Adobe Firefly Complete Guide | Learn to Use AI in Your Projects

1. Introduction: Adobe Firefly is a revolutionary tool designed to empower creators like never before, which provides a simple intuitive platform for transforming your ideas into stunning artworks. And now, we are thrilled to announce our comprehensive online course created to make you a master of Adobe firefly. This course covers a broad range of essential topics, starting with text to image. this module will teach you how to turn text into eye-catching limits, playing with the settings, such as aspect ratio, Content-Type, or even the lighting. Next, we will explore text effects. Diving into the techniques that can make your text shine or pop. With firefly, you can generate texts effects using any kind of material or objects, and you will learn how to do that. Re-color vector module will show you how to change the color of the vector image. You can do that with precision and ease, which will enable you to apply any brand color palette using generative fill feature, you will be able to replace any object with a few clicks or extent an image with the power of AI. Firefly is a new player in the industry and new features are on the way we will update this course as the main features will be launched. By the end of this course, you will be able to use Adobe Firefly to its full potential. So when you're ready, join us and become a master of Adobe firefly. 2. Course Overview: Hey everyone, welcome to the course of Adobe Firefly complete guide. My name's iPad. I being a technology educator over the past six years, I love creating things with AI and Adobe Firefly is one of the greatest tool I have used so far. In this course, we will cover all of the firefly features, starting with text to image. This module will teach you how to create digital art using keywords and phrases. You will learn to adjust the results which settings many of firefly. When it comes to AI art, choosing the right prompt is the key. You will learn to use keywords in the right way. Be sure to watch the video titled prompt engineering, tips and tricks. And there we will talk more about prompt engineering. Firefly currently launched a platform where you can publish your work, see other projects, even try and edit the works of others. You will learn to use that platform as well. Next, we will talk about the text effects tool. As you might already know, this is a unique feature for Firefly that lets you create incredible effects in just seconds. Been off the biggest strike is generated to fill since it opens a new door for the whole design industry. After learning that module, you will be able to remove, replace, or add a new object in an image, extend an image, and change the proportions as you wish. Everyone is talking about this revolutionary tool. Don't quit this course before watching that part. I believe best way of learning is by practicing. So I will have challenge for you throughout the course. Make sure that you are also creating your own projects and following the course by applying two instructions. Adobe is trying to approach this new AI industry in an ethical way. Their purpose is to protect the greater strides. So they have some different rules and policies when it comes to using AI art. That's why we created a standalone video. Four terms and policies of Adobe firefly. And don't vary. We will cover all these points during the course. 3. Firefly Platform: Now we are inside of the homepage of firefly platform at firefly.adobe.com and guarantee we can work with three features, namely text to image, text effects and re-color vectors. We will see how to use them in the next videos. For now, let's scroll down to see more about the platform. Here we can see some example verbs of other creditors. And when I hover, we can see the prompt that is used to generate these works. Down here, there's a video that explains the firefly product. And moving forward here you can see the features that are in the development process right now. E.g. in painting is a tool to remove ads or replace objects in an image. You will be able to extend an image with the power of AI and change the aspect ratio. Or you can convert your 3D mock-ups into style limits. Firefly team is currently working on these features and I will update this course as they're released, these new tools. If you click on gathered here, you can explore artworks and getting inspired by the prompts. We can also click on a project to change the prompt or create different versions. Notice that there's a little info button on the image, and that is for Content Credentials. Adobe is being sensitive and respectful to the artist's work. And all of these images are approved by creditors to upload to the firefly gallery. You can also show your work here in the platform, but you also have the option to vote privately. Later on this course, you will see a video about firefly policies and rules. And there we will talk about all of the rules you need to know. Alright, this was an overview of the firefly platform. Let's get started with text to image in the next video. 4. Your First Text to Image: Now we will use text to image. Let me click here. Again, we see some projects from other creators and all of these are generated with textbook in which feature of firefly, let's write our own merged to create the first text to image project. I will say a futuristic Castle in the Sky. There you go. We have four results and I think this is the fastest image January term yet. You can expand the image and have a closer look. And not every image is perfect. There are some mistakes on some image, but this is still in beta and I believe it will become much better in the short term and it is really good for now. And one of the best things about Adobe Firefly is that there is no coast of refreshing the prompt and see new alternatives. So if I don't like these works, just say Refresh and brand new for image are already created for me. Now I want to talk about the right side menu, basically the properties of the project. First off, you can decide the aspect ratio. It's great to have all of these options. Landscape, portrait, square, and light screen. You can use portrait and square roots for mobile apps, social media like Instagram and widescreen mode for wallpapers or video thumbnails. If you want to go back, it says the history for the aspect ratio. So you can go back to portrait mode again without waiting. Or square mode, just like in the beginning. Okay, next we have the Content-Type or this project art is selected by default, but we can switch the photo if you aren't realistic outputs. All you can switch to graphic if you want to get imaged as computer generated. And most of the pupil use Unreal Engine phrase in the prompt to get perfectly emerged in other AI tools like MIT journey. In firefly, you don't have to specify in the prompt. You can select the graphic here. You can again switch and go back to our previous selection without waiting. Moving forward, the styles field has lots of options. If I switch to all, you will see the full amount. They are categorized. These are the popular ones. And let me try while explaining so you can see the difference. This is the digital art styles are, let's see Neil. By the way, you can select multiple options, but I will remove the digital art to see the real difference. Wow, this looks really cool. And let's continue with a couple more examples. E.g. if you are working on a game for the story illustrations, you might want to make an AB test between minimalistic and steampunk styles by a fly will be really helpful preparing any kind of store visuals in the future. You can use this castle in an advertisement and steampunk for another one. And this may even help you to figure out your target audience and style of the game. And this is not just for the games, it is valid for any industry. You have plenty of options for Teams as well. But I wanted to show you that you can also change the technique with just one click. We can convert them into pencil drawings. Okay, nice. Let's also include the doodle style. Here you go. These can be really helpful for the book illustrations and comics, e.g. using multiple techniques may also produce great results, but it is risky at the same time. Let me include multiple techniques and see the result. Not bad, right? Okay, let's just delete them and we can just click color style. Here are the effects category I want to use fishery. And these are the results. I mean, look at them. This is again, really powerful tool. Next we have materials options. This is not the best example prompt for materials, and we will talk about these features more detailed in the next videos. But let me just show you the metal style of this prompt. And we expect to see a metal castle. Alright, well here are the results. Firefly again did a great job. Lastly, in the concepts tab, you have several options such as chaotic, nostalgic, or simple. I will select simple to see how it affects this job. No big difference, but I think the details becomes much simpler. As I said earlier, you can use multiple style options here, but in general, I would recommend to use up to one option for each category. So if you select both Celtic and simple, it will not make sense, right? So you can select simple for our concept. Metal foreign materials, bio luminescent for effects, concept art for teams, and minimalism for moments. Let's click Generate and see the results with multiple styles. And the results are pretty good again. Well, to be honest, in my experience, increasing the styles sometimes provides better image. I also want to test if I add fisheye right now, with all these other filters, what's going to happen? I will just add fish high right now. And where was it in techniques? No effects yet here. So I will add and click Generate again. We can see some sparkles of fisheye style, but it is not as powerful as before. This is a good point to understand that increasing the amount of styles, we will decrease the power of each individual style. And just FYI, you can add up to 13 styles if you want. Of course I am going to clear these styles and let's continue with the next features. On here we see three more fields. First one is to determine the color of the image. I will select warm tone. And good thing here is that the image are not changed a lot, but just the color effect is applied. And this looks very butter. If you want the image to change, you can click on Refresh here. We can also adjust the light pink dye will just say golden hour for a nice sunset effect. Again, the rates are almost same. It just changed the lighting or the composition field at the bottom, we have options like Bruno background, wide-angle shot from below and much more. I must say here that these options are not working properly for all of the image. But keep in mind that this is still in beta, so it will definitely improve in the future. 5. Action Buttons: In this video, I want to talk about the action buttons. And to do that, I will go into our project. So we can click on this project, e.g. and it will create a new one for us. And we'll get the same prompt, but with different limits. So here you can see the outputs. What I mean by action buttons is that when you hover on an image, the buttons up here. So from here you can rate this slide are like the image. Gary report if you think there is a problem. But I want to talk about this bottom here. So basically it shows the similar versions of that image. So e.g. say, I like this one most, but it is not perfect. So I want to see more similar versions of it. So just click this button. Three more image similar to this one I created. This is a very powerful feature. But there's more. If you'd just come to here options. You can see that you can submit your work to firefly gallery so that it can be public and accessible for everyone. Just underneath, there is an option to use this image as reference. So when you click that, a new field will appear here, and now we are using the previous image as a reference. So basically you can use this image as a reference and give a new prompt. And with this slider, you can decide the weight of the prompt and the reference image. So as you go write, the prompt will get stronger and is a go-to left in which we will get stronger. Alright, next up I want to show you how to download an image. Though that Let's close this reference in which and I want to create a new project with a new prompt. I will say riding a motorcycle on the Mars. The results are really good. And let's say we want to download this one. You can do that by clicking this button here. And Adobe here again warns us with the credentials. So it says, Adobe is committed to promoting transparency around content generated with AI tools like Adobe firefly. So Adobe's purpose here is to let people know that it was generated with AI. Once we click Continue, you will see the credentials in the image. I will just say continue. Okay, So this is the image. You can see that Adobe included the credentials just, just at the bottom-left corner. Well, this is removable with software like Photoshop, but Adobe here says in beta version, this image are not for commercial use, so you cannot use it professionally sensitive still in the Beta version. Let's also see the resolution of the image. So by default it is 1024 by 1024 pixels. And this is for the square version. I will also select white screen. Okay, this time I will download this one. We can say continue. Iron here is the white screen enrich, and this is the square. And the white screen numerous choose elision is 17 92 by 1024, which is not bad. It is enough for most of the projects. And I think we will be able to use these amounts for commercial use as well after the official lunch. 6. Text Effects: Alright, now it is time to talk about the text effects. And this is a new feature unique for Adobe firefly, and it is definitely impressive. Let's see what it is. Once you click there, you can see again some examples using this text effect future because we see different effects. And the best part about this feature is that you can create new font, new text effects using any kind of materials. You can. Great texts from cookies with fair or with donuts, with wires, anything you want. So how this works, using this text field down here, you can write the text on the left side. So I will say Firefly for the first example. And on the right, you can describe the text effects you want degenerate. So here we can say anything like pizza, donuts or I will just say green plants. And now we write this is the result. And Firefly comes with four alternatives again. And you can solve it using these buttons down here. When you click one of them, you will wait for a few seconds and Firefly to generate the image. But after that, if you enable all of them, you can easily switch without waiting. Because again, fireflies safe and save this to the memory. Okay? Now these are the results and they are ready to be used in a row. Okay, So here we also have a sidebar menu. And we can see some sample effects that are running. You click and select one of them. You can see the prompt is changed automatically. So this is the prompt that will describe the effect to reselect pink Hawaiian hibiscus flowers and leaves realistic. This is the result. So you could also say wires. And this will be the outcome. Or you don't have to give exact materials like you can just say technology. This is not a solid material, but still Firefly will use the concept of technology. So this is, this is not go. I wonder what if we say space? Okay, not bad. We can see some planets on top of the characters. I also want to show you one of the most used effect, which is colorful fear. And there you go. Alright, moving forward you can see sample effects here. And if you say v of all, you will see options like metallic pasta or popcorn. I don't want to change my effect now. So let's go back. And the next field is the text effect fit. From here, you can decide the shape of the effect. If you select lose, some parts will go beyond the boundaries and the text will be more like 3D. If you choose tight, it will more respect to the font limits and the effect will become tighter. This property is medium by default, and actually I liked the 3D appearance. So let's go with that. Next. Firefly currently provides several fonts to use in the text effects. We can say we're all here and let me show you some of them. So this is the vertice and this is poplar. And this is the new psychotic. Notice the effect change for each of the new font. And remember that you have for alternatives for every one of them. Alright, let's continue with this one. So I will go back. And finally, we can change the color properties. Just at the bottom we have the color field. And from there you can change the background color and the text color, background color. You can make it anything you want like this. But if you just select None of them, and if you select this one, you can download it without a background. So it's up to you for the most projects without background is better because you can use it anywhere you want. So let's keep it that way. And I want to explain tax color. So right now, if I select black, e.g. for the text color, what do you think will happen? So the effect is still colorful, right? The fear is not completely black. And that's because we specified in the prompt as colorful. But also be aware that this effect becomes darker. So let me switch to white first to see the difference. Okay, So this is the white text color and this is the black text color. It seems like it is affecting the light. So there is a difference, but the color is not completely changed. And that's because we just specified here colorful for if I delete this Verge and select black and generate again, you can see the fur becomes much darker and it is not colorful. It's not completely black as well. But let's select white. So now it can affect more. So this depends on the material as well. Let's say. So if you want to create a text effect using e.g. Sushi, let me select sushi. Sushi effect looks like this. So you cannot make the text of this effect completely dark, even if you select black. So it will become a bit darker, but the text will not be completely black. So all I want to say here is there should be a correlation between the text color and the materials associated cannot be all black. And Firefly will not apply this. 7. Recolor Vectors: Welcome back. In this video, we'll talk about how to recolor vectors using Adobe firefly. Fireflies or your color vector feature is a powerful tool that can apply unique colors to your erector enrich using everyday language. Let's dive in and explore how to use it. So as always, we are in firefly.adobe.com and select the re-color Vectors option here. And here you can upload your SVG file or choose a sample, a set from the website. I'm going to continue with this one. Remember the SVG file to upload will not be accessible to anyone else and remains in your ownership. We continue with the sample SVG files. And once you have uploaded to your SVG, describe your desired color palette in everyday language here. For instance, you can say sunset, that the beach. Or you can also say cool I scholars. And as you can see, these four images are generated with the description and sky is the limit here. This feature allows you to apply extraordinary color teams to your vector image instantly. Then after the initial re-coloring, you can try variations by shuffling the colors here, like this. And you can see the used colors on the left side. So this will save time by not having to re-color each element individually and easily experiment with various color options and teams for your artwork. Alright, now it's time to explore the right side menu that are nine sample prompts to choose from. By selecting these, you can apply different colors styles to our uploaded image. The available options include salmon, sushi, e.g. the swamp, and also dark blue midnight, like this, one or two p, This collect. Under the sample prompts, there's a drop-down menu. Well, color harmony. This option allows you to choose the type of color harmony. Your results offer options include default, complimentary, analogous, try it, split, complementary and square. Let me explain these options and their meaning. In color theory, different color combinations are associated with specific harmonies or relationships. The options provided in Adobe Firefly for the harmony drop-down are based on these principles. So let's go through them. Default means that there is no enforced in specific color harmony. Unlikely uses a random colors that generally work together. Complimentary colors are those that are directly opposite each other on the color wheel. E.g. red and green are complimentary colors. Using complimentary colors can create strong contrast in your image. Analogous colors are those that are next to each other on the color wheel. E.g. the yellow and green are analogous colors. Using analogous colors can give your image are harmonious and balanced field I tried to color scheme uses colors that are evenly spaced around the color wheel. E.g. if you were to draw an equilateral triangle started from red on the color wheel. The other two options will land on blue and yellow. These three colors make a triadic scheme. A split complementary color scheme is a variation of the complimentary color scheme. This provides high contrast without the strong function of the complimentary scheme. And finally, square is similar to the rectangular, but with all four colors paste even there aren't the color circle. This scheme is also called the double complementary because it uses two pairs of complementary colors. If you want one color to be dominant, the square color is a good option for you. Selecting these harmonies in Adobe Firefly will adjust the color palette of your erector image to align with the selected color terrorists. And it will provide a different look for your image. Below the harm and new menu, there's a color options. And currently Firefly color options are limited to 18 colors. You can select multiple colors here. And by selecting a number of these colors, you can put a stronger color focus on your results. Lastly, we have a switch button down here. This option allows you to preserve black and white in your illustration. That is enabled by default. And that means any black and white elements stay indoors, origin colors. However you can toggle this off. You will get more variants results. Alright, once you are happy with the color team, you can download your recolored image in SVG format. You can then import this into your design graphics program or platform for further use. Besides, if you want to use it instantly in Illustrator e.g. you can just say copy to clipboard and paste it in any program you want. And that's it with Adobe's firefly recollect vectors picture, we can easily transform your vector image with new color palettes in a matter of moments. 8. Generative Fill: January to fill feature is launched and I will explain how it works in this video. Let's dive into it. Here you can again upload your image or choose one of the examples. Let's check two examples. First, when you hover, you can see the capabilities of this feature. Notice how it adds waves, boats, or even change the whole nature of the image. Just underneath, it changed the background, even it is blurry and now she's in the forest. You can also change the clouds. And I think it is incredible. Again, here you can remove any object from image and it will generate that area using AI. Alright, without further to talk, let's try out this feature. I will continue with this image. Open this up. All right, On the left we have insert and remove options. And down here we have a toolbar. You can also select Pen and move the image. But you can do that also with a mouse pad or the scroll wheel of the mouse, and it is easier. Okay, Let's try insert. It gives you a brush that you can draw on the image. And you can also adjust the settings of the brush like size or opacity. I wanted to try to add a dark here. So I will just draw this area and you can see that a TextField appears for the prompt, I will say cute dog and click Generate. Now we wait. It takes a while but not too long. And finally, firefly, it comes with four options. And look at them. These are huge docs and let's have a closer look. Okay, not bad. But if you don't like any of them, you can always ask for more image. There is no limit of this guarantee. So you can ask for more, for more. And formal. And Firefly will continue generating difference. And there you go. Notice here that we didn't lose the previous ones. So you can always go back and select an older option. Alright, this one is great. It is still huge, like the dogs in Harry Potter. But I like how it makes this image mystical. So I will keep this one. Alright, Next I will try to add a superhero at the roof of the building. So let me just draw here like this and type superhero, click Generate. Oh, by the way here we also have additional settings. You can switch this one if you want to match the image exactly to your selection shape. Or if you want to keep the original image, you can set the strength here. And finally, if you want to make prompt stronger, you can move the slide to the right. By default, Firefly will rely on the image as well. And I will make the prompt a little stronger. And finally click Generate, Firefly now generated for superheroes. And I don't know where is the head of this one. We have a headless hero, and this one is not bad actually, but I think these are too much in cartoon style. I click More button here, and there you go. We have other options. Well honestly, I expect better heroes, but this one is not bad. So let's keep this one by clicking the button here. Now I want to show you the remove feature. This is again a great tool. You can remove anything in the image in just seconds. So let's remove this hero to show you how it works. We have another brush for removing this time. I will just cover this hill and it is no problem if you include the surroundings. By the way, if you want to be precise here you can't use substract option and delete some parts of selection in order to work more detailed. But Firefly will generate alternatives and ways. And mostly it handles perfectly about the surroundings. So let's click Remove. As you can see, it removes the object and fields the background. Notice the Cloud changed in every option. And I think this one is the best. So I say keep again. For the next example, I want to remove this mountain here. Let me draw and cover all of it. Just like this. And click Remove. Since we didn't provide any prompt, Firefly generates different suggestions based on the surrounding areas. All of them are mountains again, but maybe I want to see an ocean here. Let's see what we can do. To do that. We should use insert tool instead of remote. When I switched to insert, a good thing is that it keeps the selection here in the prompt. I will say ocean this time. And let's see. Okay, Look at them. These are great, but maybe we should be more careful during the selection because this part becomes blurry. So let's see if it fixes in other options. Yes, it did. This is perfect. Let me check the last one and this is even better. I think the last one is more realistic. So let's keep this one. Alright, that's enough for this image. Let's go back and I want to use this image to show background examples. Removing background is not a new thing and we're all familiar with it. But we can also replace the background with something new with Adobe Firefly menu click background here, it removes instantly. And down here we have a TextField again. And now I will try to change the background to Disneyland. And boom, not bad at all. Notice that Firefly understand the type of the image. And I mean this person is real, so the pupil on the background and everything is also realistic as well. Let's try something else, maybe New York. Okay, these are the results and this is insane. I don't know why this bus is so fast, but anyways, these are really impressive. Okay, next up, let's try to change the clouds. Like in the example in Star Tool is selected. I will just draw the body part just like this. That's it. Now I will say green t-shirt and click Generate. And here we go. It really works. There are some mistakes, weird parts, but in general it works like a charm. I will also try to change the hair. So let me select all of the parts. This time I want to say blonde hair. Let's see what happens. Okay, this is really, really good. And I know that there are some little mistakes, e.g. on these pixels, hair and t-shirts are blended together and it looks fake. But remember maybe we can fix it using the Move tool. Let's try. I will select, Remove, and draw this part interiorly fixed. So the opportunities are limitless here, you can fix an array you don't like. Alright, in general, I believe these features are going to save a lot of time. And I know that all of these can be done before firefly as well, but it used to take hours to make these edits. I watched the generative fuel official launch deck and they are saying that Firefly is not able to do something that was not possible before. But it does everything Tenex faster. Plus you don't need any skills to use it. So anyone can do these adjustments with a few clicks. Alright, that's enough for confluence. Let's see one thing that firefly can not do at the moment. So I tried to change the background as if this person is at the top of the tower. I'm not sure if this is a good approach, but I typed at the top of the Eiffel Tower in the prompt. And the results are not that satisfying. It did include the Eiffel Tower, but apparently the person is not at the top. Value might think that this is not a fair case. So I also tried to say at the top of the Empire State Building, and this time results are a bit better. But again, I would expect to see some blocks or parts of the building behind the character. And they're not seems to be real. Alright folks, that was degenerative field feature of Adobe firefly. You can also use this feature with Photoshop, and I will explain that in another video. 9. Prompt Engineering Tips & Tricks: Hello everyone. In this video we are diving into the exciting world of prompt engineering. It's the secret salts for getting the most out of your AIR generators like Adobe firefly. I'm attorney. So let's get started. First off, let's clarify what prompt engineering gifts in the context of AIR generators, it refers to the crafting of effective text descriptions are prompts that guide the AI in generating the specific image you desire. Proud engineering is crucial because it allows you to control the output of the AI and get closer to the visual results you are aiming for the battery or prompts, the more accurate and satisfying your generated image will be. There is already a job title called prompt engineering in big companies. So let's see some tips and tricks for your prompts. Tip number one, Be specific. Instead of writing a cat, you might say a fluffy orange tabby cat sitting on a window sill at sunset. The more details you provide, the better the AI can visualize what you are looking for. My second tip is to try different phrasings. The first prompt doesn't give you the result you want. Try expressing the same idea in a different way. Use different keywords. For instance, instead of a medieval castle at night, you could try a stone fortress under the moonlight. Tip number three. Experiments with abstract concepts. Ai generators can handle more than just literal descriptions. You could try something like the feeling of nostalgia or the concept of infinity. Far tip. Use adjectives and adverbs wisely. Words like gigantic, sparkling peacefully can significantly influence the AIs output. Final tip, you might dollar they asked me for, how can I remember all of these adverbs and fancy words before creating something while we are living in 2023. So there is an AI tool for that. You can use a language model like check GPT to create prompts. Let's see how we do that. I will go to the chat GPT and say, write me a generative art prompt for futuristic photography. Whoops, that's not what we expected. Actually. We're not prompting firefly in this way. Right? Okay, So Chet GPT apparently doesn't know how Adobe Firefly expect the prompts, but we can train it. We can provide the formula of a prompt to check GPT so that it knows how it should, right? Let me show you one of my chats which had GPT. I first define the formula for this formula subject of the photo, three descriptive keywords of the image. Camera type, camera last type, time of today, style of the photograph. It understands the structure and already gave me a prompt, which is a Golden Gate Bridge. And these are the descriptive words. The camera type and also the lungs type. Time of the day is sunset in this example and style of the photograph is long exposure. So this is perfect. Let's try this with Adobe firefly. I will just copy and paste the prompt here and wait a bit. And there you go. This is perfect, but there is more than I say, the name of this formula is AF stands for Adobe Firefly formula, lung and type. Yes. If you understand, it, says that understands the formula and I will provide a new one with domain a F2. This time we can include adjective, subject, doing an action, lighting. And you can also add some static words like detailed and realistic if you want, include in all of the prompts. And you can say in the style of a movie or artist at the end of the prompt. Now it says, I understand the second formula as well. And now you can say, give me a lion image using a IF2 and look at this prompt. An image of fears lion roaring under the dramatic twilight, detailed, realistic, just like we said, in the style of the Lion King. And I liked that part. And Salvador Dali, I am not sure Firefly is artist ever at the moment. Artists styles works better with MIT journey, but we can try, Let's try with Adobe firefly. And these are the imagery yet. By the way, you don't need to ask one prompt at a time. E.g. you can say, write me five prompts for Lionel new choosing a font. Okay, so this is the concept of defining formulas and using language models to create your prompt. So I strongly recommend you to define your own formulas according to your needs. For instance, you can define one for realistic photographs and valid for digital artworks. And that's totally up to you. All right guys, that was the tips and tricks video. I hope it was helpful and valuable. 10. FlowGPT - Firefly Prompt Generator: Okay, Here is a great platform for finding and using grid prompt. So we are talking about flow of GPT. There are many other platforms like this one. So let's cover this platform for an example so that you can understand how you can use these websites. We are in the homepage of flow of GPT here that it says find and use the best prompts. So basically this website is for finding prompts and sharing with the community. In the homepage, we can see already some prompts of other users. And on the left side we see some tabs. Alright, so for example, here is the Learn tab and there is a documentation of flow GPT. If you are interested, you can read and learn more. There are also some hackathons and talents for prompt generators so that you can, for example, currently all of these are expired. So there was a challenge, a hackathon for meat journey, and I guess there is a money award as well. So this is a rapidly growing area at this moment. So there are many chances and opportunities like this one. So in this hackathon, it says, this team is concentrated on tasks that effectively utilize the power of generative AI to create visually compelling prompts in MIT journey. So I guess the best prompt wins the $600 Bhante. There was 18 entries and these are the entries. I believe you can go into that entry and just recovered in the prompt engineering video, there is a chat GPT prompt. So this guy, Oscar chat GPT, described yet GPD how much journey works and ask it to create prompt. And that was the entry for this guy in this hackathon. So let's go back to home and see if there's any prompt for Adobe Firefly. There is one. Of course I knew it. Credits to mark Mason. This guy created a prompt to check GPT in order to get Adobe Firefly prompts. So let's see how this works. By the way, it is shared on April 12th and he says Adobe Firefly prompt generator. It will help you generate Adobe Firefly prompt that are also comments. And you can also join the Discord server if you are interested. So he says You will say is to check GPT. Actually, you will now act as a prompt generator for generative AI called Adobe firefly. Just like we do it, provided a formula. So the prompt structure is 123456 and the structure of the prompt will be in. So he just explains the formula. One is for keyword to is for detailed description of the first keyword that will include very specific image details. And the third one is the detailed description, describing the environment. Alright? So describe the fourth one is for mood and feelings and atmosphere of the scene. First one is the style photography, illustration, sculpture, artwork, paperwork, and more. Hence, sixth one is how five will be realized. This can include the camera model, painting with detailed descriptions. So this guy did a great job and creating this formula. And it is here, I believe, in order to share with the community so you can, you are free to use it, alright? And you will not write the word description or use anything. You will write each prompt in one line without using returned. So these are for syntax of the output that chat JP t will give us. And there is also an important point to note section. While using, while writing prompts, never use numbers. So don't use any brackets, just give us the plain text in English. I need those will say is please pay attention. Concepts that can't be real, would not be described as real or realistic. This is a bonus point. So he says choose the correct camera. It is very important. Finally, you will always provide prompt in Markdown code cell and only provide the prompt. And then he says, Are you ready? Okay, Let's see Moore Show example. So we see the chat here. After saying, Are you ready, check GPT, just Ashworth. I am ready. Please provide the first keyboard. And then he said space. And here is the three prompt examples. So first one is spacecraft. This is the number one. I guess. I'm hovering over a red planet with two moves visible in the sky. Second example is aloneness turnout standing on the surface of a desolate planet, looking up at a nebula fueled night sky. Alright, and this is another example. A girl. And notice here, touching beauty creates three girls in different actions. So first one is bothering, the second one is young girl and portrait of a teenage girl wearing headphones and leaning against the colourful graph to convert ball. Alright? So in this way, you can basically create your own prompt generator using a language model that had lots of examples on the web like this one. And I believe this is a great one for starting out. And again, all credits goes to mark Mason here for creating this prompt I am sharing with us 11. Text Effect Prompt Generator: Alright guys, In this video we will take one more step in the prompt engineering journey and train that GPT with new guidelines or the text effects feature of firefly. I believe this video will be a good practice for you because we will create a trainer prompt from scratch. And you will understand this process by following. I am in chat GBT and I will start typing. So first of all, let's explain what Firefly is and what is the text effect feature. I'm going to say Adobe firefly as a text effect. Richer. And notice here I just use double quotes here in order to specify it is the name of the feature, right? Okay, I'll, there'll be Firefly has a text effect feature in order to create text effects. Which user prompt. And then we'll say please act as a prompt generators or text effects. Adobe Firefly. Okay? And while doing that, please follow these guidelines. All right, now let's, alright, now let's go down here and we will write our guidelines. First one is how many prompts during eat? I can go with five. So you can adjust this in your needs. So I will just say Renee requests response with pipe, text effect, prompt and write. Nothing else. I'm writing this because sometimes that GPT response with explanations, descriptions, and all these dumb stuff. So we are just saying, right, nothing else. Okay, so the second guy is prompt. Should explain the style of the effect, right? So this is the main idea here. And I will just describe more detailed appearance of the characters. Okay, So this is on character level, because remember sometimes some character can look like a robot. So let's just let check GPD and all this and also any relevant theme and concepts. Okay, Next, I'll just warn GTP tunnel because it doesn't know at the moment that Firefly is not every row of copyrighted content. I'm going to say Firefly is not a radar of copyright, characters, products, and brands. So don't includes words like Batman, iPhone, or YouTube. Okay, great. Next up, and actually we are done here. And I will just tell one more team in order to adjust the prompts. We can say please always provides prompt in Markdown code cell and only provides prompt. We are writing this again. So you can see this sentence in all of the prompt engineering books. Most of the people use the sentence. It's a popular one. And that's because we can get the response in Markdown code, which makes things easier and easier to see, easier to copy. Alright, so then actually this might be enough, but I want to be sure that GPT understands me correctly. So I also provide some examples. Okay? I will provide five examples. And I don't want to waste your time by typing these as well, because these are the things you can find on the web. And also you can use mine. I will just leave the text in order to copy it, paste it. So I just paste these texts here. So examples are like molten lava texts, bubbling, appearance, glossy texture, volcanic team captivating designed. And in the second one we have a nice concept. In the third one we have neon concepts, doodle, and finally, I, you know, I like working with abstract concepts. So serial texts, dreamlike distortion, and so on. So that's it. Now GPT also has examples and all of the guidelines. And at the end, we will just say, are you ready? Let's go. Okay, It says, absolutely, I am ready to generate texts, effects, prompts for Adobe Firefly. Please provide your request. Okay, I'm going to try request a prompt with only one merge. And let's see if it gets it. I will just say fire All right, It is working. So now we have already four of them, and this is the first one. Here we go. We have five incredible prompts for Adobe Firefly texts effect feature. And let's test one of them. Okay, so let's try the first one. So I will go to the Adobe Firefly Text Effects tab. We can again use Firefly text here. And in the description field, I will just paste the prompt that chat GPT galas. Alright, I am excited. That was insane guys. How we did not think and look at these words. So I think this is the correct way of using these AI tools. So no one can remember all of these words. Nolan can imagine that many combination rate. So this is perfect. You can also try more effects for fire. But let's see something different because in the examples we have provided, we have provided a similar effect. So maybe it is biased, right? I will try something completely different. I will say snake. Okay, Let's not wait for it and copy the first one where it says Select during scale text, repetition, alien character, finished, nature and wildlife team. I like this one and exerting an inverse memorizing. All right, so this is awesome. Look at this F character and all of the textures of snakes, even the eyes of the snakes. This is Irene. Really, really good. Let's see this. I'll turn to you and this is again, perfect. It's a joyful snake in this time. And let's continue. I will try to, I will try to get, I don't know, maybe neon will be cool. So let's say neon. As always, I will just take the first one. Copy and paste here. Let's wait for that effect. This time we have late night city team. So Diigo, this is again, really, really good, right? Alright, for the final example, I want to test jet GBT with limited vert. So I will just say YouTube. Okay, remember we just limit chat GPT. We said fireflies not aware of copyrighted characters. So avoid words like iPhone, YouTube. Batman shouldn't include any brand name in your prompt. So I wonder what it will do when I say YouTube. Let see. Alright, You genius. It says I am sorry about the prompt should not include copyrights, characters, products, and brands like YouTube. Could you please provide another keyword? So it's shouldn't include the vert YouTube, but you can use the genre and the YouTube ecosystem, right? Create prompts. In YouTube ecosystem and genre. Read out using the var. Alright, now we have the results and they are actually good. We do description tags, detailed and informative behind the scenes team. Okay, here we have video thumbnail takes optimized for small-scale item on the reason why this is here. First impression team, this is weird. Okay, Let's see one more video title, text, bolt and attention grabbing. This one is good. Content creator team, professional, and stylish. Alright, these are not bad. Actually, I want to explain one more time so I can say, you can create prompts around the concept of online video sharing, streaming, or content creation without directly mentioning the brands. Give me one thing. Livestream, compound texts, digital timer, aesthetics, real-time interaction team, immersive and some suspenseful. Alright, so actually I don t know what I'm expecting. So I like this one most WE DO title text. I will just try this one. Diigo. I don't know if this is related to video creation. And let me try this one as well. All right, not bad, right, Nice. Try to TP T. So this is something related to Adobe firefly. So because it is, so let me just type YouTube so you can see it is not aware of the brands. So the problem is not about our prompt. For the final example. I want to try, um, something like something abstract. So let's say existence. Now it gives one prompt because we just asked or one in the last request, which is fine. It's simple. It's easier to copy. I just copy this one. So let's see our existence prompt. Quantum real texts, distorted and fluctuating characters to theoretical physics team. Mind-bending and enigmatic. Hello, right, let's see. And I go, this is our existence text effect. Alright guys, that's how we can use and train Chet GPT for your text effect feature in Adobe firefly and use these prompts in texts effects. By the way, you can also use them in adobe expressed inside of a project. So my recommendation is always heavier. Check GPT on one side. Just copy this prompt here and create a chart with this prompt. And then you will be able to ask a text effect prolonged whenever you want, which will make things much easier. 12. Photoshop Generative Fill: Hey everybody, you can now use generative fill feature of firefly with Photoshop. Right now it is available in the Beta version of Photoshop, but it will be integrated to normal version in the future. So if it is already available on your Photoshop when you watch this video, you don't have to install the beta. Otherwise, let's see how you can install Photoshop beta. In order to install it, you can go to beta apps tab on the left, and find Photoshop in this list, you should click here the Install button. And since I already have it, I can open from here. Alright, now I have a dark image in my Photoshop. And first I want to show you the Replace feature by replacing this talk with another animal. To do that, I will choose lasso tool from the left and drove the surroundings of the dark. I am leaving some gaps on purpose so that Firefly well aware of the environment. Once you select any area on the image, this box will appear and there you can see a degenerative fill option. Click on that, and then new text field appears. This is the place where you will type our prompt. In this case, I just want to say a cat, click Enter. And there you go. Firefly generated three variations of the requested prompt. And you can see the variations on the right. Just like in Firefly app, we can request more variations as we want. I guess this selected area is not perfect for cats, but let's see more variations. Alright, now we have three more variations. Again, we didn't lose the previous generations. I should say that some kids are weird, but this one is not bad. But I want to show you a trick. What we just did during the selection is to draw the borders of the dog, right? So Firefly tried to generate a cat inside of that selection. I guess that's why it's struggles to find a good pose for the cat. And we got some weird pauses. Let me delete that layer so we can go back to initial phase. And what we can do here is that we can also select the area just like this, and that will give Firefly more space and prompt the cat again. Okay, now we have the results that produced bigger cuts, but at least they are not in weird pauses anymore. But we can also try is to change the prompt by saying running cat. Um, let's see the outcomes. Okay, this one looks great except the ice. And this one is acute one, but this leg doesn't look real. And this is the final one. Again. Great. But I guess fireflies struggles with the cat eyes at the moment, right? And you know what? I will generate more. I want to find the proper cat for this image. And wanting to notice here is that since the image is realistic and not a digital art, it is always harder to generate a live tanks, which looks real. Okay after some generations, I really like this one. And only problem is again with the ice, actually the left eye. So let's try to fix it using generative field feature. Again, I will select this region and click Generate without any prompt. While normally during empty prompt will remove the object. But this case is exceptional. So I really wonder if it will remove the eye or will create new variations. Let's see. And as always, we have three options. First one removes the eye completely, and this is great if you want a pirate cat. Other radiations created ice, but the size and the color is not matching with the right height. So this was not a good idea. Maybe I also want to try one more thing here. I will select both of the eyes this time and click Generate. And there you go. If you can generate ice, we can just close them. The second variation is also good, I believe. And this is the tort one. It is so aggressive. But I hope what I'm trying to show you guys is clear. Firstly, Firefly is not perfect yet. I don't want to show you only the happy paths. Sometimes you should spend more time and generate more variations in order to get results you want. Good thing here is that there is no limit or cost of multiple generations. Besides, these tools improve Schwab fast, which means it will work way better after a couple of months. Alright, next thing we can do with that tool is to extend the limits. I have a great example here, because the front of the jet ski is cropped and this image is landscape right now. And let's see if we can convert it. Okay, say I want to post it to Instagram, so I want to change the aspect ratio as well. I will just select Crop tool, extend this while holding Option key. And you can also give specific width and height values. It's totally up to you. Alright, now collect the image thumbnail here by pressing Command Control button that will select our image, but we want to select the empty regions. So we can just click Invert here. Then you can click Generate, Fill. And here you can again type some explanations, but you can also leave it blank, which will let Firefly decide an extent the image. So let's go. Honestly, I think this is one of the best features of firefly. I mean, look at these variations. I like this one most scenes it completes the front of the jet ski. And no one even understands that this is not a real image. Next up, I want to talk about the common mistakes I see on the Internet and YouTube. As you might already know, photoshop has an Object Selection Tool. It works really, really good. E.g. you can select this book on the desk and the same box appears. Now let's try to replace this book with a coffee. So I will click, Generate, Fill, and type coffee here. Let's wait for the variations. Alright, now, these are not what we expect to attract. And I feel that you have an idea about why this happened. Yes, since we selected the image exactly without any gaps or surrounding pixels, it failed. Besides, the selected area wasn't convenient for a coffee. Replacing objects in similar shapes is not a bad idea. But if you want to replace object with a different size object, like in this example, a better approach would be delete the object first and then insert an IV object. Okay, So first you can use Lasso tool and draw the surrounding of the book with some gaps. Generate without any prompt which will delete the book from the image. These are the variations of the image without the book. And we can continue with this one. So select lasso tool again, there are other region that you want to create the coffee. So we can type coffee here and wait for awhile. Diigo. The reason that they look deleted is the layer order. So in layer one we created the empty space, and in layer two we created the coffee. If you move layer two on top, we will be able to see the full coffee. And I think this is great. Using objects selection tool is not always a bad idea and there's a way to do that properly for Firefly generative fill feature. Let me show you that. Say that we want to remove this guy on the back. Let's see the wrong way first, I will select the guy with the Object Selection, hit generative fill, and generate again. This created new people for alternatives. And this is again happened because we specifically selected the guys borders and the area that the person fits. But you can do instead, select the guy again. And before clicking generative fill, click this button. There's an option to expand the selected area. So I'll click Expand selection. And now we can expand as many pixels as we want. I will try to add two pixels for this example. Click Okay. And as you can see now, we have margin between the selection and the person cannot generate with this selection. There you go. Notice how Firefly understands the objects behind the guy. Even they are blurry and creates Blu-ray alternatives. All three variations are great, but I think this one is perfect. Alright, that was the generative field feature for Adobe Photoshop. I hope these tips will be helpful for you 13. Photoshop Remove Tool: We talked about how to remove objects using generative fill. But in Photoshop there's a new remote tool which is also powered by AI. And I want to show you that as well because it comes with a brush and make it easier to remove tanks. Alright, you can find your move tool on the left side menu. You can just right-click here on this one and select Remove tool. This will give you a brush and you can draw over the objects. The things you want to remove. For instance, this bunch looks like dark tea and we can clean it up using this tool. I will just draw this white region just like this. And you don't need to do anything else, it will disappear instantly. Normally, I speed up the video in order not to waste your time, but this tool is so fast, I don't even need to make video faster. Okay, let's also remove this course here, and there you go, it becomes much better. Maybe we can remove these details too so that it looks cleaner. Okay, Just like this. Nice. Also on the left side there are things like stickers which looks ugly. We can just draw over them like this and also this one. Alright, ty can say that this tool works perfectly on things like this, scars, scratches, and so on. One more thing I want to show you is that you can also make skin smooth. Are we using this tool? So let me get closer to the face. I just click on the points, the pimples, Agnes, and they disappeared instantly. So this is working well on the screen as well. Okay, Great. What about Jimmy when an object, for instance, let's try to remove this tree on the back and we can adjust the brush size. Up here. I will make it bigger for the tree, maybe a bigger, a little more. And this is great. Now let's draw. I need to be sure that I include all of the parts of the tree and be careful on the details. You don't want to remove the arm, so you shouldn't include the surrounding objects and it goes away. I'm not sure that this result is good enough. Notice how it becomes blurry and some areas became beard. Maybe this was my fault. I could draw more accurately to fix it. We can continue removing these areas. In general, I think degenerative feature works better in cases like this. Notice that the plant on the back is not generated. It just remove the tree and there is a gap now because it didn't generate new image on these pixels. This thing also looks weird. So as I said, I'm not sure this Remove tool worked good on this case. So let's try one more tink. Let's try to remove this swatch. I will make the brush size smaller and draw on all of the boundaries of the watch. Like this. And again, need to move the voyage but left weird texture on the wrist. Good thing here is that we can keep removing. So I will just draw the areas I don't like. And in that way we can make it look better. I will just continue removing the pixels until it looks good. And after some mutations, I made it to look like real. And this process was really fast. Alright folks, that was the new removed tool. I think that does a great job for most of the cases, maybe not on the tree. By the way, we could also continue using the Move tool to edit these parts. But in my opinion, generator field might be a better choice here. Sorry, Move tool is treatable where the background is clear and doesn't require to generate new assets while removing 14. Practice Blend: In this video, we will create a really cool image using blending technique. Let me explain what we will do beforehand. So for the ones who wants a challenge, you can do the project before watching the video. We will create a portrait cameras in Photoshop like this one in the screen, generate two inwards with Firefly, one for the top and one for the bottom. These limits will be surreal and digital art. And there will be a gap between these two image that is an empty region for AI to fill. Then we will blend these image using generative fill in Photoshop. I'm not going to show you the results now. I don't want to affect your imagination. Okay, So I will start doing my project. But firstly, I want to say that January to fill in Photoshop is not the same thing with text to image in Adobe firefly, you shouldn't use generative field for creating the standalone image from nothing. Let me show you this example. I will select the bottom of this canvas. Click generative fill and prompt underwater city in the ocean with elephants. As you can see, the results are not satisfying. This is the second one. And third one is not bad. Actually. This is interesting, but definitely these are not what we expected, right? So let's go to Adobe Firefly. When we type the same prompt to Adobe Firefly, these are the results. Besides in this platform, you can adjust many settings, and I'm going to use this one. So let me download it. Okay, Let's go back to Photoshop and let me place it to the bottom of the Canvas. Alright, now we need one more image for the top. And I want a futuristic city with buildings. And after the comma, I will say golden hour. Notice that Firefly automatically remove that phrase and add it as a style. Okay, great. Actually I didn't like that. The buildings are not completed and there is no sky. But we can fix it using generative phil. Alright, download this one. Important in Photoshop and position here like this. Maybe we carry them all the copyright from this image. We already had one at the bottom. I will select that area and use generative fill for this one as well. You can use it for anything. And guys don't use this for read purposes. I believe in you. Okay, perfect. Now I will select the empty region, but also including some parts of both images. And click Generate again. Wow, this is incredible. I mean, it looks already saw cool. So let's see some more alternatives. And I liked this one most, but there they are, rigid transition on this line. So I will select that area and generate one more time. Okay, great. Melt bad. But this one is perfect. Oh my god, How the heck it is that good? I am still getting charmed with these features, guys. Last thing I want to do is to complete these buildings and create a sky at the top. With the crop tool, we can extend the image and use generative field again in order to fight or flight complete discard. These are the alternatives. All of them are good, but as I said, I want a clear sky and complete buildings. So I will use this one. And with that, my blending in which project is this? And this is the final version. I hope you liked 15. Hands on Practice: Changing Clothes: Alright guys, we've seen some examples for replacement, but in this video, let's talk about changing the clothes and variables. Specifically. First of all, I've watched many videos on the web around this topic. And one thing I noticed is that no one is talking about the failures. In my opinion, change in quotes is still not perfect and can be better. In this video, we will talk about not only the best scenarios, but also the difficulties and things you cannot do. Alright, let's start simple. So the easiest thing here is to change this t-shirt, for example, with another shirt. To do this, we can use Lasso tool to select the t-shirt and drove around the chart. Just like this. And unclick January to fill. And I will just say tennis shirt. Okay, Let's wait for the results. And there you go. Those are the results. This is the second one. And this is the third one. I think Firefly messed up with this one. So I will click Generate one more time to see more results. Okay? This one is not bad. This one is really good. This is the most realistic one, in my opinion. Let's see the final one. Okay, again, Firefly messed up. Alright, at least we have one good result. So thumbs up for this one. In the previous videos, we have seen one example for removing the watch. And we could do that with the erase tool. But let's see if we can add an Apple Watch to disguise wrist. In order to do that, I will just select the wrist and type Apple Watch. Alright, they are here. And actually Firefly created some watch, but I think the rotation is not good. So all of the three options are in the wrong way. So I will generate again. For the other results, nothing changed. Okay, so this is the first point I want to warn you that fight or flight cannot create watch properly. Maybe the issue is about the selection. So I want to select larger area with the rectangle selection. So I will just select like this. Maybe it needs more space. Let's get one more chance. So again, I will say Apple Watch. Okay? And once again, the results are not satisfying. So these are too big, for example, That's because we just increase the area of the selection. But still they are they're not looking realistic in my opinion. And also, it is clear that Firefly understands Apple watch because these are like Apple Watch. But the position and rotation is the problem here. And once again, please notice that ai industry is improving really fast. So by the time we are watching this video, it is possible to firefly become a great watch creator. Okay, So next up, I want to show you how to replace short, for example. In order to do that, we can just draw this area. And while doing that, please be careful. You don't include the hand here. As I already said many times in the previous videos, AIR generators struggles for the hand generation. So we don't want to ruin the hand over the dark. And that's why I'm just drawing like this. And now I want to say, great tennis short. Alright, these are the results. This one messed up again, but they are good shorts as well. Okay. Now, let's try blue tennis short, okay? Gray is, gray is not so specific, so let's try blue tends short and these are again black. And maybe this happened because I just abort the process while Firefly generating something. So there was a bug, I believe. Let me try RET. Okay, let's start over. Let's try red tennis short and click Generate. Okay, so at least we have some red shirts. Alright, I will just continue with this one. About the difficulty here is that you can not decide on the style, the pattern, or the complete color of the clots. So this is another difficulty example. Let's continue. What if we want to remove this short and insert pants will set? So I will just try sports parents and click Generate. What do you think will happen? Remember the selection area? The selected area is for church. We didn't include the legs completely. And during this election, it's a question mark if it can generate Pads. Alright, there you go. These are again shorts. So we can not have parents because we didn't select that region. So Firefly all in January, it's new content on the selected areas. So if you want to create pants, you should make a new selection. So let's do that. And this time I will also include the legs, the bottom parts, all the way down to the socks and shoes. Just like this. And again, I will exclude the hand from the selection. And now I will say sports parents. Right? Let's see. And there you go. Alright, this looks weird because the nearest sides of the pan is white, I believe, and the background is white. So this is not looking good. Let's see other options. This is better. This one also better. I'm going to go with this one. So the key point here is that you should select all of the area that you want the cloud to be generated. Another example we'll be creating a sweatshirt for this guy. So if you want to create a sweatshirt, selecting only the t-shirt is not enough. So you should start drawing from the wrist, right? So just like this, I just drove selection and click generative field. I will just type sweatshirt for the prompt. Let's wait for awhile. And here is the sweatshirt. I think this looks great. This is the second option. And third one is again, a t-shirt. I don't know why this happened, but I liked the first one. It is oversized, which is perfect. I'm paid the way the pants looks ugly, very ugly. We completely ruined this guy style. So let's try to fix that. And I will select this layer again. Why we selected all of the legs for creating pants. And let's try cargo patents. Basically not this isn't basically not working. I don't know why. Maybe if I fly it doesn't know what cargo means. So I will just try genes. This guy continuously gets thinner and thinner and desired. The other options, I don't know why, but somehow firefly just crops some parts of the legs. Mostly on the leg, on the left leg because all let's create more. Alright, I've tried many times, but none of the results are that satisfying. I will continue with this one. Don't forget that we have a magic tool, right? So we can use generative field in order to fix the pant. So let's try to do that. I will just get a closer look. Select this area where it gets white. So without typing in a prompt, I will just click Generate. Let's see if it, if it can fix it. And apparently it does. So this is way better. So let's do that for the rest of the broken parts. Alright, Dr. Sam, iterations I could I could fix some of the parts but I couldn't fix some other just like this black region right here, I wanted to remove it. But Firefly consistently keeps it. But if you use generative fill again and again, so you will ruin the whole image. And after some point, I think the reality goes away and it's just hard to make this image realistic again. So I will just stop there because I want to talk about the shoes. So first off, let's select the shoe on the left and just describe our prompt as a sneaker. Let's see what it gives us. Okay, So these are the results. First one is not bad, second one is not as snicker at all. And the third one actually the best. Maybe there are problems with the ropes on me, but let's keep it because I want to show you something different. So the left one is generated. Now we want to create the right one as well. And we expect to get the same results, right? We need the same shoe on the right side. We'll select it and say sneakers again. And the results, I haven't even though the results are better. I mean, it's not the same, right? So doesn't make sense because shoes should be same on most of the cases, right? Alright, let's try some different ways to get that. So I liked the shoe on the right. So I want to keep it and then I want to change the left shoe. And this time I want to say the same thing occurs with the right foot. The results are not that satisfying. Actually. Maybe we should say same shoe with the right foot. And again, once again, it's not working. So I even try same shoe with the other food because right and left depends on the perspective, right? So I also tried that one, but once again, Firefly couldn't generate the right shoe. Alright, so all of that means that I couldn't find a way to generate the exact same shoe with the other foods. But what we can do is delete all of these. Let's go back to the initial phase. And this time I will select the rectangle tool and select both of the shoes at the same time. Now, let's say Black shoots, hit Generate and Diigo. So now we can get the same shifts. So Firefly is a way of creating the same style for both of the foods. Select both, put in the same layer. So these are the results. Even though, even though the shoes are terrible at this, they are saying, Let's try something different. I want to say futuristic snickers. Okay, Let's see, one more time. Firefly is able to generate the same shoe, applying the same style to the other foot. And this is the futuristic snicker and desired robot foods. I like them. And this is the last one, so we could keep going generating more sneakers. But please mark here that you should select both of the foods if you want to create shoots 16. Adobe Express Platform: Hey folks, I have got some great news for you today. Adobe released Adobe express beta, which includes Firefly features like text to image or text effects. So it is available at the moment at this URL, knew that experts.adobe.com. And I believe after the Beta version, it will redirect. It will be redirected to the stable express URL. Alright, so let's see what we got here. In new Adobe express. It becomes an all-in-one app with AI powered. So here we can see some taps. I'm at two Recommended tab right now. And here we can see Instagram square post Instagram Story TikTok video, and so on. Down here we have text to image feature just like in firefly, and text effects feature. So this is the place or the platform that you can adjust, edit your videos and image using Firefly features using AI, and create an export videos and images for social media, for example. So it's again all in one tool. And let's go forward. Here we have social media templates for Instagram or TikTok videos. And notice here we can create an edit videos on this platform on browser. So you don't need any GPU power or in a memory usage. Alright, we also have options like Remove Background and resize image. So these are all new, as you can see, we have video templates here for Instagram, real story or YouTube videos, mobile videos, and so on. This platform looks like Canva, right? So it is more user-friendly, easier than ever. I liked it. And next up we have photo, and here we have quick actions like texture image. And you can already use text to image feature in Adobe express. That's a new feature. And we also have document tab here if you want to create a newsletter, brochure or in Rescue Me and so on. As well as marketing is possible to create logo here, poster, invitation or cart. And lastly, January to AI top. Here we already know all tag how text to image books and texts effects works. But new team here is that you can now create text effects, for example, In inside of your project. For example, if you are recording and Instagram real video, you can create text effects with firefly and import directly into your project. Alright, On the left you have Home button and your stuff here. And it is basically your created files, Recent files, created files. And you also have access to the brands feature now, which is basically for your brand name features, for example, your color palette, your logo, and your favorite templates, and so on. In the Explore tab, you will find many templates for your projects. For example, these are for Instagram Reels and these are all editable, and these are for TikTok videos. This is for YouTube videos, for example, intro and outro and so on. You will also be able to schedule your social media posts from here, just like this. And all you need to do is to integrate your Instagram or LinkedIn or TikTok account. And then you can create your videos, pause in this platform and scheduled to post it at anytime you want. 17. Express with Firefly: Alright, but this course is about firefly, right? So let's go back to Home and under degenerate to tap. Let's talk about text image and text effects features in Adobe express, basically. And this is really similar to the firefly platform. So you can see the prompt of the image and you can type any description here for your image to be created. And this is exactly like the text in which feature in firefly. And likewise text effects feature in this one, like this one, you can describe the text effect and the texts you want to create and it will create for you. So let's see One of the image. Let's start with the image. Alright, Once you go inside of the project, you will see the properties on the left side, different time, the Adobe firefly. And it will take some time to load the project here. But let me talk about, in the meantime about the properties. Just like in firefly, we can change the content type, photographic art, which is great. You can change the style of the image as well. Diigo, we have results. So remember the prompt was modern New York's theorem with big windows, built-in, rude and shallowing with colorful books like that. Green later and arc modern Koch and solid or fluorine, four to five millimeter original editorial photo. Alright, that was the prompt and this was the result. So just like in firefly, we again have four options for alternatives for this prompt. And you can switch like this. And the reason here we have the properties on the left and alternatives is because you can now edit this image with other features of other Bay Express. For example, I can add a text on this image without using, without using any other program. So that's why I am calling this Adobe express an all-in-one tool. You can also add some media. If you add music to this, this will become a video. So let's go to media photos and you can drag and drop this image inside of this project just like that. And change all of the properties, right? One thing you should notice here, after, even after your query, even after you edit this project with others, other features, other properties, you can always go back to your textbook image, go into text to image, and you can still have access. You can still switch to the other alternatives. You can still change the style of the image. And so it just saves your settings here and it is always editable. You can always change the prompt as well. And that will create new image. So let's say we want to delete this image and we want to create a new text, right? So I will click add your text. I will just say New York wipes. And I will change the color to white, maybe. Now it is more visible. Let me just try to recommended texts shapes. I like this one. I want to add a background rectangle for this text to be more visible. So I just can go elements here and then shapes. Here we go. We have rectangles. Maybe we can select this one. Now we have this rectangle. I will just decrease the opacity. I will adjust the size. Okay, maybe like this. So we can adjust layers here as well. Like this. I will just centered the position. Now, I group them and I can edit it like this. Alright, these are the things you should be familiar if you're ever used Figma or Canva. This is just a browser-based design app, right? So what I want to talk about actually is, if you select the text, is the new text effect feature. We already solved that in the texts effect from Firefly. But now we can also edit using AI inside of the project. So if you just go here, you can see sample effects just like in firefly. And maybe we can try, I don't know. Maybe we can try blue smoke text effect. I think this will work better for our image. So let's try this trust retro. Okay, Let's have a closer look. This is our text effect, so we can maybe place it center. But I want to try something new. So I want to type newspaper. Okay, let's, let's see what's firefly algorithm will give us for the newspaper prompt. Alright, That's perfect. Maybe these are, maybe we should choose tight option here. Click Generate again, because it was spreading too much. That's so much better. Okay, I liked that output. I hope it was clear that all I'm trying to explain here is that with Adobe expressed, you can now use Firefly features together in all-in-one tool except degenerate to fill feature item no, it will be available in the future, but it is not at the moment. But you can use texts effects and tax payments pitcher together here. And that's not the only option. You can also use these features inside of an Instagram video, inside of a TikTok video inside of a story. And we will see that in the next video. 18. Use AI in Videos: Alright, now I am back in the homepage of Adobe expressed. Again. I will go into the video tab and let's try Firefly features with video and how it works. Under the video tab, we have some templates. I want to use a popular from the popular, we do templates. Maybe like this one or this one. This looks great for me. First of all, since we just opened the video, we can see the timeline in here, and it is very simple, user-friendly. And wanting to notice here is that the objects is not visible here. At first, once you click on any object In the video, you will see, you will see them in the timeline. And they are just, they just appear here. And they are also editable. You can edit the duration, you can edit the position of the object and so on, as well as the Properties menu on the left side, you will have some recommendations again, since this is a text, you have texts effects feature again. And you can convert this text into anything. And you have an animation, you have the animation option also. Here you have three kind of animation for in during the animation and animation out. So since dropping and dropout selected here, let me play the video. You see this, the text just drops in and run it. Finished, it just drops out. So the bottom arrow selected here and probably left will be selected in the dropout section. We can also delete it and add our own. For example, maybe I want a slight in animation for entering and Diigo. Okay, I will, I will go with tumble. Alright. Now let's do one more time. So this is the animation. So it is really easy. And you will edit all of these objects as you want. But what I want to do is that I will go to media and I will select text to image feature here. So again, you can choose the image size, let's say portraits. Alright, new text field appears, sorry, new image field appears. And we can try describing things. At this moment. That says Technology and Innovation. I want to create, I want to use text image, and I will just say technology and innovation. Let's see dot-coms. And maybe we can just decrease the duration of this image. Let's wait for the results. Now we have the results. Let's make the image bigger. So these are just some abstract in which we can select one of them. The first one is here as well. Let's go with the first one. Maybe we can just put it here like this. And I will add just the layer one more time so that it will be behind these texts. Maybe we can make it bigger. Okay, This is much better now I change the blend mode to screen, so this will reduce the contrast and brighten details and I believe it fits better to the image. Okay, so now we can also add some. In animation. It can grow or shrink, shrink better. It can maybe four out. And mentioned we can select fate. Alright, now we have our animation ready, which takes two image generator. We just created an image and imported inside of an Instagram video. Alright, let's see. Maybe I can just expand the duration just like this. Let's plate. Not about to write. It couldn't be better. I agree. 19. Firefly for illustrator: Adobe has very recently released and Adobe Illustrator with Adobe Firefly integration, it allows you to easily recolor any of your artwork in just seconds. So let's try out some for ourselves and let's see how it works. I have a few examples here, and I just got them off of Shutterstock, like ready Vectors. So I'm going to open it with Adobe Illustrator. And it's going to open up this canvas for us. And it's a nice illustration here. You can see it's all vectorized and you can go in and manually change the colors of each element. But that's not what we want to do with Adobe Illustrator's latest version. So we're going to select the entire artwork by pressing command a if you're on Mac or Control a if you're on Windows, then we're going to go to Edit, Edit Colors. And we're going to go to Generative recolor. It's going to show that it's only in Beta, but I think it works pretty well. So here we have the prompt box, and here we have some sample prompts that we can try out. For example, let's try out trippy disco. And it instantly rights in the prompt, in the Prompt window. And you can see it generating some options for us in just seconds, which is extremely fast. And you can see the instant changes right here. So now if we go back to where we were, let's actually try and write our own prompt. So I want to try vibrant colors. 80s mood. The more specific you are with the prom, the better results is going to give you. So let's see how experimental gets. So this one is a little too dark. Maybe this one, I feel like this one is the best one. I want to add like a red tones. And let's type in synth wave. Maybe that's going to give us a little more of an enhanced result. So let's see. Yes, I feel like this is more of a vibe I want to go with. And here we go. We just recolor the entire artwork in seconds. You can keep experimenting and it's still all customizable by hand as well. You can move around your elements, no problem whatsoever. So let's look at one more example. We're going to right-click open with Adobe Illustrator. And here we have a little more of a complicated artwork. So we're going to select the canvas. We're going to go to Edit, Edit Colors and Generative recolor. In just a few clicks, you can recolor your entire artwork. It can take a little longer depending on the complexity of the artwork. As you saw, it took a little bit of time to open this window. But once you're in here, it generates it pretty quick. So for example, summer by the sea. Prompt gives us these kinds of results. And there is actually one more thing that I want to show you. That's a really nice perk. Not only can you recolor Vectors, but let's go and open up a new canvas. Doesn't matter the size, just any, I'm going to do it 1080 by ten ADP. Just so it's a classic size for Instagram posts. And I'm going to actually drag an image in here. So we can see we have this image, Let's fit it to Canvas. That's nice. So an image you can't really recolor it, right? It's grayed out here. But if you Image trace it first and let it load for a little bit, let it think. And then we expand into shapes. You can actually see it transformed our image into shapes. Now if we select it and go to Generative Recolor, you can recolor your photo. And it's that easy. Depending on the result of the Image Trace. It can produce different results. But most of the time it works pretty well. Just play around with the image tracing if it doesn't work the way you want it to work. So yeah, this is Generative Recolor and I can't wait to see what other users do with this feature. And I can't wait to see what more integration of Adobe Firefly they implement into Adobe Illustrator 20. Advanced prompt examples for text effect generator: In this quick lesson, I wanted to go in depth about text prompts for text effects. So let's start by creating a cookie House. Let's see what it generates. And we can see it managed to do it pretty well. We can see the windows, we can see a roof, and we can see it being applied on every letter, maybe except for this one. But this can be like a facade of cookie dome of the cookie house. And there are few options that we can choose from. It's kind of has a little more detail and has the color palette that I like. So I'm going to go with this one. And now we just want to build on this prompt step-by-step, ensuring that it executes every detail well. So I want to add a cookie house with a cookie monster. And let's see what it comes up with. Because we're giving it a character. Now, I'm interested in how it generates that character. We can actually see it wasn't really able to reproduce a cookie monster the way I wanted it. So let's try something else, something a little more abstract. First. So let's say cookie house, or let's try another subject with a dog. The dogs are a little distorted, but we can see some resemblance in there. I feel like this one might be more accurate. Yes. So we can see actually more dog face is right here. Then I want to add Winter. Make it a winter cookie house. And you can see it actually added some snow textures on here. And we still have some dogs here. So let's try something completely different. Let's say. Let's start with theory. And it should generate us the characters. And we can see a beautiful fairy here, and we can see a very light kind of vibe out here. I'm really happy with this fairy song going to keep this, I'm going to add flying in the forest. And let's see what it generates when we're giving it an action instead of a noun. And you can see some beautiful, beautiful results. You can see the forest, magical forest. You can see the ferry. It also generated some more fairies, can clearly see there in the forest. And I really, really loved this prompt. Okay, so let's leave this one at that. And let's also try generating some pop culture movies, some video games. So let's try out. I want to see what vibe it gives if we write Grand Theft Auto, if it understands what I mean, and we can see that it doesn't really get the reference. I don't really know what's going on here, but I feel like this is like metal parts of the cars. So let's try something a little less niche. So let's try Shrek. Let's see if it gives us a swamp. It gives us a donkey. Here we go. This is the vibe. This is more of a Shrek. They're all, there's already green, which is a good sign. This monster does not look like Shrek. But let's look at another example. And this is, this is a little creepy, but it has the swamp and the green vibe. So I can totally get with that. Let's also try Game of Thrones. You can see it gives off that beautiful, majestic vibe of Game of Thrones. There is also another option that caught my eye. I like this one better. This one looks like a castle and it's generating pretty well. And I want to go a little bogged here. What if we did a Game of Thrones with Barbie? I am really curious how it executes this. You can see that Game of Thrones is kind of overwhelming the scene right now. But there's some pink, I think some resemblance of Barbie out here. Here's some more Barbie. Let's see, This looks interesting. It looks kind of creepy though. Maybe if we were a little more specific Game of Thrones with pink Barbie world. Pink Barbie, pink Barbie world. Let's see what that gives us a here we go. This is much more. There's all I wanted. It still has that castle feel, but it's all pink, that Barbie vibe. There's some resemblance of Barbie, but not really. But yeah, this one, this is more adequate. I feel like this is a little creepy again, but anything to do with dolls seems a little creepy to me. But this kind of has that Game of Thrones slash Barbie vibe. And I think I'm going to finish off here. Make sure to be as specific as possible with your prompt. Experiment with doing subjects and nouns or writing action and can generate some really different, some beautiful results. But it can go also not so well and not in your favor. But It's all about experimenting. Ais still knew this Adobe Firefly version is still in beta. So I'm really happy with the result for beta 21. Firefly for branding mockup: So in this video, I want to show you how to create a branding mockup for anything using Adobe firefly in Adobe Photoshop Beta. So first, we want to extend the canvas by dragging from both of the sides that I want to extend. Going to click Enter. And then I'm going to select the area that I want the AI to fill in. Here we go, I want to click on generative field generate. And by not typing into anything into the text prompt, it will just fill in the scene. So here we have a few options, and I feel like this one would be the most suitable for what I'm trying to achieve. So I think I'm happy with that. So once we have the generated wall, I'm going to make a selection, create a new layer, fill it in with any color. And then I'm going to convert this layer into a Smart Object. And then I'm going to command T on a Mac or Control T on Windows to change the perspective of this layer. Then once I have the perspective, I can go into the Smart Object. I can add my logo to the smart object. I can disable the background. Let's save the layer. And we can see it's in the perspective of that layer. So I'm just going to move it up a little bit. I'm going to fix the perspective a little bit. Because right now it's getting a little too distorted. For my liking. Here we go. I think that's better. And I can select Multiply. I can enable a background, any color. Let's fill this with an orange. Let's save it. And we can change this mode to multiply. Let's select a different color. That's not going to be too harsh. Let's change the color of the logo so it's a little bit clearer. To see the result. I'm going to save this. Here we go. We can see this color isn't really visible. So let's see, let's choose a different color. And this should be a little more visible. So here you go. You can see it's in the perspective of the building. You can see the logo here is in the perspective of the building as well. And this was, as you remember, AI generated. And you can tweak it around and do it till perfection. But I'm pretty happy with this result for now. You can apply this for any kind of branding mockup. For example, like a laptop or an iPhone or building, or a banner, huge banner on the street. So feel free to experiment with this 22. Turn your portrait into surrealistic art: In this video, I want to show you how to transform any image, any portrait image, into surrealistic Art. So I have this image of a model right here. So first I'm going to select the brush size and make it smaller than zoom in. And first I'm going to select her eyes and generate VR goggles and see where it comes up with. And these are kind of close. These are better. But I think I want to make the selection a little bigger and let's regenerate. I'm actually happy with these results. For example, this one is really good. I'm gonna leave this one. So now that we have this, I'm going to increase the brush size a lot and I'm going to select the model's body. And let's reduce the brush for some detail work. And make sure you select the entire fit. Let's write in here we have. So now let's type in space suit. And let's see, these are definitely a little more surrealistic. Like her neck here is kinda weird. With that one was the best result that we had produced a here we have some better, maybe not this one. Some better results out here. As you might know, AI has struggled with hands. So let's not select hands. Let's subtract hands. Unbelief some skin. It's all about experimentation and trying to see what works. And what doesn't generate well, sleeve like this. Let's see what this generates. And these are actually much better results than previously. Like this one is really good. It fits her. And not this one, but this one. The neck is nice. The hands are all good. The soup fits her. And I think I'm pretty happy with this result. And I'm going to keep this. And here you go. You can do this for any image that you import in Adobe Firefly, and you can generate as many pieces as you want. So feel free to experiment with it. And I'll see you in the next video. 23. Use ai for ads design, poster design, flyers and entertainment: In this lesson, I wanted to show you how you can utilize Adobe Firefly and it's features in order to create some ads. Some add graphics and flyers for that. First I'm going to go into text to image generation. We're going to generate neon concert hall. Dj in the middle, in the DJ booth, in the background. Audience in the ground. I'm trying to be as specific as possible so it generates the results. I want. I have a typo here. Let's see. We've got some great results out here. And think I'm gonna go with this one, and let's download this option. And I'm going to import the image in to Photoshop beta to extend an image. I'm just going to drag this out here. Not that much, little less right here. So it's going to extend the image in these two directions. And so on going. And so I'm going to make a selection of this area that we just created. Take a little bit of the poster so it has a reference. And I'm just going to click Generative Fill, generate. And you can see we've got a few options of continuation out here. And I feel like, I like the abstract vibe. The whole thing is kind of abstract and I like it for it poster. I think I'm going to go with this one. It actually continues the people here pretty nicely and in the same vibe. So after this, you're pretty much good to go. You can add your texts out here. Now we have some empty space to add text. And you can add your event information easily. And this is how you create an AI generated add or a flyer 24. Stopmotion animation with Firefly: Today we're going to learn how to create an animation using Adobe Firefly. So we're first going to generate our images using text to image generation. So what I'm envisioning is imagine it's space and there are two astronauts, and they're in the outerspace from the space station. And there was an explosion. So there chords got ripped away and they're flying away into the open space. So that's what we're trying to achieve today. So I'm going to start off simple. So I'm going to start by tapping and astronauts outer space. And be sure to not make any typos because Ai doesn't really understand typos and can give you really wrong results. I'm also going to change the aspect ratio straightaway to widescreen. I'm going to want it very photorealistic. I'm also going to select dramatic lighting to specify the look I want to achieve. And I just noticed a huge typo here after mentioning to not make any typos. So I'm just going to fix that. So this is more what I'm looking for, more of like a whole body. So let's work with this image and we're First, I want to add astronaut in outer space next to Space Station doing a spacewalk. So I want to see that cord attached to the ISS. So this is already much closer to what I want and I want to specify for it to be a wide angle. Now I want to generate two astronauts to astronauts, so they're kind of really close to each other. Something weird going on with the hand here. This is more like a portrait. Let's try out two astronauts refresh. Not too happy with the results that degenerated retina. This is much, much better. Perfect. Like these only have on astronauts. It's kinda weird with the hand here. This was really good, but there's only one astronaut. This one, and this one are the best ones. I think I'm going to use this one. Actually, I don't like the rock here, but I don't like that the space station is here, or maybe it's like a Sputnik. Let's go with this one. So let's use this as a reference image. And let's also add to astronauts in outer space next to each other, next to Space Station doing a spacewalk explosion flying into outer space. This is what I'm going to generate after it's used the reference image to generate some more results. So here we have the reference image generated photos. And let's now add the explosion flying into outer space. So it actually generated not what I want, it made the explosion. In the background. Think I'm going to change the prompt a little bit to astronauts. Outer space, next to space station. Next to space station Explosion. Spacewalk. Two astronauts flying away into outer space. I want like there to be for them to be impacted by the explosion. So their stomach is in their arms, are stretched out. I'm not sure if I'm going to get the result I want, but it's worth a shot. Let's think we're going to have to completely change the prompt to astronauts impacted by Space Station explosion, flying away into outer space. While doing a spacewalk. Let's try this. I'm glad we saved the reference image, so it's not going to be far off from this. No. Maybe we should remove this reference image and let's give that a go. This is a little closer. This one maybe adding in front of them will actually make a drastic difference. So we'll see, I'm going to remove the wide angle. Here we go I feel like this is a little closer to what I want to bit weird and legs and arms. Maybe doing it a close-up will generate something better. Let's put space station Explosion and the front. Here we go, simplifying it a little bit. Did its job. I think. We instead road space station Explosion, two astronauts flying away into outer space surrounded debris. I think I like this one. So let's show similar to this. I think I'm still going to go with this one. So we're going to use this as a reference image. And we're just going to start downloading our images to collect our frames for the animation. And we're also going to make the images closer to their reference image. We're also going to make the generations closer to the reference image. Now I'm just going to start downloading these images and collecting our frames. Right now we've got four. Let's do this a couple more times. So if we preview our images, we can already see some action happening here. And it can see the way the scene changes. Now I'm going to actually go ahead and import my images and drag them into the sequence. So right now it's extremely slow because every image is like 4 s long, like five. I'm going to command a, select all of the clips. Now I'm going to Command R. To change the duration of the clips. I'm going to do this two frames, and I'm going to select Ripple, Edit, shifting, Trailing Clips. Here we go. Here's what we have. And let's change this so it's a little bigger. I want it alone, more flowy. Going to add cross dissolve to everything. Just did what was copy? Command C. And then I select all the clips in, clicked on Command D. But now should be more flowy. It stops here because the same frame. So I'm just going to delete this. Then I want to nest it. Select all of the clips. Right-click nest. Going to slow down this sequence. Pressing Command R. Do it at five. Make sure to select optical flow. And here's what we have. Let's render this. And let's loop this animation and have a look. And the hand is kind of weird here, but this is supposed to be a surrealistic animation. And I kinda like the debris coming in and disappearing. I really liked the changes in this astronaut. I would have liked them to be perspective shifting, so like flying away further out from us. But we would have needed to play around with the prompt a lot longer. For now, I'm really happy with these results that we achieved in such a short time. And I can't wait to see what you come up with and what kind of short experimental movies you can create with this? 25. Terms and Policies: Well, in this video, we will talk about terms and poses of firefly and the frequently asked questions. This video is based on the information available as of May 2020, 3.4. The latest information about Adobe Firefly, please visit Adobe's official website. First things first, data collection. The first Firefly model is trained on Adobe Stock image, openly licensed content and public domain content. That's why you cannot generate image of licensed characters like Iron Man or Batman, specifically with Adobe Firefly. Adobe is all about diversity and they've got it covered when it comes to their image, datasets carefully curate and pre-process the data to make sure it is free from harmful or biased content. And guess what? They also respect the rights of artists and intellectual property. That means they are building datasets that are not only diverse, but also ethical and respectful towards their customers and the community. Next is improvement and continuing to Adobe is committed to continuous improvement. They are not just releasing fireflies and forgetting about it. Adobe provides feedback mechanisms for you to report any potential a biased outputs or share your suggestions. They want to work together with the customers to make Firefly even better? The answer of this question is simply no, copies of customer content will not be included in the firefly models. As Adobe describes. Adobe does not train on any Creative Cloud subscribers personal content. Speaking of that, content credentials. So what the heck are Content Credentials while they're like these attributions, description details that comes with content. You know, like when you export or download something, these credentials can be attached to it. They give you extra info about how that content was produced. Now why do we need contempt credentials? They're all about giving credits. They help content producers get recognized for their hard work. Plus they are great for people who lived the content because they can make more informed decisions about whether to trust it or not. That's right. Transparency is the name of the game. When you download an image generated by Firefly or share it in the firefly gallery, you will get Content Credentials. These credentials simply that people know that the image was created using AI. Specifically for Adobe firefly. It's all about being transparent and open about AI generated content. I picked three most asked questions for FAQ part. And please don't hesitate to ask me if you have further questions. So first, what is the maximum resolution that Firefly producers is 2000, 2000 pixels. And for the file types, Firefly supports the following file types, JPEG and PNG in the text effect and text to image models and SVG in vector colors. Finally, if you want to share the image degenerate with firefly, you can do it with the exception of the elbow guidelines. So if you take and credit Adobe firefly, you are good to go During the beta generally to imagery and the texts from Firefly should not be used for commercial purposes. So that was the terms and policies of Adobe firefly. And as I said earlier, if you have further questions, you can just ask in the comments. 26. Upcoming Features: Let's have a look at the upcoming features of firefly. And these are the things that I'm most excited about because most of them are brand new technology and they will make creator's life much easier. Let's start with January to fill this feature is likely to be about generating content to fill a given area in a design. In other words, Adobe firefly could create patterns, textures, or even complex objects based on the text prompt and fill that defined space with the January that the image personalized results. This feature might be more tailoring dice output to the individual users style or preferences. Potentially learning from the user's past designs or preferences or enrich e.g. to generate results that are more likely to match their aesthetic or functional needs. Tax director is simple enough. This feature would like to take a text description and generate a vector graphic based on that description. Vector graphics are scalable without loss of quality. That makes this feature incredibly useful for design work. Next, we have extended image feature. This feature called allowing users to automatically extend the content of an image. For instance, if you have a picture of a forest or a lake and you need a wider image for your design. Adobe Firefly might be able to generate additional pixels to the sides of your original image. One other example is 3D to image. This might refer to a feature that can take a 3D model or seen and gender it as a 2D image. This could be useful for creating realistic scenes are objects without needing to manually draw or design them. Again, according to this video, you will be also able to change their environment, change the style and properties of the image. Texture pattern is nothing new. You can also do this with other AI tools as well. But yet it is really useful for any usage of textures, patterns that are numerous. And we also know that Firefly will assist other programs such as Photoshop or Premier Pro. Though texture brush will generate brushes for Photoshop and fresco from a detailed text description. E.g. if you write dotted lines, Firefly my generate a brush that paints dotted lines. Next is one of my favorites sketch to image. This feature called potential to transform my rough sketch into a more polished and detailed image. This might involve interpreting the sketch and filling in details, colors, and textures, that is on the AI's training data. Last but not least, thanks to template, this feature might allow users to generate the template for a design based on a text description. For instance, if you input a business card with contact info on the left and logo underwrite, firefly, cool degenerate the template matching that description and you will be able to use that template in your further projects. Alright, these are the upcoming features of Adobe firefly. As I said, I am really excited about these features and I will update this course as they launched these new features. 27. Course Project: Hey folks, we have come a long way in our journey exploring Adobe Firefly. Now we have a special challenge for you, of course project that will test your understanding and creativity. Let's get into our course project resource around to key features we have learned, texture, image and text effects. You can create amazing background image with what you learned in this course. I encourage everyone to create a background image for your computers. First, you should use text to image feature of firefly. And here the prompt is up to your imagination. I will also do the challenge. So before you start, you can watch mine or just pause the video here. Do your challenge and watch later. I will say very angular, cute robot sitting on a chair looking at the camera. There you go. These are great, but we want to change the aspect ratio. So select white screen here. And I also want to add some styles. Let's select digital art. And since we are going to combine with a text, I will select minimalism because I don't want objects in the image. Instead, we need empty spaces. Okay, not bad. And let me try to create an empty bowl on the right side so that we can put the text on there. Great, I like this one most, but let's see the similar versions. I will click on Show similars button here. And here are the results. These limits are good, but actually they are not looking angry except this one. So I'm going to say show similar one more time using this image. And I hope you will get some angry robots this time. There you go. The first one is perfect. Maybe there is no enough space at the right, but we can fix it in Photoshop. And actually that will be a great example to use generative field feature of firefly. Let me download this one. Then it is time to create the text effect. I go back, go back one more time and down here, select text effects. Okay, I want to type that away from my computer. And it looks like there is a character limit currently, so I can not write anymore. Okay, no problem. Let's say PC instead of computer. So get away from my PC for the prompt, let's say robotic. And we can maybe include angry faces inside of the characters for fun. Alright, these are the results. Let's see other versions. Okay, not wet. They are actually anger faces, but I am not sure if you use this effect. Let's try angry robots instead of faces. Okay, not bad. Look at this. In this version, there's a robot in shape of f character. This is fantastic. I loved this one. Should we downloaded or wait. Now I remember that our image background was also gray, so this techs can't be visible. Before downloading. Let's try to change the color. Firstly, I will try changing the color here. And apparently it doesn't affect at all. Let's select another color. Or maybe we can provide the color in the prompt. I will say red here and there you go. Now we have red robots. So keep in mind that the prompt is stronger currently. I think Adobe will fix these problems in the short term. Alright, anyways, we got the text effect and it is red. And this is the text we want to use in the project. So let's go ahead and download without the background. Here we are. I open the knee Photoshop project and import the image, regenerate it. We will also import the text. But first I want to show you another example of January to fill. As you remember, we want an empty wall on the right side and robot to be on the left. Let's see what we can do. Either move this image elder left. And now we have this white region on the right. Select this region, and after you select generative filled box will appear. And remember that currently this feature is only available in Photoshop beta. Click on January to fill and click Generate again without typing any prompt. We'll extend the image without adding anything. And Firefly, artificial intelligence complete the image looking at the left side. And boom, this is incredible. I cannot even tell the difference. Okay, now the image is ready and next step is to import the text. Let me edit the text position and size. I will speed up this part. And that's it folks. This is my background wallpaper for this challenge. Here you can see how it looks on my computer and please share your work with the community. And I am excited to see your projects. 28. Comparison of Ai Art Generators: As a generator AIR test, you should know the alternatives for Adobe Firefly. There are many other tools in industry, but in this video, I want to compare the most popular ones. Namely, we will be talking about firefly, meet journey, stable diffusion, and Dalley, there's known best peak. Each tool has advantage in different areas. The best way to figure out which one works best for you is to try them out and see which fits your workflow of the bust. And I will try to explain the differences, pros and cons according to my research. Let me introduce the tools first, starting with Dolly. This AI tool was launched by OpenAI in 2021. I need quickly gained attention due to its ability to generate unique image from textual descriptions. The software's name is a mix of the names of animated robot Pixar character volley, and the Spanish surrealist artists Salvador Dali. Delhi is the first molar. Meet your new, launched in July 2022. This art generator was designed by a team of AI enthusiast who aim to make AIR accessible for everyone. My journey is currently only accessible through at this court boat on their official discord server. You can direct message to the vote or invite the bot to a third party server. Stable diffusion is a deep-learning text or image model released in August 2022. It's unique feature is the capability to gradually evolve the artwork in real time, allowing the artist to observe the creation process and make alterations at any stage. It was developed by the startups table DEI in collaboration with a number of academic researchers and non-profit organizations. Last but not least, Adobe launched firefly in March 2023, releasing two beta text frame which allows users to input a natural language problems and receive an image similar to other generative AI in which models, but there are more features in Firefly. At launch, the firefly beta is only available through the website. But Adobe stated it will soon be integrated across its apps and it is already integrated with Photoshop. All of these models are trained using the dataset. Now Lee was developed with something called clipped. It is by OpenAI. Clip is a separate model that was trained on 400 million pairs of images with texts captions scrapped from the Internet. Make journey uses many datasets like Liam, which was scraped from the web as well. In Forbes interview, majority of founder David halts has admitted that meet journey did not receive consent for the hundreds of millions of images used to train the AI. Stable diffusion is no different. It again used pairs of images and captions taken from layer on five before training, which includes 5 billion payers of image and text. The dataset was created by Leon, a German non-profit which receives funding from stable it to as many time status in this course, adobe Firefly trained on a dataset of Adobe Stock. They are saying that they did not train on any Creative Cloud subscribers personal content. Next up, capabilities. In terms of capabilities, there is no official way of comparing these tools, besides they are better in different areas. So I'm going to share my own thoughts based on my own experience. So I decided to say the key features of each. Delhi is the first mover and still good for photo-realistic Emirates. Also have API access, which lets you create new enrich in any number of different styles as specified by the users prompt and include in your projects. Meet journey, in my opinion, is the best in terms of image quality. As you might guess from its popularity, the best outputs comes from MIT journey. E.g. MIT journey is the only model that can handle hand generation. While others still struggles with hand drawings. You can define different art styles and mediums. However, MIT journal always creates image in Ralin meet journeys style. And you can easily differentiate the image generated with MIT journey. Somehow it generates aesthetically pleasing results. There is no API or standalone platform for meat journey at the moment. On the way to use it is through this chord up. Stable diffusion is best for developers since it is open source and can be run locally on your own computer, which means your system capabilities is important as well. E.g. tangible bite or more RAM is recommended. On the positive side, stable diffusion basically gives you more options and control. You can set the number of steps, the initial seed and the prompt strength. Lastly, Adobe Firefly is really competitive when it comes to image quality. But there is more than text to image, like text effects and January to fill. Generator field is one of the most revolutionary thing about firefly. Since it opens a new door in AI art, it will be integrated in Creative Cloud programs, like it is already been in Photoshop. But I also asked the firefly team if it's going to be a standalone platform and that's where it is not clear at the moment. They said they are still testing and considering the roadmap. Alright, that was the wrap-up of capabilities. And please consider that an ecosystem is changing fast. By the time you watch this video, some things may have already changed. Lastly, let's see the pricing models of these tools. Dolly gives you free 25 credits. There's a rolling monthly, 15 free credit allowance. If you need more, it costs $15 per 115. Let's meet journey is a subscription service which prices starting at $8 per month annually and $10 per month monthly. Stable diffusion is free to use on the web app. And you can run the tool without paint as well. But if you want premium advantage that our packet starts from $9. Adobe firefly can be accessed for free using debate access. To use Photoshop beta, however, you must be a Creative Cloud member with a subscription or trial that includes Photoshop. All right guys, that was the comparison of AIR generators. Thank you for watching