 5.2 is a small upgrade from five point 1.5 for one is a minor upgrade from version five. Now, that shouldn't concern you if you're new users just use the newest one. That raw mode mode is interesting. You don't have to do it. If you want to do it. Raw mode is when you enter a prompt. For example, if I type in blues, blue skies, that blue skies might come up with blue skies or it might come up with its own creative interpretation of it. So with Ron load on, it would be more literal and it would be more towards the blue sky versus having Midjourney come up with something completely random and crazy. Now, they've also included niger version four and Nietzsche version five. And these versions are basically this way of creating more animated style images. So if you're generating midjourney and you want something more cartoonish, something more anime style. Nietzsche is the way to go. So you would click on those. But I prefer to create something that's more universal, not too much animated styles. So I just don't have mine on. Now, there's stylized versions as well. And this is basically, you can have very low where Midjourney doesn't go super crazy or stylized very high where a Midjourney puts its creative interpretations of what you're saying into it. And with very high, you're going to see a lot more different variations versus on very low and public mode is all of these renders that we're making is available to be seen on public. If you want to go in private mode, you need to have a very high-paying account for it. That would be their professional account that most people are most people will have it on public Mote Marine mixed mode is it lets you have the capability of remixing your renders. I always have it on. High variation mode versus low variation mode is, as the images are being rendered, are they going to be a lot different from each other or are they going to be a bit similar? So as I like to test and see how many different variations that my prompt has, I keep it on high variation because from there you have the most flexibility and you can still get what low with variation would have. There's Turtle, turbo mode, fast mode, and relaxed mode. These are the render speeds. Turbo mode is if you want to have the fastest runner speed, like crazy fast, fast mode is what we typically all have. It's like regular speed, but this takes out from your hours, like you're a fast hours. And then there's relaxed mode where this is the slowest mode, but at the same time it doesn't take out any of your time, your fast hours and reset settings is to reset all the settings. Now in the next lecture, we're going to talk about formula for Prompt 5. Formula for Prompt: In this section, we're we going to learn about prompts and Midjourney. Now, Midjourney uses a language model to understand the prompt. It understands what you're trying to tell it to make based on a language to model. Now, what does that mean? It means it uses everyday speech, like how we talk to generate the images based on what you're telling it. You can speak to it as if you're talking to another person, which is a lot more helpful because you can actually just speak to it normally and It'll figure out what you're trying to say. Now, the structure of how these prompts work is, the first part of the prompt is where you are giving it the concept. Now, the tint and the tip that I like to give for that is, as you are building this concept, give it a setting, like define what is going on. The first part you would have whatever is most important. If you are trying to make something that's animated, maybe you have anime first. If you're trying to have something that's focused on a cat, have the cat first, because that's the first thing it's going to build on. It's going to build first on the thing that you are giving it and then after that following that, it'll be coming up with what else you are giving it as well. You first off start by building the scene, and then after that comes the parameters. These are parameters such as the size, the references, the type of systems it's using. Maybe it's using an old version a newer version of Midjourney. Basically, you're building from giving it the scene. You're building this image out, and then finally you're putting these parameters outside of what you're building so that it follows these rules. Now, once you understand that creating Midjourney images out of prompts is going to come like second nature. Now in the next part, we're going to talk about exactly how to create an image and I'm going to run you through making an image based on a prompt that I give it and actually, many prompts that I give it because you're going to learn that mid journey sometimes take many different prompts and editing of prompts to finally get to the image that you want to create [MUSIC] 6. How to Write Prompts on Midjourney: In this lecture, we're going to talk about the structure, the basic anatomy of a prompt, and this is the most important part of Midjourney because your main job is to craft a powerful prompts so that Midjourney understands what you want to do so then it can generate a beautiful image out for you. We're going to go on Midjourney here and we're going to type in just a basic prompts so you can get an idea of what's going on. I'm going to type in imagine, and then with a prompt, I'm going to set the scene and say, cat at the beach, surfing. Now, I do want to emphasize it's surfing. Cat surfing at the beach in Hawaii. Then that is the main part, so you have two parts to notice in here, is that this first part is going to be your concept. Then once you finish the concept, there are parameters that you can give the prompt. I would add these parameters and we'll go over some more in the future. But this one I'm going to first do is the aspect ratio. Aspect ratio is the size of it. Generally by default, the aspect ratio is a square, so one to one, but I do like a wider screen, so I'm going to make it three to two, so that's three on the bottom and then two up. Now it's going to come up with a wide image based on this concept that I gave it. We'll go from there and then we'll press "Enter" and then it's going to start generating this image for us. This is what we got, cat surfing at the beach in Hawaii, and here's the image that it generated out for us. Now, this is great, I like what I see here. The thing to kick note or the thing to notice is that if you get an image that you don't like, this is where you go back and use the same prompt and continue generating different iterations of it, so you can go and either remix this and then reroll the dice and get a four other ones, or if you like any of these, you can press on one of these and then upscale it, so it'll be one, then two, then three and four. You can click one of these and then it will give you a high resolution photo of one of these. But if you do like one of them and you want to expand on them a bit more, you can click on this, so it's like variation 1, 2, 3 and 4, and then it can build on top of these ones. I do like maybe this third one, so I'm going to press V3 and then it's going to build on top of it. Now, when you do that, you have this option of adding some more in here if you want, so maybe I want to make it more cartoonish, so I'll add in another parameter, I'll give it, niji. Niji is the mode that is in midjourney. They recently added this type of plug-in or this type of mode that you can activate, so when it does niji mode is going to look more cartoonish and more animated like. Now, I could also go to the front of it and add in an animated style Cat surfing at the beach, or I could just leave it in niji if I feel like this is enough to make it look cartoonish. I think it's enough, so I'm going to submit this and just see what it prompts us with. I've put in the niji mode and then now it is waiting to start to generate our image out for us. Now it is regenerating what we see here in the third image. As this is generating, we're seeing it being put together 71 percent and it's going to finish. Now, it just finished. Sometimes it turns out weird as you can see here, this one turned out weird, this one is closing his eyes. You get some queue. You have rendition sometimes. None of these turned out the way that I want it, it's not really cartoonish or anime. I'm going to re-roll it, and hopefully something comes out that's right. Now, if it doesn't come out right, we're going to go back and we're going to do it again. Just to be safe, I'll also do it again. But let's wait for it to start with a re-roll and see what we get. Now, it's unlikely that it's going to change so drastically or so differently from just the re-roll, but I wanted to show you for example what a re-role looks like. Now, this is 75 percent complete, looks super cute, and here's what we got. Cat's doing some weird things. This fourth one is the only one that looks okay. I don't like this at all. I'm going to go back here and then I want to change this up again. I'm going to add an array style cat surfing at the beach in Hawaii. Aspect ratio 3:2. Take out there before, I'm going to add niji mode again just for safe measures, and we're going to submit and see what we get. We are waiting for it to start. With this, I put anime style in front, so it prioritizes anime style, and then after that, it's going to add the other prompt a bit, so cat surfing at the beach. We do see a bit of change. Now it looks a little less photography like it's more drawn on. I do still think I would like a bit more cartoonish, so I could edit it and go on the same one that we used anime style, cartoon drawing just to give it as much parameters are as much of a guidance as possible to make it something that is drawing cartoonish. Then we're going to go here, I want to delete the V4, and then I'm going to add a niji mode, and then I'm going to let it do its thing submit. Now that we got this back, this is for sure more drawing and more anime style. This one for sure I really like the second one. As you can see, sometimes these things take renditions, so you have to do a couple of renditions to get to what you want. Now, this is great. If I were to choose the second photo, I would press U2 and this would upscale that photo into something that is high resolution that you can use wherever you want on the Internet. Now these are cool, I actually like how these turned out. They're really cute. Here's what we got, cat surfing at the beach in anime style. Next, we're going to talk more about the basic parameters that you can add onto your prop. These are parameters that most people should know as you're getting into creating art on the Midjourney, you should know these parameters because they're going to help you a lot significantly with the way that you are creating. Then after that, we're going to go into some more advanced things that you can add and ways to manipulate your image into something that you want [MUSIC] 7. Midjourney Basic Parameters: We've got the basic instructions of how to generate an image. Now we're going to go through these specific parameters that you can add at the back of your prompt so that it controls and gives it a better control over what your images turn out looking. Let's start by creating a prompt again and I'm going to do cat surfing at the beach just to be simple. Now we'll add in the first parameter that we talked about, which was aspect ratio. Aspect ratio, like I said, can either be wide, square, or vertical. Back in the day with the other Midjourney versions like Version 3, you're able to have a wider range of different aspect ratios but as of current, in Version 4, there is only 3 width and 2 up. 2 width, and 3 up or 1:1, which is a square. We saw already 3 width and 2, 3 by 2. Now let's make a vertical image and make it 2 by 3. We're going to press "Enter" and it's going to start generating a 2 by 3 image of this cat surfing at the beach. Here we have our cat surfing at the beach images and you get to see a couple of cool little renditions. Right here, we're going to go and edit some of the background. Let's say you don't want any clouds in here. Let's go and add another parameter, which is the no parameter. We're going to imagine prompt is cat surfing at the beach and then for this parameter, we're going to add a no. Now the no parameter let's you make sure it doesn't have something. For this example, we're going to have no cloud, and depends what else you want on the image. Maybe it's no boogie board, no designs on the surfboard or something, but we're going to have no clouds because there are some clouds on these images and it's an easy way for it to just come up with a new image that shows exactly what it does. We're going to add no clouds and we already did the aspect ratio so I'm just going to show you no clouds and then we're going to generate it and see what comes up. Here is what our image came up with and it is a bit more sandy, it's on the beach. This one does contain a little bit of cloud. This one doesn't, this one doesn't, and this last one doesn't as well. So it did follow what we said. We didn't want any clouds and we still wanted cats surfing at the beach, so it gave us what we wanted. Now there's also another function that we should know, which is the quality function and quality adjusts how these images come out looking. I'm going to show you two different ones and then we can compare the quality. Now, I'm going to start with opening up prompt again and then I'm going to type it cat surfing at the beach and then we would do double dash and then q and that is quality and we're going to do 0.25 and then launch it. Then we're also going to do one that is five. Cat surfing at the beach 5 and go. Now we have two loading and we're going to compare the difference once it finishes rendering. This is the first one. This is 0.25 quality, and then this is the second one with five quality. Now you can see the difference is that there's a lot more detail in the images. Like there are details on the fur, the face of the cat, this cat has a reaction like an emotion to it whereas the other one, the 0.25 quality is more of dreamlike and more smooth in a way. Depending on what type of image you want to do, take note of the quality that you want to add into your prompt and to your parameters. Now I'm going to show you one called weight and weight is what puts emphasis on certain parts of your prompt. Now, for example, the ones that we've been making have been cat at the beach, but maybe I want to do an emphasis on the beach with the cat just being part of the beach. I'm going to do a prompt where I would imagine, now if the focus is on the beach, let's put beach first. Beach with a cat surfing. Now the cat surfing we'll put a smaller emphasis on a cat surfing, and then we'll put a bigger emphasis on the beach. The way it works here is that there is four emphasis on the beach and then there's one emphasis on the cat. What that is, is like the beach has four times as much emphasis as there is cat. The thing with these two is that I had the beach show up first so that it understands that beach is more important than having cat first and your goal really is to make it as easily understandable by the Midjourney Bot so that it can understand what you want to do and spit out an image that works for you. Here we have the emphasis on beach with a bit of emphasis on cat and surfing. We're going to press "Enter" and see where this prompt takes us. Now it shows us this prompt with a couple of interesting images that cannot really relate to what we wanted to do. We wanted a beach scene with a surfing cat in it and I guess having with a cat surfing is confusing. We should rephrase this. Now, it did put weight on the beach so as you can see, there is a lot of beach in here, but there is no more surfing cat. We're going to go back and do this again. We're going to do beach with a four emphasis with surfing cat. Now surfing cat maybe we'll give it a 1.5 and we're also going to add the no parameter. We don't want any humans in there. No people or no humans, so we'll put no people and we'll make sure that everything is correct. So beach emphasis and surfing cat. Now, let's see what we get. This is the prompt that we got. We ended up getting a couple of interesting images. This one has a cat holding a surfboard. This one just a cat or dog at a beach. None of these really have that cat that is surfing at the beach. Let's change this up a little bit more so beach scene and then we're going to put an emphasis on that with surfer cat. Now we want it to be as easily understood by the AI as possible. Now we'll put an emphasis on the cat. Then we'll also add no people and no dogs because we want the focus to be on the cat. Then we'll let it run again. Then here we get this scene with a cat on a board, the scene with a person holding a board, and a whopping cat. Now the closest one is this cat on a board but there is a human over there. We need to get rid of that human somehow. Why is there a human there? We're going to go and revise this and then we're going to make sure no people, no dogs, no humans, characters, or surfer cat and then we're going to add only surfing cat in water. I guess we put emphasis three on the cat this time for a 50/50 weight between the beach scene and the cat now because I guess earlier there wasn't enough cat. Now we have a surfing cat with three in water and only cats, no people, no humans, no characters, no dogs. Now, let's see what we got. Now we have a cool rendition. This is what it came out with. There is still some humans on these ones. Strangely, this one is maybe a boat in the distance. That's nice. This one is a rock in the distance, so that makes for the most sense. Now, this does work, but it has something weird, which is the hat. We're going to re-roll. We're going to re-roll that one. The second one. We'd like that one the most, so we're going to re-roll it and see what else we come up with. We're also just going to throw no hats. Surfing cat. Now a cat in the water would be nice. Beach scene with surfing cat. Now, maybe we do a dash surfing cat in the water. Now we probably don't need only cats, no people, no human, no hats, no characters, no dogs. Then we'll submit just to see what we get. See sometimes it does take multiple renditions to get to where we want to get. Well, this came out a little weird. I don't know what's going on here. Sometimes it does get really weird with these renders From here, we might need to just go back and redo it. Now we see cat surfing at the shore break and there is an emphasis on the sandbar. So you see a bit more sand on here. As you can see, I put some emphasis on the sandbar, and these images that are generated have a bit more sand than the previous one that it came from while also showing this cat surfing, which is what we wanted. Now this first one is my favorite. He looks super cute. If we want, we could do some more versions of it or we can upscale it. I think this is already really nice. I do want a wide version of it though, so I am going to do a variation and I'm going to make this aspect ratio 3:2 and I'm going to submit it and see what we come up with. As a reminder, there is weight on sandbar, so it's prioritizing sandbar a little bit more and adding a sandbar to the image. As you can see, more sand here and here it is just building up our image. Here we go. There's our surfing cat. It's so huge. There's a cat just getting smacked by the water. Then if we want here too, we can add some more quality to it. We can upscale it, whatever we want with it and it looks pretty good. I like how it came out, so I will upscale the second one and if you feel like there is not enough quality, we're going to go back one last time and we can add another one to this, another parameter. This one we went over earlier too but it is the quality parameters. I can put quality at five. Then I'm going to submit it again and then see what we get. Right now it's working on two different jobs. It's one, upscaling one of our images, and it's also generating another one. Here is our newly generated image with quality five. This cat is getting smacked by the water. It's getting scooped out. This one is jumping. First one is my favorite. Actually, I still do like the ones that aren't as upscaled so much, so the ones back here were my favorite. We've got the version 1 wide aspect ratio. These are some of the parameters that you need to keep in mind as you build your first Midjourney prompt. 8. AI Art Generation Process: Let's get to the funnest part and that is creating your AI art. Now, how are we going to start? I'm going to guide you through the process, the thought process that you have to go through as you're creating your art. Now, the first thing to keep in mind is we need to create our center of attention or main attraction. What is our main attraction? Now, through this example, I'm going to be making my own as well. I'm going to be typing up a prompt so you can follow along and see my thought process and work through it together so that we can together make an amazing art prompt that comes out to be something amazing. For our center of attention or main attraction, we're going to start with maybe a man in the jungle walking through a tropical forest, waterfalls in the background. Now what you want to do here is you really want to think about what you want to create and the more detailed, the better. Here I am doing a man in the jungle. I'm not doing a jungle with a man. I'm want to focus on the man in the jungle and he's walking. You have to think about what he's doing specifically and he's walking with waterfalls and a tropical forest type of scene. Now you could even get more specific that he's man in the jungle with a hat or with goggles. You really want to imagine this as best as you can. The more specific the better. But if it's too specific, sometimes it could throw the AI off. It's a balance of how specific you can get without confusing the AI and that's where the art process is. We'll continue from here. Waterfalls in the background. Small pond. Now see, this might confuse the AI, the small pond, but let's see if it read it and does well with it. We have that. Now, the next part you want to consider is the composition or the viewpoint. Now, I want something that's grandiose feelings. I'm going to write massive scale and I'm going to add intricate details too because I want to see lot of intricate details in this piece of work. You also want to think about lighting, like what type of lighting you want this to be? Is it twilight? Is it afternoon light? Is it maybe lit by artificial light because you're doing a scene in the city for example or city light, night light? Think about how you want this scene to be lit. For this part, I probably want an afternoon scene. I liked it being in an afternoon. Afternoon and sunny just to make it less confusing for the model. Then one of my favorite concepts is, as you could even see here, is this animated type concept. I'm from the '90s, I was born in '92 so I want to do a '90s animate concept because it evokes a sense of nostalgia for me. I want to put '90s animate a concept. Now I have the angle, I have a massive scale angle. I'll have details on what the forest street and what the scene looks like. It's afternoon, and the art is done in a '90s concept. Now you could even get more in-depth and specific. Some people do that by having maybe artists. I could have in the style of, sorry if I pronounce his name wrong but Hayao Miyazaki, who is the person in charge of Studio Ghibli. We could also add style of Hayao Miyazaki comma, style of Studio Ghibli, that has these parameters or actually not parameters because we'll get to that in a second. But these guidelines to reference as it is creating this art. I liked that. Then I'm going to finish this off with the parameters of aspect ratio being 3.2 and you don't have to add it, but it is Version 4. Now I do like this a lot. We're going to go and make this happen. Now, I want to copy it at all and slash, imagine the prompt pop up, and here we go. We came up with this art and it looks nice. Let's open in browser just to see. It's in their browser. This is what it looks like. Which one do you guys like best? Nothing with a waterfall. I do like a waterfall. Let's go with a little reroll and see if we like anything with the reroll. Here's our rerolls and still does not. This one right here has waterfalls. Not the type I like. Let's go with another reroll. Sometimes we have to do that. Anyways, this is what the images look like. Actually feeling these look pretty nice. The second one has waterfalls in the distance but barely see it. First one is cool. Third one is interesting. Fourth one is cool too. One thing we could do is we could put some emphasis on the waterfall. We can copy all of this and add dots. Imagine, I'm going to add emphasis on the waterfalls. I'm also going to edit this prompt. I don't need background probably because the main subject is already man in the jungle walking through a tropical forest. But now that I'm looking at it, I don't actually need to write in the jungle. I could probably do man walking through a tropical forest with waterfalls, small pond and a massive scale with intricate detail, afternoon and sunny and '90s animate concept art and the style of Studio Ghibli and the aspect ratio of 3.2. I like this. Let's see what we get with a re-edit on our prompt. Here is the re-edit of our prompt and cool, the second one has the waterfall. Let's go in browser just to take a look in more detail. This one too the first one has waterfalls in the distance. But this second one is really cool. I actually do like this second one a lot. When we find something that we like, we can upscale it. The second one, I like that, so I'm going to upscale the second one. Now, if you do want to see some variations, you can also click on one of the variations and then go and choose from the variations which one you like. But I do like this second one already, so I'm going to upscale it and then see what we get with an upscale. We got this with our upscale and it looks pretty good aside from the fact that the protagonist, it looks pretty good too, I was going to say he looks a bit stretched out. But if you do feel he does look a bit stretched out, we could go back and give it a another upscale or some variations. Now, I do like how it looks, but let's make some variations just to be safe. In this part, we could add small man. As we just did that, our first upscale came in. We had to do it again since it was paused and our first upscale actually looks really good. I could definitely use this for something. Although it doesn't hurt to look at what our new variations did and it is not the style that I like so much. I'm glad we did the upscale and that the upscale here worked out really well. Now, this is amazing. This is a piece that I declare finished. Here it is. Here's our finished piece [MUSIC]. 9. Generating on Web Browser Interface: Okay, this is one of the most highly anticipated update that Mid journey has had for us. And this is being able to imagine on the browser. So just the other day, mid journey announced that the mobile or the web browser features are available. So now you can go on the browser and you can create images with this bar up on top here. So right now I'm on the mid journey alpha explorer page, or mid journey Explorer page. And I can type in here to imagine whatever I would type in previously into Discord. So here for example, I would type a cat riding a spaceship in space with a view of Earth O, with a view of Earth right here. We also have these parameter controls on the right. And let's go through them so you can understand what is in here. It's very similar to what we used to have or what we still have in Discord, except now it's a lot more user friendly. So we'll start with the left. This is the image size. You see a couple defaults, which is portrait mode, square, and landscape. So with portrait mode you can do three by four, which is standard Instagram size two by three, or nine by 16, or one by two, this one's really vertical. And then on the other end, you could also do six by five, and this will make it landscape. And then four by three, which is one, I typically use three by two, 60 by nine, which I also use a lot for video edits. And then there's two to one, which if you do cinematic stuff a lot of times is one to 2.5 So you can use something like this for cinematic type of generations. Now moving down there's the mode and just like with the Discord version, you have standard or you have raw with standard, it's going to do its own interpretation of what you wrote down. With raw, it's more going to listen to exactly what you wrote as in what's in your prompt. And then on the right side with aesthetics stylization is how much style mid journey is going to put in it. If you have it at zero, it's not going to go too wild and put its own interpretation of what you said onto your generation. But if you put it all the way up at 1,000 it's going to go wild and crazy and have its own interpretation of what you put in there. Now, weirdness is different. It's kind of the same, but different Weirdness is also its own interpretation of what you're putting into the prompt. But this is putting its own sauce on how creative and wild all of this gets. Now moving down, we have variety. And variety will show up as something different actually. But I'll mention that in a second variety. If you have it at zero, then the images are going to be a little bit similar. But if you have it all the way up max, the four images that you generate out would look very different from each other. So you also have a couple options here for speed, which is relaxed mode if you have the plan that lets you have relaxed mode and then there's fast mode and there's turbo mode, obviously. Fast mode and turbo mode takes up more credits. So you might finish a run out of credits sooner. So I'm going to have a variety of max to 100 and then I'm going to have just stylization at 500. And I like to keep it raw so it listens to what I write exactly. And then weirdness. We'll keep it here at 300 just to be on the safe side of things so it doesn't get too wild. And then we're going to generate this, and then I'll press Enter. And then once I press Enter, it's going to show up in the create side. Over here on the create side, you'll see the progress of the image that you're rendering. And you'll also see here that the chaos parameter is actually 100. Chaos is what is on the variety side, as I mentioned earlier, variety here is at 100, it will show up as chaos because this is what we had on the discord when we had chaos at 100, that would give the most variety. And here's the image that it generated for me, a cat riding a spaceship in space with a view of Earth. So we've got a couple interesting cats here. They're off super different. There's like a little baby cat. I like it. These generations are cute. As you can see, you have been doing a lot of generation. So now let's try again and do something with a lot of weirdness. And we'll turn the variety down to 50 just to see if the options that we have are wild and super mixed up and different. Or if it's kind of 50, 50 in between, like the last render we did had a lot of very different images. And then I'm also going to change the image size just so you can see how it turns out Square was the default on the discord, so we would always have to have a suffix to be a certain aspect ratio if you wanted to change it. Otherwise you can just do it here which will keep it at that level. So it's actually a lot more helpful. Okay, so for this one I'm going to do something more cinematic and I'm going to have this aspect ratio as 235 to 100, which essentially translates to one by 2.35 up in here. Let's generate something that would be a cinematic scene of a cat, cinematic scene running on Mars with a space suit shot on cinematic camera. Okay, so let's run this and see where we get. All right, so here's the generations that we got and as you can see they aren't too clear under Y. So let's just rerun this again with this it's also similar to how we have it on our discord where you can re run it, copy the prompt, hide it, or a couple of other options here. So if you press Copy, you'll see that you can copy the prompt. If you press this, I'll copy the prompt. If you press this, it'll copy the job ID and also this one for the seed. These are the two things that you typically see when you would use the envelope function as you react to a generation on discord. And you have the envelope function, it'll send you the job ID and Ed. And then the bottom two is the image and image URL. And if you want to download these images, you could also press on more and download. Now, I thought this generation wasn't the best, so I'm going to re run it. Okay, so this one looks better. I like this, this looks a lot more sharp. And then here are the other images. These are just kind of strange looking, definitely the weirdness kicking in. But let's go with this first one. So we'll have a couple different options here, which is also very similar to Discords Mid Journey. First, you have vary. So if you want to vary this image and have different variations, you can have a subtle variation, a strong variation. If you want to upscale it, you could have a subtle or a creative upscale. Typically, I would just use the subtle upscale which doubles the image size. But if you want to add a bit more details, then the creative upscale is a good option. Again, you have remix if you like, this image remix. Remix is good for if you want to use this initial prompt and just change it a little bit. It'll take the structure of this and then it will change it into whatever you add in here. It's a bit similar to vary, but you have more control over what type of remix that you want to do. As we go down, we'll see Pan is when you want to extend the image. If you want to extend a top portion of it, you can press this up button. If you want to extend a bottom portion of it, you can press this down. And then same with left and same with right. It's very similar to zoom, which is right under it. If you want to expand all four corners, then you can with these two, X would double out the edges of this image to fill in, in this example, more terrain. Then you could also have the aspect ratio. And this is important if you want to make your image a square because sometimes square images are necessary. Or if you want to just change your image to a different aspect ratio. And moving on down here we have re run to re, run this. And very region, very region is actually one of my favorite features because with very region you have the ability to change something in the image. For example, Va wanted to change the sun into a moon or a different object for example. Or a spaceship. Which will probably be more relevant in this context. And so I just submitted my job to change that into a spaceship. Now of course, this isn't always going to be 100% perfect. That's what I found my experience to be. But they're updating and upgrading every single day. Now, on the bottom two, you have these three which is to reference this image. If for example, you want to run another prompt and you want to reference this image, you could press Image and then el go here. It's essentially having slash imagine and then the link to the image. Now you can also copy the style which is like the style reference feature which is REF. Then if you want to copy the prop, you can press the prompt to button here. And then the prompt will pop up here. Now we like this realistic style, so I pressed on style here. We'll just run something else that's similar scene of a cinematic interior scene of cat astronaut in spaceship shot on cinematic R camera. And let's run this. Here are some of our other renders. Like I mentioned, this is our remix. With the remix I put astronaut cat and as we see here, there's more of the cat wearing an astronaut suit. And I love this, this is so cute. Moving up, we have also the expansion thing where we changed the aspect ratio. So here you see that it filled in the top and bottom and made this a vertical video or image. So now it's nine x 16. And then here, like I said, with the very region, it doesn't always work 100% correct. Let's try again, and this time let's try on this right side and have a spaceship there. So here's the one we made, copying the style, and it also referenced that image. These are prettier, right? Except this one is bit blurry, so I don't like that. But this one looks cool, got some potential to use this. And this one actually looks pretty magical actually. I love to style this and how, how actually like CGI cinematic this looks. So that's a cool generation and as I said, the very region doesn't work perfect all the time. And as you see here, we tried spaceship multiple times and it hasn't come out. But sometimes you just got to keep trying. Maybe we'll do sunglasses. I just want to show you an example that it does work. So we submitted this job, let's see how it shows up. After a couple more tests, we finally got something. But I had to do astronaut and it completely replaced the Cat character and gave me an astronaut, which is fine, but this is how the future worked. So I did a very region and I selected this entire area which was the Cat. And then I also tried to do Cat astronaut instead, but it just gave me a lone cat. And as you scroll here, you could kind of see that editing magic in action. So this is like more of the original ones which we asked for sunglasses but got this instead. And here there is not much change at all. So definitely this feature still needs to be worked on, but everything else essentially is there. And that's about it for using Mid Journey on the Web. 10. Examples of AI Art : In this section, I'm going to show you the midjourney.com feed and how to get inspiration from it. So on the feed, this is the hot new photos at the moment. If you scroll around, you can see different stuff that's going on. You see all these different styles that you get, these really realistic style like this. You could also get some very creative stuff like this. You can also get very painted work like these two. Full of color stuff like this. This is a beautiful one here too. If you scroll around, you'll find many different types of inspiration, and it shows you the capabilities of this application. Now, cars are some of my favorite and animate stuff, so I really like these, but the main thing on this website that is important to note is that with a paid membership, fortunately it is paid, you do have access to seeing what the prompt is. For example, if we scroll down and we see something we like, then we can go and click on the information and look at the prompt. For example, let's look at this one. This one is amazing because it looks like a photo. Now, if you look at this, you can click on that and open in a new tab, and then it'll show you the whole photo, but also you can just click on it and then I'll show you the different crop that are used. Here, let's just break down what the prompt uses. It starts off by highlighting the mustachioed general inspecting the rank and file troops. Prussia 1900, Hasselblad photo. Now, what this does is it centers its main subject is the mustachioed general, and then around it is the inspecting the rank and file troops. Then this is the Prussian uniform and 1900s. Now, the Hasselblad part of it gives it this look and feel. This is like shooting it out of a Hasselblad camera. The photo just makes it seem like a photo element. So this is a very interesting way to use mid journey to create something like this, like a photorealistic type of image. Now, if we scroll down, we'll see some more fantasy type stuff, so let's go and take a look at this one. This one's super fantastical. Just to see what they use to build it, let's look at this prompt. Sometimes you're going to see super long prompts like this. This is a very interesting one for us to look at and to break down. Let me zoom out a bit so you can get an idea of the image. First off, I'll start off by describing this girl, the protagonist in here. It says flirtatious, frisky, anime girl in her living room in a futuristic space. So that's the setting, space station. That's the setting. She's at her room in a space station and she's got the flirtatious look on her, the anime girl look. Then they start adding rank and file, cinematic shot, dynamic composition, perfectionalism, ultra detailed hair, ultra detailed skin, ultra detailed eye, and closed mouth. It's starting to go in and add all these little details. It gives the cinematic look. You can see there's a bit of blur in the background, a dynamic composition, so if she's posing, not just resting. Perfectionism, this project gives it a perfectionist look, and also detailed hair, skin and all these facial features. It gives this very detailed hair in here. I wouldn't be surprised if they add detail later down in the prop as well, although it's not mandatory, and a closed mouth. That's what the prop here did. Then there's a dress and super hot culture and future dress for women, high fashion makeup. It's giving this whole look based on that. Then finally, it adds art by Emil Memmoth, Giger, Marcin Nargaba, Rebecca Millen, and award winning realism and close-up shots. This pulls inspiration from a couple of these different artists and it adds the element of award winning realism. That's why you can see that it's real. It doesn't look completely anime, which was the original reference. It gives it a close-up shot-ish, but it's super close-up. This is probably a 50 millimeter, so you could also probably type 50 mm in there to get a similar type of framing. Then the composition is balanced and dynamic, futuristic soul in the background. Now, these are more of the extra details and sometimes the majority doesn't use them. As you can see, there's no soul in the background, so was in part of Korea, but the composition is pretty balanced. It has the subject in the center. It covers about a third of it, but it also expands to the side a little, so it does have a pretty balanced framing. Now, the image is shot with a Canon EOS 5D Mark IV and Fujifilm Superia X-TRA 400 film. Now, this is two things that don't really belong together because the 5D is a digital camera, while this is film. What I take the Midjourney took from this is that it gave a color processing that the Canon 5D Mark IV had and it gave also the look of the X-TRA 400 film as a color emulation based on what the 5D Mark IV would shoot. This is what it came out with, beautiful work of art. We could also scroll down a bit more, and let's look for something that's cartoonish because we have been looking at all the other stuff already. Now, these are cartoonish, this is cartoonish, this is very nice little cartoonish looking thing. Let's go in and look at this. This one has a street wear girl with a pearl earring painting in street wear clothing in Tokyo, raining lightly. A very simple prop, just street wear outfit. As you can see, these pop their colorways in the street where we're all right now. Pearl earrings might have been implemented. It's not super obvious. Street wear clothing, yeah, this is a street where design, and in Tokyo probably gives it this back city lighting and light rain. This doesn't get too in depth, but we can find something they gets more in-depth because we can that way learn more from how things were built. Now, this is a Ninja Pikachu in watercolor art style, very interesting one. This is also very interesting night vast landscape award winning concept, so let's take a look at what they used in here. Not surprised that they use a lot of different props for it, so let's break it down. Now, the first part is they used a night vast landscape. This is a landscape type of photo. Setting the scene for that, award winning concept are in a highly detailed angularium huge gothic castle in exploding seawater, stormy windy sea, huge waves, floating lights all around. Great. So that is building this entire scene. Nebula sky, chiaroscuro, hyper real surrealism, luminism, ultra-wide angle. Now, this is giving the framing of how this is coming out. This is a wide angle type of shot, and this is what's creating for that. Then now the effect of it, ultra detailed, realistic, ultra photorealistic, hasselblad, high-definition 64K cinematic in color grading. So now this is what's giving it this effect, what it looks like. The ultra detailed and photo-realistic look is giving it this drawn on, but also photo-realistic look. Now, it also adds the cinematic and color grading aspect, which gives it this type of dark blue and teal orange effect, they would call it or teal and orange sometimes, which is what this look is, and then it goes into add depth of field for realism again, film lighting, rim lighting, intricate. This is giving it all the different types of lighting, and it adds in some realism, maximalist detail, very realistic. Photography by Carli Davidson, Elke Vogelsang, Holy Roman, and warm colors. Those last few were probably not highly referenced, but it does play an effect and how this came out looking. So this is a beautiful piece of work with a lot of intricate detail into it. There you go. If you go and reference the mid journey website, you have a lot of different things to reference and check out and to look at how they made it. It's super important to look at different prompts just to show what's possible. In doing so, you can get some inspiration yourself. They're things like this with a lot of interesting colors. Now, this doesn't have too much going on, so it probably did rely on a bit of luck, but there are other very colorful pieces that don't rely on luck and that are very much crafted in the way that it's crafted. So, yeah, play around on the website, see what you like, and then draw some inspiration from it. It's one of the best resources out there. [MUSIC] 11. Replicating Art Styles: Now that we have some inspiration that we just found, let's use that inspiration to create our own version of something similar. We'll start off with the first one, and we'll start and imagine and then prompt, so let's go and say, an anime girl in a futuristic space station or space office or laboratory vehicle, and then we'll add cinematic shot and a dynamic composition. Will also add ultra detailed hair, ultra detailed skin, ultra detailed eyes, smile, and then we'll give her a dress. We'll maybe we'll add a they had hot culture, which I'm not too sure what it is, but we can add our own style of dress, so maybe our own futuristic will do dressed in futuristic uniform, high fashion makeup. Then we'll do reference some of these artists that they had, so I'm not quite sure what the three artists they had was Emil Melmoth, and Geiger and Mason OGrabber, but we'll reference one of themself. Would do art it by Emil Melmoth. Then we'll also add award winning realism, so maybe that gives a cool touch to it. Then it will go close portrait, composition is balanced and a dynamic, futuristic Tokyo vibes, and we'll do Fujifilm superior extra 400. That'll be like the film colors that it had, so let's see what it comes up with. Cool so here are what we rendered up and we got our own anime girls are futuristic, anime girls. Now, let's go and upscale one of these, so I do like maybe the second one, it's got some really cool details within her hair. Maybe it's a uniform thing the first one is also really cool too, me let's go with this first one. We'll upscale the first one, so we'll press U and we'll upscale it. Then the journey is working on it, and then it's going to come back with an upscale version of it. We have our upscaled version. Open it in browser just to be sure. Now this is what it looks like. Super cool. As we're looking at it, let's talk about some of those parts that were used, so as you notice, it referenced the anime girl part, and I think it also referenced the Tokyo virus because it is an Asian character. As you can see too, it's placed in a laboratory, so it's an anime girl in a futuristic space for laboratory. Now, it's also cinematic shots, as you can see, it's cinematic lighting, and then the hair is detailed, skin is detailed, and eyes are detailed. Not quite as smile, but it's okay. Sometimes it doesn't get exactly it. We could go back and put some weight on the smile but for now, I do like how it looks. Then dress in futuristic uniform, high-fashion makeup, and then art by Emil Melmoth, not too sure what she looks like, but we can look her up just to reference, and then award winning realism. It does look pretty, it's got that realism field to it, and then it is also a close-up portrait, like I mentioned here, and composition is balanced and dynamic. It does seem very dynamic. It's in a laboratory that is a fast-moving hallway and she stationary, then futuristic Tokyo vibes and also the Fujifilm superior color, so not exactly sure how the superior works in there, but it does have filmish tones, which could also have gone with the Tokyo vibes or the cinematic and dynamic composition. But I do like how this looks, I'm really happy with it. But let's move on to the other inspiration that we have, which is a landscape type. We're going to go and imagine another prompt, so slash imagine. Then we're going to go and write vast landscape of the ocean. Award winning concept art highly detailed castle. Maybe should we do an Asian castle? We could do a Japanese castle or it's probably pagoda and then exploding seawater. Huge waves, floating lights, nebulous guy, and these are just examples that were taken from the inspiration. The sky stuff is going to be pretty interesting. Then we'll do some more hyper realism, luminism, and then will give it the ultra-wide angle, and ultra detailed, and photo realistic. Hasselblad is also another cool one as well because they are some amazing cameras. It does come up to be a very beautiful looking type of image that will do HD, cinematic color grading maybe, that the field. Then we'll do photo-realistic and film lighting so that we'd get some cool colors, and then realism, and then we'll go from there. Warm colors even. Let's see what we get, and also let's make it a wide. Give the aspect ratio three by two, and then let's go. We're starting to see it generate, see the outline of everything. Looking pretty cool from this preview so far , 93% almost done. Getting there looks pretty awesome, excited to see what it looks like. This is what it came out to look like beautiful, got this pagoda looking thing in the middle of a crazy roaring ocean. Now, my favorite one is these first two, so let's go and upscale it and see what we can get. Now, I do like this lighting here, this volcanic looking lighting. The second one is the Hasselblad film looking light, which is also beautiful, but I do like the first one, so let's upscale the first one. This first one was upscaled, and this is what it looks like. It's super realistic, beautiful, got a lot of things going on at the same time, It's pretty amazing. Now, we'll also run it through a remaster and a detailed upscale redo just to see what it ends up looking like. The first one is going to be a remaster and the second one is a detailed upscaled redo. This is the remastered. This is what the remaster looks like. Very cool, very artsy and no weird like fantastical away. Very beautiful. I really like this. Then for the other one, we have this other thrashing water has that gray texture to it. It looks like an actual painting with real paint. Very awesome in its own unique way as well. These were what we got from our previews. This is what we got from our inspiration. Hopefully you can draw some awesome inspiration as well, and I'd love to see what you make. [MUSIC] 12. Expanding Images: We're going to talk about expanding our image based on what mid journey gives us. We can add directions to it, left, right, and make our image a lot bigger. First of all, let's start by generating an image that we like. And it could really be as simple as portrait of a cat. Now once we have that portrait, I'm going to show you what it's like when we expand that image. And the expansion has many different ways to expand. Okay, so as you see here, we have these generations of cats, the portraits. So I'm going to pick this one right here on the top right, because it's the only real looking one. The rest are kind of fictional. Like Circa Cat. So yeah, let's just go with the most real one. I'm going to upscale it, and once you upscale it, you're going to have the option to expand this image I'm going to show you right now, which is here. So you have the option to zoom out, and that means it's going to build a scene around this cat. And there's a couple different ways to zoom. You can zoom out 1.5 times, you can zoom out two times, or you can zoom out a certain custom amount. So if zooming out 1.5 x is too much, then maybe you can do 0.5 x. Now just for example, I'm just going to zoom out two x so you can get an idea of what it looks like. And when it is doing this zoom, it's actually basing off of what is already in this image. So it's going to add onto this image as well. So it's using AI to understand what's going on in this image and then build on top of it. So as you're seeing here, you could see that it's built in the center and then it's expanding outside. And as we're starting to see this image come together, you see the legs, you know, some more legs on these ones. A bit of a tail in the back. So it gave us like a more of a full view of this cat. So some of it he's standing on a box. Some of it is him just standing, you know, But now we have legs. And that was built, that was built off of the one portrait that we have of just a face. So my favorite one here is also the top right, this one. It's the most realistic. But let's go back into our expansion section, which is right here. And then explore the other options as well. So as we're here, you also see these arrows. So what does these arrows mean? Well, what it means is you can expand that image. But instead of expanding out like we just did here, you know how it was at first just a portion of the cat, but it expanded on the top. It expanded on the left and right side and it expanded on the bottom. Now if we use just this arrow, it would only expand down that certain direction. So if I press down, and then it gives me this pen with the new propped. So you can add what you want here. I would just do furry paws and then it will expand into some furry paws down here. Now you could also add whatever customization that you want for it, but for example, then we're just going to use something simple just so you can get an idea of what it's going to look like. Okay, so as we can see, it generated some paws. Now these look nuts. Like these are kind of unrealistic. The first two are very unrealistic. The bottom two are also, you know, this one looks all right, but the bottom right one looks very unrealistic. So if I were to choose one of these photos, I would upscale the third one. And then I would have this one with the cat, the portrait that we had earlier with his expanded paw on the bottom. And so at this point, I have this one done and if you want, you can also zoom out on this. And then there's also the neat function to just make it a square if you want to make it into a square to post somewhere online. Since sometimes it could get very complicated. Say for example, you had zoomed down a couple times. If you're making like a scene where it's like a building, a very tall skyscraper or something, you might have zoomed up or down many different times. For this example, we only zoomed down, but depends on the situation that you're using it for. So as you can see here, since we use the make square option, it's building out this image to become a square because there's relatively not much information on the sides. It didn't build much on the outsides, but if you were doing a very busy scene, it would also build on top of that scene. So it will look at what's around and then it would add on top to the sides and have something that is fitting to what you already have. So here's our cat, and he's looking really cute. So just to summarize this section, you learned about how to do expansions in your up scales on mid journey. 13. How to Upscale Images: Now in this section, we're going to talk about how to Upscale an image. So what we have to do is first go on Discord. Alright, let's talk about upscaling on discord. Now, we have this image here, these four that we have just rendered. And you have the option of U1, U2, U3 for now, what this is, is when you click on one of those, it would correspond to the image here and then make it a higher-quality. Now these images aren't exactly high-quality yet. It's just a preview for you to see and understand what was just rendered. But for you to pick one and want it to be high-quality, you have to Upscale it. Now, the first image on the top-left is the number one, the top right, this number two. The bottom left is number three, and the bottom right is number four. So say for example, we want to Upscale the bottom left one will press Upscale three. And then as that's happening, you press that this is going to pop up and your Upscale is going to pop up right afterwards. And then you have this image. This is a lot higher quality than if you had just downloaded straight from this. Okay, So after you Upscale it, you have these extra options as well. I'm going to explain it as well so you can understand what's going on. Now, you have two options here, very, where you can change this image up. It's gonna be kinda similar, but it's got to change up a lot, which is very strong. Or you can vary it again and have it be subtle changes. Now, Midjourney has had some exciting updates. And these are some of the most exciting update features. The zooms right here, which makes it a certain zoom tool out, is to zoom out of this scene. I'll just zoom out so you can see what's happening. But when you press that, it's going to zoom out and build on top of this scene. So it's going to add to the different edges. So the sides and the top and bottom and it's going to bring this image out. So this would have use cases wherever you want to make an ad and you want to put some text on this image, you could, because you have extra room to now. Then these arrows, these arrows correspond to increasing a side. So if you want to increase more, like build out more to the left side, you would press the left arrow. If you want to build out more to the top, you would press the up arrow. If you want to build out more on the bottom. Like if you want to add his legs and his hands, you would press the bottom button here. Or if you wanted to build out on the right and just add more detail over here, then you can press the right side one. Now with this one, we just zoomed out and you can see that it's showing more of the door on the sides and a little bit more of the tree here. That will give it a few more seconds and this has got to finish loading. So this is where we got. And it's added a lot more different details on the sides. And you can choose whichever one you want to Upscale. Now, let's look at it and I'll show you some things. Now the hands aren't always going to be perfect right here. So you have to look around and phi the best ones. Midjourney is still developing. This one looks like the best one. So if I were to keep one, I would use this. But it does help sometimes to go into a different software and then do the edits on there just to clean it up, like Photoshop, for example. Now, that's all you need to know about upscaling and the different functions in the Upscale section. Now, I just mentioned these two, but the Custom Zoom is if you want to do zoom out 1.7, you would enter it in Custom Zoom right here. And then with make square, this would make this into a square. And that's it. Up next is our conclusion. 14. Conclusion: Welcome to the final lecture of this course. Before we get started though, make sure you post your project into the project gallery. That way I can help you out by looking and giving you feedback on your work, and also congratulations. We went through three important concepts, the first one is how to set up Midjourney and navigate around the platform. You should have a clear understanding of the settings and how to use the basic functionalities of the program. Two, we went through looking at inspiration and learning from others how to use their prompts in our own work as well. That way we can take inspiration from everywhere and use it in something that we want to create. Then finally, the most important part, we learned about how to write prompts, prompt engineering, and as a quick summary for you, writing a prompt involves having your main image, building the scene around it, and then you add in your parameters. The key big takeaway from all of these, is that you should have learned how to craft an image using Midjourney. This will give you a significant leg up against everyone else that is only just learning Midjourney and creating random images. You have the power now to create images exactly like how your mind imagines them. Again, congratulations on finishing the course and remember to submit your project in the project gallery [MUSIC]