How I EASILY created an AI SHORT FILM in a single day!

Today I'm going to show you how I was able to make this 90-cond short film using some AI video tools. I used Design AI, who are the sponsor of this video, and I leaned very heavily on Sora 2 and

Vo 3.1, so I could generate sound with some of my videos. Now, making a video like this in the days before AI would have taken weeks, if not absolutely months to create, and it's only a little bit over a minute for this short film.

But I'm going to run through exactly how I made it with some of the techniques.

But I want to start off by letting you know, first thing I did was I created some storyboards. I come up with a

some storyboards. I come up with a concept. I drew up a bunch of story

concept. I drew up a bunch of story boards onto some paper so I had a bit of a guideline to follow when creating each shot before I pieced it all together.

Now, from there, the next organic step would be to try and use AI to create images that match up to those storyboards so that we can then go ahead and animate them with AI video tools.

But the first shot didn't really require me to create an image first because none of my main characters were in that initial shot. So instead, what I did is

initial shot. So instead, what I did is I decided to generate a video using text to video to start the animation off.

Now, so that I can come back to some of these shots, I will be creating various different projects. So, I'm going to

different projects. So, I'm going to start off by creating a new one. And I'm

going to come up and just give it a quick name. And I'm going to try and

quick name. And I'm going to try and give it a bit of a numbering system. So,

I'm going to say one temple shot because I'm going to start off with a shot of a temple. And this is going to be pixar

temple. And this is going to be pixar style animation. And there's no

style animation. And there's no characters in this shot. So, I'm going to come down 16 to9 and all of my projects will be 16 to9. Hit apply. And

now I'm ready to get started. So, the

first shot's probably going to be the simplest. So, I'm going to come into AI

simplest. So, I'm going to come into AI video and watch this. Cling 2.1

standard. I'm actually going to change over to Google Vo 3.1. I'm going to type in a prompt. Have a pixar style animation

prompt. Have a pixar style animation shot slowly zooming in towards temple from far away. Pixar style shall monks are working in various areas in front of the temple as a camera approaches. Now,

I want to have a little bit to play with here. So, the first thing I'm going to

here. So, the first thing I'm going to do is come here to where it says 720p.

Obviously, I want to be a higher resolution. I'm going to go with 8

resolution. I'm going to go with 8 seconds because I want to be able to remove or chop or find the best part of the shot as opposed to relying on VA 3.1 to create exactly what I need. So, I've

got that set up. I'm going to turn sound on cuz I do want to generate a little bit of sound. And of course, I can always turn it off, but I have a prompt.

I'm not going to have any images for this particular uh shot cuz I want to sort of show you how you can increase the the complexity of each shot. So,

starting off with the simplest first, VA3.1.

And because it's V3.1, it's going to be quite pricey at 800 credits, but we're going to get some really good results. I

hit generate. And this is the shot we get.

And I think that's a strong start. So,

we're going to run with it. And so, that already shows you uh with one single attempt, I got a pretty good shot to start off with. And that's pretty lucky because the Sora models and the VA3

models are a bit pricier on the credit side. But the fact that I got that shot

side. But the fact that I got that shot straight away is pretty good. And it's

exactly what I had in mind. So having a good solid prompt that really describes what you want in the scene as well as the camera movement really does a great job. But now we're going to move on and

job. But now we're going to move on and start creating some of the images we need to match up with our storyboards so that we can then convert those images into AI video. The first one I'm going to use some consistent characters. So

I've logged into design AI and I come here to the character section.

I'm going to generate images of my character. I'm going to choose my Pixar

character. I'm going to choose my Pixar monk and type in a basic prompt. And now I have a simple prompt which includes my character. But I'm not going to run

character. But I'm not going to run through the whole process of how to create a character. But I do have full tutorials available which I'm going to link to in the description below that can help you run through this whole process. I'll also add a card at the end

process. I'll also add a card at the end of this video. So if you want to actually use consistent characters in your short animation, you can follow that process and that way you can tie things together again with a bit more

consistency. But otherwise, I create my

consistency. But otherwise, I create my first image by coming down and hitting generate. Now, it was able to generate

generate. Now, it was able to generate this image. But the problem is that his

this image. But the problem is that his eyes are open instead of closed. But

that's where design AI is really powerful because it has this canvas and all these tools uh sort of baked into it. I can click on the image and I can

it. I can click on the image and I can actually fix up and edit small areas of the image to get exactly what I'm after.

So, I've added this image to the canvas here and I come down to the AI editor section and then up to local edit. And

from there, I can select the eyes. And

from there, I add a simple prompt asking for closed eyes. And then when I hit generate, it will actually give me a few different options. And I can close the

different options. And I can close the eyes on my character. But on top of that, because I have my storyboards here for reference, I can comb through and see that I have various different shots of him in the exact same position, in

the same pose, but the camera's either much closer or much further away. So,

I've actually got a few different tools I can use within Design AI in order to make those shots up. Coming back to our image here. If I just simply size it

image here. If I just simply size it down into position, I can come over to the AI editor section again and I come over to generative expand. And from

there, I can actually use the remainder of the canvas as long as it's transparent. And I give it a little bit

transparent. And I give it a little bit of information of what I want to see in that area. And I can then generate the

that area. And I can then generate the remaining sort of like space around our monk. But I also need to zoom in on the

monk. But I also need to zoom in on the image as well. So, I click on it, come up to the crop tool, simply move the handles into place, holding down shift to keep the aspect ratio the same. I

then click done once I'm happy with that. And because of the reduced

that. And because of the reduced quality, I'm going to actually enlarge it to cover the entire canvas. I'm going

to come down to upscale and enhance.

Head over to upscale and upscale this image in order to smooth out some of the issues with the resolution since we're cropping into the image. And by

repeating that process, we're able to zoom in, zoom in, and create various levels of zoom for the different shots of our animation. But then I want to try something a little bit trickier, which

was to take a frame from our text to video clip and just crop in and add our monk into the doorway as a way of sort of transitioning from the outside of the temple to the inside. I'll show you how

I managed to do that. Now, in order to get a screenshot from our video, I'm going to come down here and type in screenshot from video. Hit search. And

from here, any one of these will do the trick. I'm just going to pick the top

trick. I'm just going to pick the top one. Click to add your video. I've

one. Click to add your video. I've

renamed my video to one to sequence the order of the shots. I hit open.

Find the shot that I want to take maybe somewhere around here.

Hit start processing.

And the image will appear below. And I

can download it. From there, I have a new project. I drag the image of my

new project. I drag the image of my video in and I use the same tools to crop into that level of the doorway and then upscale it to improve the quality.

But also want to remove the man from the image. So I went to the AI editor, came

image. So I went to the AI editor, came up to local edit and of course just simply painted him out and had the AI remove him from the image. But on top of that, the doors looked a little bit funky. So I decided to go through just

funky. So I decided to go through just with a simple canvas eraser tool and erase the areas of the doors to have that area of the layer transparent and bring the original image through. Even

though it's lower resolution, it just matches a little bit better with the doors. So now we got a shot of the front

doors. So now we got a shot of the front of the temple. We remove the old man and we can see the doorway. But the next thing we have to do is actually get the monk to fit into the doorway as part of this transitional shot. So that way it looks like we're going from the outside

and moving in. And so the first step is again to use the generative expand tool to add in the extra information around the image so it looks like something we might see from the outside looking

through the doorway. I then download this image and bring it in to the other project of our temple door and just kind of size it into position.

Now, it's not going to be 100% the right shape and it's going to be a little bit too bright, but that's okay. We've got a little bit of a solution. I think that this is lined up pretty well. Although,

I'm going to have to sort of cut off and as a bit of reference, I'm going to need to cut off around where the uh bottom of the that sort of top row of the pillar starts. So going to come up to crop,

starts. So going to come up to crop, bring it down, hit done, and I can hold in shift and kind of warp it into place if I need to. Otherwise, that looks pretty good. It's just a little bit too

pretty good. It's just a little bit too bright. So I'm going to do is click on

bright. So I'm going to do is click on it, go up to my layers, and where it says opacity, I'm actually going to bring it down a little bit to darken the room so it does look a little bit more natural. And from there, I now have this

natural. And from there, I now have this image, which is part of the original video we had, and also our monk. So I

can use that as a starting frame for one of our videos to sort of go between the outside and the inside of the temple.

And then I continue to reuse the methods I've just covered to create more of the shots that I need for my animation, just the images for now before I go ahead and use them as start frames or even in some cases end frames for my videos. But I

also need to introduce another character. So again, I use the same

character. So again, I use the same methods of editing and I use the same consistent character. But sometimes I

consistent character. But sometimes I need both characters in one scene. So

I'm going to cover a few more techniques to make that happen. What I'm going to do is head over to character here.

And because I have an image with a character already in, I'm going to use the insert character into images tool. I

choose my character, which is my Pixar Samurai.

Then I describe how I want him to be in the scene. So I'm going to say we see

the scene. So I'm going to say we see the back of Pixar Samurai to the right of the frame looking at the meditating Shaolin monk. Now I need to also just

Shaolin monk. Now I need to also just get my lasso tool and I'm going to draw a nice big section here to the right to add this guy in. I then hit generate.

And from there I can fix up the little issues by going in to my local editor and I can add the beard back in or make some finer adjustments if I think there's anything there that needs to be fixed up. And now for the rest of the

fixed up. And now for the rest of the images in this project, it's rinse and repeat. I go in, I generate images.

repeat. I go in, I generate images.

Sometimes I will generate elements. I'll

take sections of the background from certain images, add them in, piece in various different elements or even edit things like facial expressions. It's a

simple case of using the tools to fine-tune the images you need and simply plan them out and go through one by one and create them until you get some images much like what I'm showing on screen right now. Because once we have those images, it's very easy to go into

those projects and simply use them as start frames for the videos we need to generate for our animation. Now, the

reason we go through and do this and cradle these images is because we know exactly where we stand when the animation starts when using these images as a start frame. We have a lot more control over various elements of our

images. This means we get to really nail

images. This means we get to really nail things down, use up less credits, use up less processing time, and really get what we want and what we're looking for.

So, we can start to go into the next step, which is want to start converting them into videos so we can start to piece together our animation. Now,

there's a few different methods I've used to create the shots in this video, and I'm going to cover some of those here. Not all of them because there's

here. Not all of them because there's too many shots. But to give you an idea, I have this shot here, and is the only line of dialogue this character has. So,

rather than using something like say lip sync, I'm going to head to AI video.

I'm going to switch from Google V3.1 to something like Sora 2. Take a quick look at the settings.

I only need 4 seconds for this shot.

It's not a very long shot. And because

some of the shots that establish the the actual video were a little bit more on the pricey side. For some of these easier shots, I'm using a uh a simpler model, but I'm going to hit the start

frame, add the shot, and then I'm going to add in my prompt, which is So, I have the boy looks down at a slice of pizza, smiles, and says pizza. I have the sound turned on. We're using Sora 2, and we

turned on. We're using Sora 2, and we got 4 seconds. I hit generate, and then we get >> pizza. There is an odd pause at the end

>> pizza. There is an odd pause at the end after he says pizza, but I believe I can cut it pretty early and work with that shot. So, we've got our line of dialogue

shot. So, we've got our line of dialogue for the Shaolin monk, but our other character, the samurai, has more than one line of dialogue, and I want his voice to be consistent from shot to shot. So, simply using Sora or V3 may

shot. So, simply using Sora or V3 may not be the best option because it might change the voice from each shot. So,

instead, what I'm going to do is we're going to turn to lip sync. And with

lip-sync, we can take our images and just simply apply an AI generated voice to them and it will be consistent from shot to shot. But while we're on the topic of lip sync, this time I'm just going to take a straight up image

because Design AI's lip sync for image is actually an incredibly powerful tool and does a really good job. So I come down here, I've got my image selected. I

come down to lip sync, pick a face image, and choose this image here. And

I'm only looking to add dialogue for this character because he has a few lines. I need it to be consistent. So I

lines. I need it to be consistent. So I

click next and choose 16 to9. Next

again. And of course I want to go 1080p.

And I'm going to pick my voice. Now I

need to find a voice I think is going to work across the board. The James voice is actually not too bad. So I hit play to test it.

>> I must not fear.

>> However, maybe that's a little bit too mature. So again, we can try Finn.

mature. So again, we can try Finn.

>> Hey, what's up guys? This is Finn. I

think that's not too bad. Let's select

Finn and we're going to add in our dialogue. So, hey, let's go. Generate

dialogue. So, hey, let's go. Generate

audio.

>> Hey, let's go.

>> Now, I do need to change the dialogue.

It's not quite reading correctly. So,

instead of having comma, I'm going to have an exclamation mark, and I'm going to separate it into two sentences. So,

now that should change the delivery of the dialogue. I regenerate. Hey, let's

the dialogue. I regenerate. Hey, let's

go.

>> That is perfect. So, for the rest of this process, I'm going to run with Finn for our samurai character. Click apply.

And I'm going to give it a little bit of space by moving it across.

And so, now we have this clip here.

Maybe even bring it over a bit more. And

now I hit generate to see how that turns out. Now, I want to just quickly mention

out. Now, I want to just quickly mention that I actually didn't end up using that shot. I actually wanted to remove the

shot. I actually wanted to remove the monk because I actually had that dialogue mixed up with the wrong shot.

But following that exact same process, I was able to create this shot with the same dialogue.

>> Let's go.

>> So, I think that does the job. I

actually removed the hey from the start cuz I didn't think it flowed that well.

So, I just left it at let's go. And but

there also is that issue of what happens if I have a little bit of movement in the video as well as a lip sync. Well,

we can also add lip sync to existing videos. So, I'm going to show you how

videos. So, I'm going to show you how you can do that as well. Now, when you have an image like this one where there's actually a little bit of action that goes with the speech, what I'm going to do is generate a video first

and then add the lip sync because I need his voice to be consistent. So, I'm

going to come down here to AI video, add in a simple prompt, and I make sure I mention what I want to see, but also that the bald monk doesn't open his eyes. He keeps his eyes closed. Also,

eyes. He keeps his eyes closed. Also,

need to make sure I add my start frame.

Hit generate. Then, we get this video here with just the action but no speech.

So, I'll come over and choose lip sync.

And then I go through the same process as we did earlier with the image, but apply it to our video. And what we get is, hey, and the video doesn't quite do exactly what I want to, but it gets the

job done. And again, I can cut into a

job done. And again, I can cut into a certain point of that video to get what I want out of it. So, it is a good idea to retry some of these videos in order to get what it is you're after.

Sometimes it's going to take a few attempts, but sometimes it's a matter of looking at the video and thinking, what area of this could I use? And if there's some errors, is it worth cutting them out? So, you need to think not just

out? So, you need to think not just about the video you're creating, but how you can use it and how it will fit into the flow of the animation because there's always going to be that big chance of something weird happening with your video. But, uh, with this one, the

your video. But, uh, with this one, the way he rubs his shoulder a bit, I don't really want that. But, the little bit of movement around the speech, I think, works quite well. So, considering that will help you in this whole process.

Now, just while you're working, one thing that may help is if you have a lot of different projects, I only have these four here on this particular account, is you can also create a new folder. So, if

I click on that, I can give it a name.

And from there, what I can do is simply drag these into the folder.

And that way I don't have to contend with the rest of my projects. I can just come straight in and they're all in there and I can continue building from here. And so you get the idea. We're

here. And so you get the idea. We're

able to use text to video for some pretty fresh shots that don't have characters in them. But when we have our characters, we've already got our images made. All we need to do is go into image

made. All we need to do is go into image to video. Simply add our image in as a

to video. Simply add our image in as a start frame. top up our prompt and

start frame. top up our prompt and create the video from there. Or if we need the dialogue, we can use the image to create a lip-s sync video by going into lip sync, adding in an AI generated voice, and it will generate lip-s sync

and also animate the rest of the image to produce that dialogue. But like I said, if you need some movement in there, you might want to generate a video first and then add the lip-s sync to the video. But one thing I want to

mention, too, is that there are some occasions here and there where you may want to use a start and an end frame.

For example, I have this shot of the Shaolin's eyes closed and this shot of the Shaolin's eyes open. So, what I've done is I've used a start and an end frame in my video prompt. I mentioned

that I want the eyes to open and that's how I was able to produce this shot for the actual video. So, it's a case of determining which technique is best for which shot based on the image you're

able to generate. And it's just a matter of going through and generating all those videos over time. But something

funny happened with one of the videos I uh generated with Sora when I wanted to create an action shot is it actually added in some dialogue. So if you see this >> look at these moves and a jump can't stop the rhythm.

>> Now I really liked the dialogue on that and when I played it alongside some of the lip sync it actually sounded not too bad. It kind of blended in enough that I

bad. It kind of blended in enough that I thought I could probably use it. So

sometimes you might want to accept a few happy accidents when it comes to using AI because it is a little bit of a gamble. You kind of roll the dice a

gamble. You kind of roll the dice a little bit whenever you're creating something with this technology. But the

next step is to edit all these clips together. Because once we have created

together. Because once we have created our storyboard and converted those into images and then made those images into videos, we now have all these separate clips. And what I do is I look at that

clips. And what I do is I look at that storyboard for reference and I number all my clips in order that I want to pop them into the editor. This is a pretty basic way of working. It's probably not the way the professionals work, but I

pop them into my editor timeline one after the other. And then I go through and I make sure I trim out the bits and pieces at the start or the end of each clip that I don't want because I need it to be a nice sort of rhythm. I need it

to cut nicely so there's not really weird pauses in there. But you can use pretty much any editor for this. I'm

using Camtasia here, but you can go and use something free like Cap Cut to sort of piece these all together. Now, a few other little tricks is you can add in some ambient noise into the background,

which I've gotten some from the YouTube audio library. I've added that in as a

audio library. I've added that in as a background audio track in order to just create something that kind of ties it all together. And [snorts] sometimes you

all together. And [snorts] sometimes you get a few little weird sound effects that are missing. So again, the audio library has some sound effects in there that you can download and use in your videos. And finally, I did add a little

videos. And finally, I did add a little bit of music, had a few little extra sort of uh issues that I ran into, and I'll show you what I did to fix those up. The cool thing about timelines is I

up. The cool thing about timelines is I have this guy jumping around here making a lot of noise. And what I'm actually going to do is I'm going to take one of these still shots. I'm going to copy it and just paste it above because I kind

of want him to keep making that noise while we're focusing on our uh monk character. And it creates a nice

character. And it creates a nice interesting effect where we come in.

Now, in this particular shot, we can see our character's eyes are open. So, one

thing I want to do is I'm actually just going to bring it up.

I'm going to enlarge it a little bit more just to kind of crop those eyes out. Sometimes you have to do these

out. Sometimes you have to do these little workarounds if you're not quite getting the results you want. If you can find a way to work around it. Another

one is when he opens his eyes in this shot here. So again, I can go back to

shot here. So again, I can go back to the start of the video where his eyes are shut. I can create a screenshot of

are shut. I can create a screenshot of an image. So now nothing's moving, but

an image. So now nothing's moving, but this side of the shot was perfectly fine. The camera doesn't move. So, I can

fine. The camera doesn't move. So, I can hold down alt in Camtasia to crop a section of the video and keep him perfectly still. So, as long as I line

perfectly still. So, as long as I line it up down here on my timeline, I can now have him stay still.

>> Hey, come on.

>> And overcome that problem pretty easily.

And so, to summarize that, again, you want to pick an editor, something like this, or even Cap Cut, which is free.

Put all the clips in there, trim them up, add any sound effects, review everything. if there's a way to fix up

everything. if there's a way to fix up certain clips or maybe you have to simply regenerate them, you just want to go through and build that narrative the way you design in your storyboard. And

of course, you can always review and change things as you go if you feel like they're going to work out better whilst you're making these edits. So, this is a whole different skill in itself, but this is a pretty basic way of looking at it, just lining things up one after

another. It's uh pretty easy to do when

another. It's uh pretty easy to do when you've got a whole bunch of these clips generated with AI. But the final video is now ready and I'll show you how it turned out. So you can see exactly what

turned out. So you can see exactly what we can do with AI in a little over a day by producing and using this process to create an animation.

Hey, let's go.

[cheering] >> Wake up.

Look at these moves. Ha! And a jump.

Can't stop the rhythm.

>> Hey, come on.

pizza.

>> Yeah, it's time for dinner.

[music] And that's the final result. Now, I

might be a guy who makes YouTube videos for a living, but I'm definitely not a filmmaker, but this was a ton of fun to experiment with and uh just a few different problems uh trying to solve little issues, but it was all pretty

easy. The power of AI when you have

easy. The power of AI when you have things planned out, especially when you have things like Sora V3, where you can generate videos with sound effects, and you can probably see which ones in this clip were actually made with sound

effects using Sora and V3. It's just

really easy to cradle this stuff, use consistent characters to piece it together, and even some of the shots of the monk's face were just simple still images that I use the zoom function on my video editor to achieve. So, you

don't even need to convert it all to video. Sometimes still shots can be used

video. Sometimes still shots can be used quite effectively. But overall, it comes

quite effectively. But overall, it comes down to practice, problem solving, and just having a bit of fun playing with this stuff and just working out what you think is possible and putting it into action. So, I hope you enjoyed this

action. So, I hope you enjoyed this video. I hope you enjoyed the animation.

video. I hope you enjoyed the animation.

And don't forget to check out Design AI.

There's a link in the description if you would like to create something much like this yourself. And if you do, please

this yourself. And if you do, please leave a comment below with a link to the video so we can check it out and see what you've created. Otherwise, that's

the video for today, guys. I'm going to pop up some tutorials on the screen here which may help you with some of the more in-depth processes within this video.

There's also [music] some tutorials linked to in the description below. But

otherwise, that's the video for today.

Thanks for watching. Hope to see you again soon and have a great day.