ChatGPT + Adobe After Effects: Animation Made by AI Expressions

1. Introduction: Hi, I'm Jake. This less will teach you how to use a HA GPT free version to create cool expressions for Adobe After effects. You don't need any experience working with Adobe After effects or CHAD GPT. I'll explain each step we will take to create this kind of animation using just three sliders. Learn which prompts to use to add features to your project and how to interact with hat GPT, to build cool animations in Adobe After effects. You can always write your questions in the discussion sections under any of my classes. To get help for me, while creating cool effects, I think, is the best way to learn the software while creating interesting effects for fun. And to complete this class, our goal is to create this cool animation using just three sliders built with expressions provided by CHAT GPT. I'll be happy to see in my class as also your result after following all the steps. So let's get started. 2. Describing Idea: Okay, so as you can see, I'm going to use Adobatter Effects 2024, and you can pretty much use any version of Adobet effects. It will work fine. Also, I'm going to use Edge browser because here I didn't login in any of my accounts, and I want to use a free version of HIDGBT. So basically, I just went to chgbt.com, and here we what we're going to get. Okay, so first, let's describe what we want to achieve with our template. Okay, so with my poor knowledge of English, as you know, I'm from Ukraine so English is not my first language. I've tried to describe what I want to see here. You can just pass this video and read and see how well I've described what I want. But for you, I'm going to show you what I want inside of dobiotreffect, because visually it will be a lot easier to understand. So as you can see, I just want to have this kind of result. I want to select this layer, for example, and I just want to stretch it, something like this, and these other layers should squish and make the space for this to be open. Want just to stretch it like this, and all of these should be responsive and squish it to make a room for this main photo and then squish it back here and then to have this ability for this one, also to extend it from the middle, and this should be squished in the other parts. And same goes to this one. So this is pretty ambitious project, and if you see this class, then it means that I've got something that I want to share with you. And the next video, we are going to see what Cha GPT free version will led us to do. 3. Adding Expressions: By the way, this is first for me, I didn't use this expression and I will do everything on your eyes, and you can see that all of my steps which are going to take to create this kind of template. So now we can see that it already understood the assignment and it described even better than I wrote here. So we can just follow this step by step, all of these extractions which it gives us. So first of all, I want to create a new project and start over. Let's click here on standard layout. So my layout should be similar to your and I want to start over because all of my visual representation which I just showed you, it can mess up because HA GPT expect us not to have anything and it gives you all the steps just to start from scratch, which is really handy. Now we can just follow along with the CHA GPT, which also understood the assignment, and we need to follow step by step guidance. And from there, we are going to add some features and try to see what we can get even more out of our idea. First of all, we need to create placeholders. So let's do exactly this. We can just click here on New Composition. It's a good idea to use the same naming because it will refer in the next prompts, which we are going to ask to the same naming which it already provides us. It asks us to create three compositions, placeholders, placeholder one, two, and three with full HD resolution. Let's go to after effects and let's call it placeholder one and we can get this width and we need to type here 1920 by ten eights to get the HD resolution 30 frames per second and let's leave it 15 seconds long. Let's click Okay. So basically, we need to have three of these. So let's select here and Project panel and press Control D two times to get three placeholders, as HGPT asked us to do. Also, it asks us to use image as it for testing. So I'm going to go to File, Import File, and these three images will be also provided in resource step to the slabs for you to download. So let's import in these images which I provided in resource step to the slash, click Import also, you're going to see this chat text file. Here I'm going to copy all of my conversation with CHA GPT. And here you're going to see all the conversation which I had with HI GPT. You can see this is my prompt, then hGPT set. Why this is handy because when I'm going to copy these expressions, you can just locate it here in text file and just copy from here and use it, and I'm going to use this interface because it's much easier. I'm going to show you how I'm going to use it. So you can see how you can create your own templates using hGPT and workflow itself. To follow along with this class, you just need to open this chat txt file and just follow along. And as you can see, we have all these steps one, two, three, and as you can see, it's exactly these steps here, one, two, three, and then expression, and just copy all of these expressions which we are going to use. So we dropped our images into our Adobe After effect file. I'm going to save it as class, so you can just save as, so you can save your after effects project wherever you want. Let's open all of these three compositions. Let's select this first one, then hold Shift, select the last one, and leftmost double Kaken it to open each of these compositions. Let's drop in each of them one image at a time and press S to resize it. Let's open the second one, press in S to resize it, and third one also a and resize it. Then let's follow along to see what we need to do next. So the second step is to create a main composition. So it's a good idea to use the same this main composition naming, which ChIPT provises. So let's go to composition, new composition, and let's call it main composition. And click Okay. Close this render Queue. Let's go to main composition. I'm going to drag it here, so we would know that this first one is our main one, and we can just simply drag and drop our placeholders. And let's see what we have next. So what we need to do is to add shapes layers as a Mt. Create three shape layers as rectangle. So we need to create a three shape layers. Let's select MT the name to use the same one. So I'm going to select Mt let's go to After Effects. And what we can do is actually to select this rectangle tool here, you can just simply let mouse double click on it, and you can see it created a mask on placeholder three. I'm going to press Control Z to undo this step because we need to make sure that we didn't select any of these compositions. So let's click here to unselect any of then leftmost click on this rectangle tool. If you see some of others, make sure to left most lick and hold and choose rectangle tool. And then when you've selected your rectangle tool, leftmost double clicking it, and it will create this rectangle. Let's select this shape layer and press Enter and press ContolV to paste the name met one. We can just simply select this met one and press Console D two times to get these three copies of the met layers. So we've created these rectangle we need to use Alpha met for corresponding placeholder, how to do this. Put each met above its placeholder layer and then set placeholder check Mt to Alpha met or Alpha inverted if needed. Let's do exactly this. It asks us to place it like this. So the placeholder should be beneath this met and accordingly here and here, third to third, second to second, or the first. We need to click here to Toggle Switches until we will see this track Mt. So we need just to select the Speck web, alongside this track mat, elect most lick and drag to this mad above. So this should be connected to this one, and this should be connected to this one. So we've created Alpha met trick which GPT asks us to do. Then let's see the next step, responsive scaling with expressions. So what we need to do? It seems that we need to go to our placeholder layer, open transition, and then scale. So let's do exactly this. Placeholder, open transition and see the scale. So here we need to copy this expression. From the TXT file which I provided, you can just locate this expression which goes under fourth step and just simply select it and copy. In my case, I can click here to copy, and when you're going to use HA GPT, you can also do the same here. And let's go to After Effects and hold Old key on keyboard to be able to left mostly on this stopwatch, and it will allow us to write expression. In our case, we're going to press Control V to paste expression which HGPT provides and we've paste it in here and let's go to the next one. It also asks us to paste this expression to the position. So let's copy this expression to the position, and let's go here and hold Alt key on keyboard, left most click on the position, and press Control V to paste. So same thing we need to do to this position and scale. What you can do is to select these placeholders and press P to see the position, hold Shift, and press S to see the scale. Then you can go to Cha JPT and copy the scale, and let's go to the each of them. Hold Alt, left most click on the stopwas, press Control V to paste this expression. Also scale, hold Alt, leftmost l on the stopwatche along the scale, control V, and same we can do with this position. Copy. Let's go to after effects, Alt, leftmost l here, paste it in, here also, hold Alt, leftmost lick on the stopwatche and paste this expression. Okay, so we did this setup with the placeholder and position scaling. Let's see what we need to do next. So fifth step will be make layout adaptive across all three panels. It gives us these two ways what we can do. Use controllers and use expressions, and Option A is use expression controls is easiest to manage. Let's trust Cha GPT and go with this option A, and let's create all object and name controls. Let's copy this name, and let's go to after effects and let's go to layer. Now, no object. As you can see, we've created this Noll object, and to rename it, we can just press Enter and press Control V to base this name. But the trick is that it changed the layer name, but it didn't change the source name, and sometimes it could be a problem. So it's a good idea to go here to the project panel. And open the solid and select this null object here. Is this exactly the same null object as you can see? But if you will rename it here, press Enter and press Control V to paste this controls or just type controls. Make sure to not miss any letters because it will use the same name in the expression itself. So it's a good idea to type exactly controls the right name. As you can see, it renamed it here as well, because now the source name is also Control. So we've did this step. Now we need to add three slider controls on our null layer. Okay, find our effects and presets panel. In my case, it hide it here. You can just click on these ones and see where we have this effects and precess panel. If you cannot see it, for example, if I'm going to close it, you can go to Window and enable it from here, Effects and Presets. And let's type here slider Control and select it and drop it on our Ctrols layer. And let's also name it the same name as ChaPT proviss. So with one, let's go to Aftereffects select it, press Enter and paste it in. Let's press Control D two times. And it's a good idea not to use a simple values here and change it to something else. For example, to 100 or 210, whichever value you want. But sometimes these expressions doesn't link well if you will leave at zero. So just change to any value which you like, so I'm going to set to 100. It's a good idea to keep in mind that when you're using expressions which HGPT provides you, it may not work well if you will leave as a zero. And you can ask HAGPT why it didn't work, but my tip to you, it's probably because you didn't set the sliders to higher value than zero. 4. Asking For Help: Okay, so let's do this next step. So as you can see, it said that we need to use this expression and use similar logic to met two and met three. Adjust position X of each met based on the previous ones, so they are placed side by side. And here is result that I need to have. And basically, after when I'll apply this expression, it should be working fine. But for me, it's not really obvious where I should put this expression. When we're going to try to create your own templates, you may step into this kind of problem where you can see that you need to do something, but you don't really understand where to put certain expressions. So in these cases, it's a good idea to write a prompt to hav GPT. So this exactly what I did here. I just asked, I don't understand where I should put this expression from this part of your message, example math. So I referred to this part of this message. I said that I didn't know where to put this expression. It gives us this really handy step by step guide where you need to place this expression. So now you can see that we need to select our met then go to contents, then to rectangle, then to rectangle path one, and then size. Now it's super obvious that I need to open this met one, then open the contents, then open this rectangle one and rectangle path, and now I see the size. Now it's super obvious where I need to put this expression. Also it asked me, would you like me to generate the full set of expressions for all three meds? So as you can see if I ask these questions like, I'm a beginner and I don't know where to put my expressions, it will offer you to make even more step by step tutorial, which is great. I said, yes, and now it gives me this step by step setup. I repeated how I needed to create my null object, how I needed to name it, and how I needed to create our slider controls, which we already did. But now it says that I need to place this size expression into met one, met two into rectangle path size. So now it's more obvious which expression where it needs to go. But also, I see this problem that it showed me this name slider, and I know that in different languages of adobearoefect, it might not work if it uses not built in names of Adobe after effects but uses English names. For example, if you have French version of Adobe after effects, it will not work for you. So this is why it's a good idea also to ask, please write the same steps step by step guide, but with the expression that would work in any language of Adobe after effects. So as you can see, it said, absolutely, it will work in English, French, German, Japanese, et cetera. And rewrite all of these expressions which we saw earlier here, all of them. But now instead of slider, you can see that it uses this built in coding for each of these effects and now it will work in any language of adobat effects. Now I can safely copy this expression to rectangle size met one. Let's go to met one and go to this rectangle path one and size. I left mostly here and press Control V to paste it. Then let's go to this met two. Let's copy it and let's go to this. Meet two contents, rectangle, and size can show V to paste it. So let's go to Mth three, copy, and let's go to this. My three here, as well as size called Old left mostly on the stopwatch and press Control V to paste. So now as you can see, we already have something going on here, which is great. Let's see what other expression we need to apply here. So the third step after applying these expressions is we need to apply this position property for each met layers. Okay, so let's copy this position to met position. As you can see it different kind of expressions. I'm going to copy this met one. Let's go to met one, press P to see the position. Al left mostly on the stopwatch Control V to past it. And as you can see it jumped to the left, which is already coming together, let's press P because we know that we need to place this position to met two. Here we can see our position, selected met two, press P to see the position, old leftmost lick. Control V to paste it, and let's open here in advanced, this met three position, and copy this last expression. Let's go to after effects, leftmost lick on this stopwatch and press Control V to paste. Okay, so as you can see, it automatically arranged this and let's go to Controls and change this size. And now as you can see, we already getting this script to work, which looks pretty interesting and fun. Amazing how it works really well without knowing any of the expression writing. So let's see what is the next step, set up placeholder and track mats. And this fourth step and fifth step, it repeated us like we are beginners, but we already did this. We created our placeholders and set our alpha Mt. As also, as you know, we already change this scale for placeholders and also we've changed this position for our position expression. As you can see, we already created this really cool expressions that works well. 5. Adding New Features: So we get this really fun result. But as you can see, now it stretches the image itself. In some cases, it could be helpful because maybe this is the look which you would want to use, where you would want to, like, squish it and have this kind of transition. But I want to make this template. Like, these images doesn't stretch. It's only this mating technique is working. And the image, I would prefer that it will just move to the left or to the right, just keeping in the center. Exactly this, I tried to describe to Cha GPT. I understood my goal, my new request that we don't scale and stretch the placeholder to fit the mat just move, offset the image, and it stays centered inside of the shape regardless of the size. So here is updated setup. So what we need to change? We need to go to our placeholder and remove our scale expressions, which is pretty obvious. Let's go to this. These all placeholders, press S, and just simply hold lt and leftmost ly on the scale. It will remove these expressions. As I can understand, it already did the half of the chop, which already gives you this really interesting result. As you can see, you can get this really dynamic thing going here, and also it prompts us to add this to placeholder layer, this position expression. So let's copy, go to After Effects, and let's select these placeholders and press P, and let's add new expression. This one and then select this one and paste it in and this one, please folder one. Also, let's paste it in right here as well. Now let's go to Controls and see what happens. And now as you can see, it keeps our images pretty much in the center, which is exactly what I wanted. So now, as you can see, we can get really nice kind of animations, which we could try it out. So for example, we can just animate this will appear right here. Watch it a bit and get really interesting kind of animation from just these three sliders, which is pretty handy and pretty interesting to watch. As you can see, we even have this optional fit to place holder proportionally. So here we can get this kind of explanation if your image are large, and you still want them to fit within the shape without distortion. Can use scale to fit without stretch like this. So optional, we can just scale expression to fit without squash stretch. So let's try to copy and see how it would work. So I'm going to go to art effects and press here to scale, and let's paste this expression right here. And as you can see, it would work like this, and it's not exactly what I want, but it's a good idea to have this option. So I'm going to press Control Z. And control Z one more time because I don't want to have this kind of expression here. So this one looks much more professional to me. It gives this parallax effect, which I really like. It is pretty interesting to work with. So as you can see, it's pretty easy to jump from ha GPT, ask for the questions, and clarify where you need to put certain things and get this fully automated thing happening in a side of Awefect. And basically, you can just click here on the horizontal type tool and type any text. And cool thing that you can also check Matt to this center one and get this text inside of this one and you can use just your own text and make some reveals of your text. So finally, let's create a little animation. So let's create this first value like really big. And this we can make it smaller so it will fit our screen just like this. And let's create the keyframes for all of them and select the controls, press you on keyboard, and just move it here now let's make this smaller to zero. And extend this center one and see how it looks. It looks way too fast. So let's maybe let's make sure to select this one, press F nine, and go to this graph editor. By left mostly in here, make sure that you set to di speed graph and drag by these handles, so it will start pretty fast. Then model here, and we can get this kind of reveal. And we can even change this one to make final animation, just like this, go to Graph Editor and ragu these handles to make sure that it has this kind of animation. And let's also select this keyframe, press F nine, and play around with these keyframes. Interesting kind of looking animation. We can just save these keyframes just like this. Here, making sure that this also gets to zero or close to zero to fill our screen with the final image. And just by using these sliders, you can create this kind of animation where it stretch out and leaves a bit and goes out of the frame. So main takeaway from this class, which I wanted to share with you, that you can pretty much create any tool inside of Adoboerfex just by using a CHA GPT, even free version. I'm not sponsored by HAGPT, but I see it as potential and try to create your own tools using HIGPT. Can pretty much continue tweaking with this and ask for more features. And if you didn't get what you want, it's a good idea to save your project as iterations. So for example, save as working class Version one and then ask for new features, add them, and if it didn't work, just save as Version two, and then just go back to the version which you like the most because from time to time it get the results which you didn't really expect and didn't really want. In these cases, it's a good idea to write something like this. Let's forget this prompt that I wrote then write your prompt, which didn't result in a good outcome, and it will ignore those changes or all changes which comes after them, and you can just start over from the point of your conversation which you liked. So basically, you can just type that forget everything that we done after this prompt where it worked fine. Just copy from the moment in your conversation where it work fine and just try to add new things until you'll be happy with your new tools using HGPT and after effects. Feel free to follow me here on Skillshare. I have a lot of classes on adobo ter effects on VX inside of Adobe After efect. I don't really use HGPT in all of these classes. I just simply teach you how to use Adobe After Effects to create cool VFX, and you can click here on CmorT CMR. And it's pretty much conventional classes where I just teach you how to use dobiotor effects without any AI. Also, I have new classes this month, and you can see the newest classes which I've created, fundamentals of adobo effects, logo and text animations here, and I even have a classes on free AI tools if you want. Then I have complex Vat using Adobe After Effects, the WIT resolve class, and even more really short class, you can get this really cool animation in just 13 minutes. You can create a really awesome VFX shot in 12 minutes from this class. Choose the class which you like. Submit the project. As you can see, I'm checking all of them, and I can give you tips if you want. And also, you can ask the questions in discussion section to any of my classes. I read all of them and answer as quick as I can. Follow me here on Skillshare, and I'll be happy to see you in my next class. Thank you for watching.