Google Veo 3 Masterclass: Create Cinematic AI Videos from Prompts

1. About Course and Google VEO 3: The Welcome to Google Veo three AI Video Creation master class where you will get the expert knowledge of the world's leading generative AI video creation tool. My name is Sad and I'll be leading you throughout this course. So as with all of our trainings, we will follow a practical project based learning. That means we won't just learn the AI tools theoretically, but we will work on real projects and create real viral videos, practically, and I'm going to guide you step by step how to create those videos. So whether it's a product ad, any viral video idea, creating realistic AI characters or even using your own avatar to create awesome AI videos. This course has everything you need. Google Veo three is here, and AI generated videos are becoming more realistic every day. From realistic interviews and conversations to cinepatic storytelling, creating any kind of UGC content, product videos, advertisement, or completely imaginary scenes. You can bring almost any idea to life with the right. Whether you want to create cinematic stories, create any kind of viral social media content, create any kind of UGC ads, or you just want to publish viral content on the social media platforms. This course gives you the practical tools to convert your imagination into reality. So let's get started and discover what's possible with Google Veo three. This is the best VEO course ever. Come on. We will take you to it. 2. Your First AI Video in Minutes: Video, we are going to see how we can quickly make the realistic viral videos for our YouTube content for our channel or for whatever brand representation we want. And we want to take an inspiration here because the best way to start working on AI is to see what other people are doing. And first of all, just try to recreate that. If you're able to recreate that specific video with the same concept, then you can apply your own concepts as well. So this is a quick example, channel that is impossible tales. You can search it on YouTube. And it's about EI made animals that talk and reply them in the interview. Look at this video. Oh, in the world? A leaky trash can behind. So as you can see, it is so amazing. The videos are very realistic. Whats weakness as a predator. I have great startup speed, but terrible. And just look at the detailing, guys. Let's see how we can create this. First of all, whatever you want to create, let's say that we want to make the cheetah speak. So what we do is, first of all, take the screenshot of that. So let's use a snipping tool for this. Just make sure to get a good shot of the reals. And let's actually select this screen right here. You can select the whole frame as well. And let me just save it somewhere. Okay, so I have Save it, and once that is done, you will go and work on a new tool. Besides VEO three, that is WISC. Now, WISC is the platform that is offered by Google Labs, and it's a very effective tool as well. So we just scroll down and it says Wisk labs dot Google FX. We click on that, and basically, you can simply sign in with Google. So click on sign in with Google and get started with Flow. If you want to receive some marketing emails, you can continue with that. Like next, scroll all the way down to the bottom, and then you should be able to click on the Continue button. And as you can see, it has some amazing video creation. Tools here. So let's create a new project. Click on Get Started. And once we are on this main screen, you might see a directly upload button or the start button itself, and it will look just like Tab JBD, but we will take the separate support of hat JBT as well to define a complete prompt and what we want the cheetah to do. Because if we define it ourselves, it might make mistakes. So instead, what we did is we just uploaded the screenshot. And we asked define the scene and write a prompt for this because I want the cheetah to say join my AI course now. And as you can see, it has defined a complete scene definition with some prompts. So whatever prompt you want to use, you can use a shorter version of it, voice expressions only or a complete DDL prompt. I will use this one. Let's click on Start, click on Upload Image, and we will upload the screenshot that we have saved earlier. We paste this. It contains everything. It defines the complete shot, and it says, Join my AI course now. Limited time. Offer. And we click on the video, 16 by nine is fine. And if you want to fast produce this, you click on X two, actually keep it on it, and the fast is fine. It will use a little bit more credits, but that is okay. Click on Generate, and let's wait for it. And it has generated two results, basically. So XT is not only to produce it faster, but it also generate two samples of it. Let's just listen to it. Join my AI course now limited time offer. So pretty awesome, right? And let's look at the second one Join my AI course now. Limited time offer, isn't it? It's amazing. So this is how you can create powerful, engaging videos to whatever scene you want using VO three and whisk and tools like that. Estry this by yourself, and I will see you in the next one. 3. Choosing a Google Flow Plan: Guys, before moving forward, if you want to continue with a free plan, you can. But since I want to show you a lot of examples, I'm going to show you in this video how we can buy the subscription. As you can see, this little message appeared that you need more EI credits to complete this request. So we click on Upgrade because it is every penny worth it to get this subscription and learn this properly. So let's actually continue with the basic package per month. Click except, and let's continue with the card, click on subscribe. Mm hmm. And now, as you can see, it is working because we have the plan purchased. So this is how you can purchase a plan, the basic one in Vu three. 4. Ultra-Realistic AI Characters: Video, we are going to see how we can create a unique, realistic looking AI character just like this person. Yes, this person is completely AI. It is completely generated by AI using a simple command. But let me show you how we can make those characters that looks ultra realistic. First of all, let's see what people are doing. If we go to YouTube. We just first of all, search for the AI influencers. And you will see a lot of tutorials here, how we can create those influencers. We are going to see all of them in detail. Don't worry about that. We will see and learn every single thing, because this is making a lot of money nowadays. But I'll tell you an other way of doing this so that it will be even more realistic than this one. We do is we search for a real interview. Let's say street interview, I'll just write that in YouTube. And this is an interview with learn English with street interviews, simply street English. So we just open that and we take any character for the sake of example that we can use as an inspiration. We will not use those characters and make them speak because that would be illegal. But we will take the inspiration of the scene, the sunlight, the real looking footage, and the background and the style and all. We can take the inspiration of that. So we just use any character, and we take the screenshot of that person. We will not use that person. We'll just take the sample to create our unique AI character. We just take a sample like that, save it to the location, and then we have these samples from the real persons. Which we'll use as an inspiration. Now, what we do is we go to Shad DBD and we just paste this character, and I just spoke it here. Let me just do it again. I'll just upload this sample interview, and simply you can dictate it. You don't need to write anything. Just use this image and create a totally separate EI character person that should be different from this person, but the scene should be very realistic looking. The sunlight should be in the same way, but the background should be different. Just make a complete unique character, but the scene should look ultra realistic as this one. Create a full prompt on how we can create that. Actually, let's change to create a full prompt for this. Otherwise, it will start explaining that. So it will analyze the image. Create a unique person. As you can see, this is a separate person. It is totally AI created. You won't be able to judge from any view that it is AI created because the sunlight, everything, the background looks ultra realistic. And this one is even better if you want to use an Asian or Indian looking guy. So let's use this one actually. Download that and let's use this person and make him speak now. And we will just speak the prompt now. Create a full prompt for this person to speak with the interviewer, and the interviewer asks that you have recently completed this AI course with accounting training and solutions. How was your experience? And he just says that experience was amazing. I couldn't believe I could create that many EI videos which are realistic looking, and you know the best part, I'm AI created as well, and he laughs. And this is the end of the scene. So that's how easy it is to define any situation. Obviously, you need to tweak it a little bit after that. I repeated this two times, so I just remove that. Now, to create a fully functional prompt that when we can use in flow, we need to send this prompt so that the chat jibe can define it. It started creating the image, write the prompt. I'll just ask again. And now it writes a complete prompt about it. With the style requirements and all, we just copy this and we create a new project in flow, click on start, click on upload the image. We'll use our sample character for it. And now we just open that. We have already uploaded the image right. So we use the image here, make sure to attach the image. And then we copy this complete prompt, paste it here, and to use lesser credit, we need a separate setting. We'll use video. We'll use frames. For the images, it almost use no credits as you can see. But for videos, it uses 20 credits or whatever you define, it will define it like that. So you might need to tweak it a little bit in order to reduce the credits. So if you want the light version, it will only use ten credits. So let's try this now. So light will obviously need some time and the quality will not be that good. And if we generate two X, it will generate two results for the videos. So let's only generate with one results. It will only use ten credits. So if you want to use your basic subscription for a longer time, you can use and try out these settings. Let's try to generate it with the basic setings. It might take some time to generate it, so just wait for it to complete. And here is our video. Let's try to play it. So you recently completed the AI course with account tech training and Solutions. How was your experience? The experience was amazing, honestly, eigens So Okay, it flipped it around. But the footage was very nice. Let me just customize it. Okay, the interviewer is at the back end, and the person that is shown in the image is actually the person who is answering to the interviewer. So he should say that the experience was amazing. I can't believe I created many AI videos by just learning this course and you know the best part. I'm AI created, as well. So just recreate this. As we can see, it made a mistake, but that's not a problem that happens sometimes. And let's look at this first. The interview is not visible on the screen and stands behind the camera holding a small handheld. The interview asks naturally off the camera. How was your experience? And the young wang looking towards the camera naturally, now this should work fine. So let's try it again. Click on Start, use this image, and we paste this. Let's just generate again, make sure the credits are only used to the minimum. That is ten creats only. And let's look at this one. The experience was amazing. Honestly, I can't believe I created so many AI videos just by learning this course. And you know the best part, I'm AI created, as well. The experience was amazing. And now you can see the results, guys. It is so amazing. It used just the minimum credits, and no one can believe that it is AI generated. The experience was amazing. Honestly, I can't believe I created so many AI videos just by learning this course. And you know the best part, I'm AI created, as well. The experience so this is what we are going to learn here because if you look at the AI influencers and all, although it is very good, but this does not look as realistic as I have shown you because that is the best technique on how we can do it nowadays to create the ultra realistic looking videos. So just try this by yourself, create those unique AI influences just by taking the real inspiration, recreate those persons, and you just create a realistic looking video out of that. So just try this by yourself, and I will see you in the next one. 5. Using Veo 3 in Flow and Gemini: This video, we are going to see how we can access the VEO three model in different ways. Let's say that we want to try it with Gemini, try it with Flow or build with VEO itself. Here's how you can do that. First of all, we head over to the Google and we just write here Deep Mind Google Veo three. And it will show you the official website of Google Deep Mind. You click on that, and this is basically one stop shop where you can access any interface to access VEO. So if you click on Try and Flow, you click on Create with and you land on the same area where we have de created one. Let's close out of here. Let's now click on Try in Gemini, and you can pick a template to create. Let's say I want to create a mammo, dinosaur vlog, T rex saying, Hey, all. I just want to let you know I love VEO three, something like that. And we can select the pro model. You can select Google Ultra, or let's actually use this fast model, whatever is selected. And make sure down here the create video is selected. So let's generate and the third option is build with Veo. So if you click on that, it's all about how you can generate that. You can use it in Python and you can use different styles, but that is a bit complex. So instead, we can keep it to flow or we can use it in Gemini. So some people prefer gemini, some people prefer to create in flow, but both of them are perfect. So hopefully, our video is created. Let's view it. So here's our video, dinosaur log. Hey, A, I just want to let you know I Louv VEO three. Okay, it changed it to Louv. We can edit this prompt. We can make it I Love VEO three. And immediately it will create a new video for you or create a prompt and then create a video for you. So as you can see, it's generating an updated video now. But you might have seen that if we define a prompt in a detailed way like extracting it from Chat GBD, it generates much better results, much better details, especially if you take an inspiration, it generates much, much better results. So what we can do is, let's say that you want to generate an AI influencer, you just go to the Pinterest and after going to the Pinterest, we just write here model. You will see a lot of models here. We can use any one of them and just recreate those models to create our own unique character. Or you can say athlete and here are different photos of them, and you can recreate your own character. That will be totally unique and you can use that whenever, wherever you want without getting any right because the AI or VEO three will only take the inspiration from what style and what physique and what kind of person it should be. And then it will create a totally unique character that you can use in your videos. So that's a part for later on, we have a lot of things to explore. But meanwhile, our video is ready. Let's see this. All. I just want to let you know I love VEO three. And still it says VEO three, but you can improve these things. I'm going to show you how to pronounce the words in a better way just by using a simple trick. And now this time, let's take the support of Chat GPT. And here, let's extract a detailed prompt. This time we are not using any image for the inspiration. Not providing any sample image. We want the AI or VEO three to totally create it from scratch. And the best way to define this is to just dictate. You just click on this and say, Hi, just make a complete prompt of a person, an old man sitting at a restaurant and saying, Enjoy your life before it's too late. I enjoyed my life a lot, and I can still dance at this age. And he laughs and dance a little bit, and the video ends. So whatever comes to your mind, you just create a video out of it. Actually, let's not change anything. You just send a prompt and let's see how it creates this prompt. We'll just copy it simply. And this time we will paste it in Gemini, click on a new chat. Paste this. In the tools. We click on Creative Video. Fast is fine. Click on Submit, and let's wait for it to complete. And our video is ready. Let's play it. Enjoy your life before it is too late. I enjoyed my life a lot. And I can still dance at this age. Enjoy your life. This is amazing, guys. Let's download it and let's view it in full. Now, let's view it. Enjoy your life before it is too late. I enjoyed my life a lot, and I can still dance at this age. So this is how you can convert any idea into video. Whatever comes to your mind, you can simply create an awesome video out of it. So this is how you can either use Gemini or Flow to create your AI videos. Just try this by yourself, and I will see you in the next one. 6. Creating Unique AI Characters: Sudo, we are going to see how we can create our AI character directly in flow. So instead of a GPT, this time, we will use nano Banana to create it to create our unique AI character, and we will do the same process. We'll just upload our reference image. Let's say this one. But unfortunately, here, we cannot dictate it. But if we want to generate a character, we can select image and we can select 16 by nine. We can select four by three. We can use nanobnaP. We can use nanobnana two, image in four. So let's use nano banana Pro. Even if we generate two results or three results, it will not use any credits and that is an amazing part. Let's compare it with hat GBD as well. Unique AI character. So obviously, it is a little difficult to create the character by defining it by writing everything. But I've just wrote some of the commands, create a unique AI character that's based on these looks, but different than this one, it should be completely unique so that I can use it as my AI character anywhere. I should look ultra realistic. Let's see what Google Flow has to offer. We can see that it has created two unique AI characters. But if you compare it to hat JBT it looks a little bit. You can say like AI character, not that much, but it's your choice if you want to generate it with Flow, or you can generate it with hat JBT. But the good thing is you can generate it with absolutely no credits used at all in Google Flow. So if you don't have a Pro access to hat JBT, it might limit you to upload the images. I'm using the Pro version. That's why I can upload multiple images. But if you upload like two to three image, it will tell you that the limit has been reached for today. So instead, if you have only bought one subscription for Google Flow, you can use it for all purpose to generate the image and then let this person speak by using a simple prompt. So that's about it. This is how you can generate the AI character images with nano Banana in Google Flow. Just try this by yourself, and I will see you in the next. 7. Getting Ideas from Flow TV: This video, we are going to explore the Flow TV and it's amazing use case. If you click on this, you will see all the amazing stuff that people have created from Flow and you can get a complete prompt for it. So if you just change this, you will see all the amazing things that people have created. So this is a crash test, all these amazing things that people have tried. So if you want to use any of these prompts, what you will do is you just select that and you click on Create with Flow, Tick on the new project, and you simply paste it here. And we can change it according to our needs a continuous ultra white overhead shot walking on a world that is made up of switches in spheres of lights on the ground. Let's make it soft lava on the ground, and we can refine it using chat GBT. Let me just refine it, refine it for soft lava seen. Let's copy it. I just want to make it quick because we already have explored these functionalities. So click on the video, we generate a 16 by nine, and to keep it to the minimum, we will use One x and view light. It might take some time, but I'll just fast forward the process. So no worries. Just to see that if you're not getting much ideas, you can always check the flow TV and you can refine your way of thinking what people are generating and with prompts. Now, let's explore this. And this video is amazing. Obviously, we haven't changed it to lights. That's why it's blowing up lights like after each step. But look at the footage, how realistic it looks. And this is how you can use different inspirations or even use the same footage as this one. You have a lot of options here. You can check the ASMR videos and so on, if you want to search for ESMR video, you can search for that as well. As you can see, this is an ASMR video. Look how amazing it is. So if you want to generate an ASMR videos like this, you can just use a prompt and refine it for your idea and just paste it and generate a new unique video. So that's about Flow TV. Just try this by yourself and I will see you in the next one. 8. Getting Free Veo 3 Credits: In this video, we are going to see how we can get the Google Veo 3.1 ultra credits for free because right now, whatever we are working on is whether it's Flow, whether it's Gemini, it's basically Google Veo three, and we already have a pro subscription. But I'm going to show you that if you can't afford it right now or you just want it for a hobby or just a practice, here's a quick solution for you in which you can get it for free. So let's actually try it on a new ID. I've just log in with a new Google ID, and here we just write in the Google search, Bunny flow. Hit Enter. And here is official link of Bunyflow That is bonyflow.com. Click on that, and you will lend to this section which says AI videos Unlimited creation zero hassle. So we click on Get Started on the top right hand corner. Okay, so here we have to just enter the user name, email and password. Now, if you want to use it for unlimited time period, what we'll use is we will use a TEM Mail for it. So basically TEM Mail is disposable email address, which will allow you to create a quick email ID for you that can receive and sign up for the platform. So if we mention my name. And in the email ID, I mentioned that. We can use a simple password and we click on Create Account. It says email or temporary email is not permitted, so we have to use our official email address. That is the email ID. So we click on Create Account, and basically, we have to create a unique username. We resend the code. And let's use that verification code that we have received. Click on Verify email. And now we called the excess. And that's not all about it. Just watch carefully till the end, otherwise, you will skip some important updates. Now, as you can see that I have got an unlimited plan for free. This is unlimited for these many days. But after the completion of these days, we can create a new account or register with a new account to use it again. But since this is an unlimited credits available, you can use it freely and you can bring your creativity to life. Otherwise, when you have limited credits, you always have the stress of using all the credits. So that will be a huge time and money saver for you. Okay, after that, we click on the flow excess on the left hand corner, and here we will write any simple prompt. Let's try it with the hat GPT. I'll just mention a simple scene, copy it here. The temple is not working, so that's fine. We past the prompt here, and you can see ultra is written everywhere, but we'll use and keep the settings, whatever is selected, we will not change anything. And at the bottom, we will see open on Google Flow. So click on that and we click on Create with Flow, click on next, scroll all the way down to the bottom, click on Continue. And as you can see, we have caught 150 AI credits, but this credits refresh daily, so you don't have to worry about this. And let's try generating this prom now. And because we get the credits refresh daily, so we don't have to worry about anything. And you might already know the least consumption method as well. Image is for free. It uses zero credits, but if you want to generate a video, you keep all the settings as it is, but right now it's using 40 credits, and that is too much. So I'll choose one X or generating only one sample video and turn it to light, so it will only use ten credits. That's it. Click on next and it will create your video in minutes by using the free credits on Google Flow. And here is our video. Let's play. And look at the quality of this guise. And if you want to download, you can download it into any size. We need to upgrade the plan to get it to 1080. But otherwise, the 720 is also fine. It will immediately download it for you. We can play. And that's how it looks, ultra realistic, right? So this is the best way to use Google Flow or VEO three absolutely for free, using a free trial plan. We still have 140 AA credits. Now, let me just show you that in BnniFlow why you're not able to use the Ultra plan? That's because we need to click on the extension on the left hand side. From here, if you're using Google Chrome, you can click on Download the extension. If you're not, you need to download and open it with Google Chrome first. So we click on the download extension, and it will be downloaded right here. Now, the extension is downloaded, but it is in this zip file. So we click on the three dots on the top right hand corner and click on the settings at the end. And on the left hand side, we will see extensions option. Click on that and we need to enable the developer mode in order to add a new extension. And since this is a packed extension, that means it is a zip folder. So we'll click on packed extensions, click on Browse. You choose the location, or what you can do is just simply drag and drop it here. That will do the job as well. Okay, now after that, once the extension is installed, we click on the spin bars so the extension shows outside in the URL bar. And from here, we just simply need to log in with the same email ID. So if you're unsure about what is your password, what you need to do is just click on the profile icon on the left hand side, and we click on the account settings. From here, you can select your email, paste it here, and here you can see your password as well. It will suggest the password. We can look for that password and use it here. So once we sign in, now we are connected to the free plan. So if you click on overview now and we click on the Flow xs, just make sure to keep all the settings as it is. If it is set to nano banana, you will keep it to nano banana. You will set it to lower priority in all of these settings, and just write a symbol prompt here. Click on open with Google Flow, and you will already notice that you can see this ultra option here, but it will be great out. That doesn't mean it's not activated. That means it is activated with a limited excess. So basically, you can use this but in the lower priority mode. So if you click on, let's say, the light version, you won't be able to use it. But if we choose this lower priority method, we can use it. This will use zero credits, it says. So if we compare it to our original plan, we click on the new project. Here we will notice that we don't have all of that options. So zero credit is definitely the thing that you would want to avail. We also do have fast lower priority methods, so you can try any one of them. Let's try to generate it. And it doesn't matter even if you use or generate four videos, it will still generate it without using a single credit. Sometimes it crashes, so just don't worry about that. You might need to try it a little bit. But if let's say it is showing some problem, you can try it with let's say two or even one video, 6 seconds. Let's tnrate it. Let me just click on New project. Lower priority method and all is fine. Let's set it to 8 seconds. Let's see if we can generate the maximum length of video. And it sometimes might say, you're requesting the generations too quickly, please wait a moment and try again. So that's fine. Let's wait for some time, and then we will generate it again. I've opened an other tab, and I've set it to View three ultra in high priority just to check with the KrankdgPmpt. Click on Open and Google Flow. So at first, it shows cue, and then it start generating it. So let's see if it can generate it or not. And you can see that we are able to create this for free. Let's try to play it. With each other. You go ahead. So this is perfect. If we try to download it, we can even download it in ten ETP. And that's the proof that we are using the ultra version. Obviously, you need to try it a couple of times because this is the free method of activating it. But as we can see, on the other hand, we also generated an other video of a dancing dog. Let's see this. And it's also of 8 seconds, so we can download that as well. Let's download it in the original size. And if you choose original size, it will be downloaded right away. And still very good quality. Upscaling is complete. Let's download it. And now it looks like this. Look at the quality of this video, guys. So this is how you can use this Ultra version, absolutely for free, but just keep one thing in mind. You won't be able to see the credits here. That's the only problem. And if you activate Ultra, you might need to wait for a little bit for generating some prompts. So if you prefer to use that 150 free AI credits, that will be generated instantly. You can use that or you can convert your package to Ultra. But in that case, you might get some error sometimes. So that's about it. Just try this by yourself, and I will see you in the next one. 9. Anatomy of a Perfect Prompt: In this video, we are going to see how we can write the proper prompts in order to get the best results from Google Veo three because as we already know, prompting is the most essential thing in generating good results from Chat GPT, from AI or from Veo three in general. So how do you define your character, your scene? It's very important in order to get the best results. So, for example, if we are defining the character, we should define that in detail. What is the gender of the character? What is the age? What is the ethnicity, facial features? You can write about the hairstyle type, body type, clothing, accessories, if any, expressions and emotions. So, for example, you can write a prompt, young South Asian man. So we have defined the age. South Asian is the ethnicity, gender, hairstyle, beard style, it's costume, the expressions and all, and then you define the scene description. Explain where the scene takes place. So if it is indoor in the outdoor environment, in a mall, in a cinema, in a concert, in a restaurant, wherever, so you can define something like this, luxury rooftop restaurants at night with city lights and soft rain in the background. So that defines the scene completely. Similarly, the character, what the character, what the person is doing? Is it walking, talking, dancing, sitting, fighting, whatever. So it defines like this. He walks towards the camera while adjusting his watch. So this is a clear description of what the character is actually doing. Similarly, you can define the camera direction. You can define the camera angle. Is it a close up shot, white shot, overhead shot, close focus shot, drawn shot or whatever. And that is very, very important in defining a cinematic scene if you want to use this Google Veo three professionally for your content designing. I can really create some movie like footages, if you define it properly, what to do. Camera movement is it handheld, it is tracking shot, zoom in or whatever. So it goes like this cinematic tracking shot slowly moving towards the character face. So that defines clearly how the camera is moving. Similarly, you can define the lighting. You can define the mood and emotion of the video. Is it a dark, horror comedy, whatever. So this defines it clearly. Video style is a cinematic documentary, Hollywood style action, and the environment details you can define any crowd, fire, wherever this scene is taking place, you can define that. The voice tone, you can define that as well. Dialect structure, you can define the dialogues. What is the tone? What is the emotion, the pausing style. Each and everything you can easily define in words. Will generate your character and scene accordingly. You can even define the special effects. How do you want the special effects to see a cinematic feel in your videos? If is there any negative prompting negative prompting means that you just guide AI what not to do. Don't create extra fingers, deform face, distorted eyes. That usually happens with AI portages. But nowadays, VEO three and all of these offers are very advanced. They don't do that mistakes, but you can include that negative prompting so that it avoids this cartoonist logs, overexposed lighting and so on. So you can write a complete sentence like this. A spect ratio, you can define how the footage should be like. Is it a ten ATP, two K, four K, and so on, and there is a complete formula. So if you want to generate best results, always define your footage like this. First of all, you have to define the character plus action, scene, camera, lighting, mood style, et cetera. So example of a complete prom structure here is how it goes. Let's say we are defining a businessman sitting at the rooftop and he's talking to someone softly. So it goes like this. A confident young man wearing a black luxury suit stands on the rooftop at night overlooking the city skyline. Of rain falls around him, while cinematic on lights reflections on the wet floor. You can read through all of this. Actually, I'll provide you this prompting guide completely with the resources section of this video. So this is how you do it. Now, the problem is, how would you define all of these scenes? Now, because it is so much of information, I have created you this project file, where you can define your character just by choosing from here. So character name is, let's say, Joseph. We can define, is it male, old male, simple, old woman, child boy, something like that. So age group is young adult ethnicity, you can select black body type you can select muscular hairstyle and all of these things. You can define it like that. Let me just quickly define all of these things. Okay, so I have defined these characters. Let's rendit a scene, actually. Now what we do is we simply go to Chat JBD, we paste it here, create a scene, prompt, based on the above, prompt characters for a man planning our revenge. Write dialogues. Accordingly, let's see the result. So it includes everything. Joseph, a muscular young man dressed in modern street wear, black hoodie, heavy bomber jacket, and the dialogue is they thought I was finished. They took everything from me. Now they're going to learn what happens when you push a man too far. Let's see that if we define it like this, what results do we get? I know it involved a little bit of theory, but that is a very important part. Even we are going to see some more prompting experiences where we will create some professional ads and so on. This is just a quick overview of how the prompting works. Let's actually use minimum credits. And for these kind of AI footages, where you want the flow or VEO three to completely create a new unique character, you don't have to upload any photo four reference. So as we can see, according to our character, just first of all, let's take the reference. It is a complete guide that you can get with the description of this video. So the more you see all of this, the more perfect you will become at prompting because you already know the essential ingredients of it, the real formula, how the prompting should work, because at the end, I have mentioned the prompt formula for the best results, as well. So this is one file and the other one that you get is this one where you can define the character name, the project name, and the expressions and all the details about them. And here is a drop down list. If you want to create a new dropdown list for the expressions like color tone, action, I've created all of them. But let's say for the location, I don't have any drop down list, you can create a separate dropdown list here as well. So let's say that location is indoor, outdoor street camp restaurant. Let's say this is all about the location, and I want to include that dropdown here. So what we will do is we will create this in the drop down list, and all we have to do is we select this complete list, and we mention at the corner here in the name boox that it is the location, and now you just have to attach it here. Location next to it, we will attach this drop down list. Now to attach this drop down list, we go to the cell, select it. We go to the data tab, click on the data validation, and we will allow the list. Now, what is the list? List is equal to location. Hit Enter, and you should be able to see this drop down list here. So you can create the separate drop down list for all the scenes that defines it. So according to our instructions, it has completely created a separate prompt in chat JBD. Now let's see how the results look like. They thought I was finished. They took everything from me. Okay. Since VEO three creates the scene by scene up till the maximum of seven to 8 seconds, let's see the timing of it. They thought I was finished. They took everything from me. So this scene is of maximum length. So if you want to extend this, we click on Extend. What happens next? Let's define what happens next. They took everything from me, and let's paste the remaining one because this time we are trying to actually create a cinematic scene. So let's generate it now. Okay, let's run this. They thought I was finished. They took everything from me. Now they're going to learn what happens when you push a man too far. Okay, here we can notice that although it extended the clip to 14 seconds, but it has changed the character. So we are going to see the character consistency later on. We can just take the sample of the first picture and then upload it and then extend it. So there are a lot of techniques in order to maintain the character consistency. We'll see that later on. But if, let's say, want to download this 14 seconds of clip, let's see if it is able to download this complete clip or it download it at separate Clips because we just extended the clip into two clips. So let's click on full Video and 720 P. Or otherwise, if you want to upscale both of them, you upscale it and you separately compile it. So let's generate a 720 P video, and this is a final video. They thought I was finished. They took everything from me. Now they're going to learn what happens when you push a man too far. Tonight, it ends. So, this means that we are able to extend the clip in a single video. We don't have to edit it by using a separate third party software, and that is awesome. Just we have to learn how to maintain the arrector consistency and that you will learn in the future videos. Now, let's try creating one with Flow. We click on New Project, and we straightaway, paste this based on the characters and the theme I just told you. First of all, whatever is your character, is it an Asian man or what is the ethnicity? What is the age, how and where in the scene, it's sitting, it's clothing, the background, and all I have defined it like this. Asian man in his 50s sitting on the worn out steps of an old building, a busy city street. He's homeless, dressed in torn, dirty clothes and all. People continuously walking, passing by him. And this is all busy world and everything. I've defined it clearly. And basically, if you want to define it that far, you can easily take the support of Chat JBT. You just speak through it. Considering all of the characters that I just mentioned, and it will design a complete prom for you. And the Asian man says, funny how they all listen to me once. Now, nobody even sees me. So this is a very emotional clip. Let's see if Google Veo three is able to generate and maintain this emotion as well in the real video. Let's see if Google Veo three picks up this emotion and generate a complete cinematic clip out of it. Okay, so these are our results. Looks very nice from the initial picture. Let's see, and listen to it. Funny how they all listen to me once. Now nobody even sees me. Funny how they all listen to me once. Now nobody even sees me. So this is very, very good. You can see the emotions, the clothes, how the people are passing by him. Funny how they all listen to me one. And it has completely picked up the right character. He's an Indian Asian guy in his 50s. Now nobody even sees me. M. Let's download it and let's actually click on the upscale to ten ATP because it will cost you nothing extra to upgrade this. Let's see the final results of it. And if you want to, meanwhile, get the prompt, whatever you have generated the results with, you will get it on your right. So if you just expand it like this from the right hand side, you can copy it all. What I'll do is I'll mention this with the resources section of this video as well so that you can use tweak this according to your needs and generate a different scene. That will be ultra realistic like this one. So this file, you will get it in the description of this video. And meanwhile, our upscaled project is now available. Let's download it and let's see the final results. Funny how they all listen to me once. Now nobody even sees me. This is perfect. Look at the quality, look at the detailing and everything, and we're just getting started with it. So this is how the perfect prompts works in Google Veo three. Just try this by yourself, and I will see you in the next one. 10. Text-to-Video Prompting: This video, we are going to see how we can use the Text-to-Video creation function in Google Veo three or Google Flow. As we have already explored how we can write a perfect prompt. Here's the first functionality wherewith. You can write a simple text as simple as a dog, a cat, or a beer, and it will create the footage for you. And then if we extend it further to define it with a proper prompt, it can generate wonderful results. So we are going to explore all the later functionalities like what is ingredients, what is frame by frame, videos, what and how you can extend the results of the video because as you already might know, that the first result that is created for any video is of 8 seconds in length. So we have generated these footages, if we open any one of them, Funny how they all listen to me once. Now nobody even sees me. If you check the length of this, it's almost of 7 seconds to 8 seconds of the clip. But let's say that you want to create your content and you want to add different frames to it, you can use the extend feature. You can use insert, remove, and camera. So these are all the functions that we will explore later on. But let's just explore how the text function works. Click on the new project. So first of all, if you want to generate the Text-to-Video results in all of the methods, you can generate it using the light version, which only uses ten credits in the pro version. We have this fast version which will produce the results faster and better, but it will cost you 20 grades like double. And if you produce quality results, which will use 100 credits plus. So if you're using this free version that I told you before, using bunny Flow and all, you can use the quality version because you still have the ultra plan and you don't have to worry about a lot of credits used. But otherwise, if you're just practicing it on a pro version, I always recommend you to choose the light version. So we can literally write anything to generate a text to video results. And the best part of this is you don't need any reference image for this. It will create it from scratch using the EI technology. So let's try to write a simple command, and let's use the minimum credits. We'll keep it to video, we will keep it to ingredients. And this time, let's say that we want to create a reel. So if you want to create a reel, you can switch from 16 by nine to nine by 16, which is a vertical video. One X means only one kind of video or one version of video will be generated. If you select two X, it will cost you more credits. It will cost you 20 credits, three X will cost you 30, four X will cost you 40. So let's actually generate only one image with a light version using the minimum credits, and let's see how it come up with a real. We generated and let's say that you want to be in 16 by nine, as well. So we quickly select 16 by nine and send it again. So it will generate two results together. Obviously, this will cost you a separate credit. But on our right, we can see that even though the footage is very nice and very realistic, it has generated a panda video. So that might happen when you just mentioned a couple of prompts. Or a very brief of it. And the second footage is completely changed. It's also very nice, but this one is off a beer. So this is the results generated with this very simple command. Let's say that if we go back and under the same project, we write an emotional bear or let's say, white, polar bear looking directly into the camera as if he is sad about the death of its family member. So let's say that we write a little bit of detailed text prompt just to see how it turns out to be. If we mention more text for the video creation using AI in VO three. Let's generate it, and you will notice that all of these are generated in the same project. Although I've made a spelling mistake in SAD, it should be SAD. Let's see the result. So as we can see, it's very realistic. Let's try to generate again just to see that if we correct this spelling, do we see any change result or not? Just remember, guys, unless you use a photo reference or you use little bit of tools that we are going to discuss later for image consistency, it might generate different results every single time, even if you're working in the same project. So the image and character consistency, we will see it later in the lectures. But as we can see, this one is fully changed from this one. Very realistic. You can see the tears in his eyes. And if you download it, you can simply upscale it to ten ATP. Click on that. It will take some time to upscale this. So we have to wait for some minutes. So my download is ready, and when we simply click on it, it will download the video for you. So very nice quality. That's how we do it. But let's say that we want to create an animation kind of character for this bear, how do we do it? But this time, you will see an excited polar bear in space, in an astronaut costume. Make sure the spellings are correct, okay? In outer space. Must be in pixel style animations. Let's try to say that. So what we'll do is we will just copy this prompt and paste it in chat GBT. Let me just show it you from the scratch. We paste it here, right, better prompt. Now it's writing a brighter prompt for it. Let's copy this and paste it right here. Although if you use that prompt, let me just show you the result. If we just keep the prompt as it is, it is generating pretty nice animation style pictures, so it should be similar in the flow or even better. But let's just try this with the minimum credits. That is OX and Light version. And this is a result. Let's see it, but I think that the more detailed we mention, it might not be generating that much of good results, but actually it's very good No, it's not very good because there are two extra legs behind him. So let's try our simple prom because it's generating pretty nice footage in Chat JBT. But this time, what we'll do is we'll use a separate project because the existing project might conflict with the pictures that is already generated. So if you're not getting the proper results, you always try a new project. On generate. And just remember, guys, this prompting and this text generation, it's the foundation of learning. That's why I'm telling you this and seeing examples again and again so that we will not have any problem when we see the extended clips and all the advanced features that we do in the course. Okay, so it has generated the similar results like we did in the hat JBD, but even that GBD version is more nice. Let's generate this. Let's see the result. And this one is much better. Now, if we want, we can use the reference image as well. We just created this with the hat GBD. So if you're able to create some reference polar bear image, we should be able to create some more results out of it. This time, let's load the reference image, actually. Here it is. Let's try to generate the video based on the reference image. So once it loads, let's send it. And this time it has generated the video. According to the reference image, let's see the results. Wow. Look at that. To the moon. Here we go. Wow. So this is probably the best video generated so far. So as we have seen an other thing that if you use hat GPT, to create the reference image, or even if you use nanobnana here, That's completely for free in flow as well. Let me just show you. So if you try to generate it with the image, using the pro version or basically any version, you can generate a free image. So after that, whatever gives you the best results, you can use that as a reference image and then create your videos. It will be more accurate. You also need less prompting. So it's completely up to you if you want to use the image to video or if you want to use the Text-to-Video feature, whatever works best for you. So as you can see, the nanobnana has almost generated a similar picture like chat DBT, so you even don't need chat dibD here because it will generate pretty nice results for you and use zero credits to make it. So again, let's view it finally. Let's click on this and actually upscale this too, one K is fine. Okay, that is for the picture. Let's generate this one. Wow. That it upscale. And now let's see the videos. So amazing results you can generate with VEO three, using the image prompting or the text baasePmpting, or even both. And one more thing that if you want to generate a text prompt, using an image reference because we have seen somewhere on the Internet and we want to generate a prompt to generate a video out of it. So what we do is we just use the reference image, identify the image and write a complete prompt to generate a similar kind of image or create a video for it. Define a complete scene, define this object, define the subject and all of the things. So whatever comes to your mind, you can just simply write it or define dictated in the prompt. Let's strand read the results. It's analysing the image, and then it will give you a complete prompt for it. As you can see, complete image rendition, prompt highly defined. So you can use this or the Video generation prompt. Let's try this. So even if you see something on the Internet, you don't need to use that image as a reference. You can identify that image and generate a video prompt using the chat Jeopardy, and then generate a simple results. This time, let's actually make a real for it. It's one X using only ten credits a light version, let's generate it. And these are the results we get, although it's a little bit changed, but let's see. So it looks kind of similar, but maybe that if we change this, paste it here again. And it was originally set to ingredients. So let's set it to 16 by nine in the ingredients. And let's see how it turned out to be. Okay, so it has generated this results. Obviously, when we mention a cued beer or something like that, it changed the facial expressions and all the things. So it's completely up to you if you want to rely it on the text based prompt using Google Veo three. But that will cause you one single issue, and that is it might change the facial expressions and look every single time. So you have to use the reference image. You might try nanobnana for a couple of times, and when you generate some results, you just attach that image, and it's always nice to use it with a prompt. So that's about it. This is how you can use both Text-to-Video prompting in Veo three and also text combined with the image prompting in Google Veo three. Just try this by yourself, and I will see you in the next one. 11. Frames to Video: Video, we are going to see how we can do the Frames to Video in Google Veo three or Google Flow. So basically, Frames to Video simply means creating videos out of your still images that you have generated either from Flow, you have your own images, or you have generated it from Chat Jeopardy. It doesn't matter. If you want to convert different images like a start and end scene, and you want to define the complete picture or how the animation should look, this is what you do. First of all, you click on the new project, and then you have this start and end buttons where basically you can add different images. So if you click on the start, you will find some of the images that you have recently generated or mostly used. You can search for it in the assets. So for example, I didn't generate anything for 31st May, but if I check 15th May, I have generated a couple of images that I can add to the prom dictly and it will be added right away like this. So this is one example, but let's just generate some new images for our project. So click on the new project. And once we click on it, we will get this couple of options. If you click on the agent, we will see all the options of start and end. But we won't be using it. Instead, what we'll do is we'll click on the image, we'll switch to Pro. 16 by nine is fine and one X is fine. So just remember, you click on the agent. You click on and choose Image. 16 by nine, one or two, it's your choice. It will use zero credits to generate it. So nanobanana Pro is also fine for it. So let's generate. Now, what command do we write? A picture animated woman walking in the supermarket, searching for products let's give a simple prompt and generate. So this is the result. Pretty nice. Whatever you want to use, you can use that. And we can refine it image of a woman walking in the supermarket searching for products, or you can say walking or walk through. The Als looking at products on the shelves because I don't want her to pick any of the items, I just want her to see the products randomly. Let's generate this result. And now this one looks much, much better. So if you want to use any of the character like let's say this one, this one is nice, we can add it two favorites and we can also animate it. So if you click on Animate, it will be added to the starting scene, but let's just generate one ending scene as well. Use the same character and take a back shot as if she is on the she is in the other section of grocery store, let's say, cosmetics. I don't want the back scene to be generated in the same shelf. So we'll use the image. 16 min nine is fine, and the reference image is already attached. We want to generate two results, and we can use NanobnaP because it will cost us zero credits. Let's generate So as you can see, both of these characters are very consistent. We'll use that or we can use this other wide angle shot as well. Let's add this one to favorites as well. So now, if you want to generate a complete new scene, you just remove the prompt, whatever you see here, just click Close or cross. Let's click on plus sign. Or, first of all, let's click on Agent and we will select actually video, click on Frames 16 by nine, and one result is fine. This one we'll use only the light version because it will use us minimum credits of ten, and then when you switch it to video, then you will see the start and end scene. So let's click on the start. Define the first scene. We choose this one. You can also search it two favorites. If you mark them as favorites, you will definitely find it here. So this one is first image, add to prompt. This one is the second let's add it to Prompt. And let's define a very simple prompt woman looking for products on shelves, camera follows. And since the camera angle is already defined in the first and last image, I think we don't need to define how the camera should follow from the front and back and so on. Let's keep it simple. Let's generate it. Basically, we are going to test at if both images are generated in a simple, continuous shot, or it changes it to the second shot. And also, if Google Veo three is able to maintain the character consistency here. So here's the video. Let's generate it. Let's actually play it. And because she's walking, and now this one is the second scene. But as you can see here, the animation changes like this as if it is not a continued shot. She doesn't put anything. She doesn't stop anywhere or put anything in the cart. It's already filled with products. This scene should be continuous. Shot. Let's see. But as you can see, this one is also very nice. We can use it in our cartoon characters or videos for our YouTube channel or for any of our animated story. Because nowadays, generating this kind of animations is very simple and very easy with Google Veo three. Let's see this one. Again, she puts a product in the card, and this is not a continue shot. Let's say, if we say it like this, woman walking from grocery to cosmetic section, something like this. Let's render it again. And this is our final video. She puts a product and she moves to the other section. So that's also good. But this is how you can generate a start and end shot using the frames to video in Google Veo three. Now, that is not compulsory. Even if we use only one reference image, we can generate similar shots. Let's just go back and we move to the new project. This is our new project. We cling on the agent in the starting shot. We just choose one. It's all sorted by dates. And this time, let's generate something else, kind of a real life scene. And again, this time as well, we will generate our images directly in VEO three instead of using hat pity for it. So we'll paste it here. We want the image to be generated. We will use nano Banana Pro because that is the most advanced one. We can also use nano Banana two or image in four, but let's just use this one. Also, we want to generate, let's say, two results, 16 by nine is fine, and it will use zero credits to generate it. It says, full body shot of a young aesthetic woman walking in the trail of forest soft natural light as under an overcast sky with a diffused lighting, gentle highlights and subtle shadows, you can use the chat GPD to define or refine your prompt just like so, and let's generate this results. Whatever we like, we'll use it in our video. So from here, you can choose whatever you like. There are two results. You can view it on the full screen just by clicking on it. And if you want to just refine this, you can click on reuse the prompt. Even if you have, let's say deleted it, you can reuse the prompt here and you can add something else. And let's say she is not carrying any luggage. Let's wait for the results. So this time, it has changed the image a little bit. So let's continue with this one. Now, this is our final shot. We will open this up, and let's say that we download it and use it as a first shot, or otherwise, we will generate a second shot as well. We just click to open it in the full screen mode, and we just want to define how our second scene should be. So we paste something like this, and let's grinderate it. We'll also see in this way, whether it's able to maintain this character consistency here or not. So as we can see, the character is very consistent here. That's how you can generate the character consistency in VEO three. You just use that image to open up and then just define the second scene here. Once you do that, you can also save it to a project, save this one to a project, and also you can mark them as favorites like this. You also might have noticed that the character is very consistent in the background scenes in the clothing and in every possible way, it is very, very consistent. And that's very important for the frames to video because we are generating a video of a same person in a same scene or video. Now, once we got the images, we get back to the main screen. We'll move all of this and we basically click on settings, and this time, we'll switch to this Video generation tool that is light. We click on this, once again, click on the agent. Make sure it is not highlighted, and then we switch to video that we will select frames 16 by nine, we only want to generate one results using the light version so that it can use minimum and you will get the start and the end point for it. So let's save these two favorites. Both of them are now saved, click on Start. And this is the starting point. Click on Add to Prompt, and this is the ending point. We'll click on Add to Prompt. Now, if you want the starting scene to be this one, you can flip it around just by clicking on this one. So this one is the starting scene now and this one is the ending scene. But let's just flip it around. So this one is the first frame, this one is the last frame. Now we just want a simple prompt on how this overall scene should be. So let's write our third prompt. Let's just piece whatever we have generated. And the more you define your shot or your scene, the more perfect results it gets. So it says a woman is walking in the forest the camera follows her with slow cinematic movement as she walks towards the camera. She happily smile and says, welcome back. And this is all about the visuals and all. Let's just generate the result. And just remember, guys, that Video three automatically gives some music and some subtitles and text on the screen. So I have to define no music, no subtitles and no texture on the screen. Keep it natural ambient forest sounds only. So let's see how it turns out. So this is my video. Let's play Welcome back. So this is perfect, guys. If we download this, let's say, here's our video downloaded. Welcome back. So this is perfect. No extra stuff is added. No weird sounds or noises. Only the forest sounds, and the walking sound is here. Welcome back. Now, you might have noticed that when we click on Download by just opening it from here, it just downloaded it right away. It doesn't ask for the quality and something else. You just go back one step, and if you click from here on the three dots, we click on Download, we get this ten ATP option here, so it will take some time, and then you should be able to download the upscaled video. Now, this one is downloaded. Again, this should be a good quality, as you can see. The quality is much, much better. B. And we can see that it is generated in two different shots. Let's write Continuous shot. Or maybe let's use the fast version. Now, it's not compulsory that even if you define this continue shot or anything like that, or you are using some other version like the fast or quality or something else or even Omni flash. The key to generating perfect results is you might need to generate with different prompts until you get it right. And I've also added one more condition with it. It should be a continuous shot from start to end scene without a cut. So let's see if we can generate Now, it didn't generate it in audio. As you can see, it is tagged as no audio generated. That's because I've changed the setting recently. In the settings, I have changed it to return silent videos. So we turn this off, and then we can generate it. But let's just see that if we successfully manage to generate a single shot for it without a cut, let's see. Let's play it on the full screen. So this is perfect. Let's generate it again. And if you see an error of image generation field, you might need to wait for like five to 10 minutes return after some time, if you have generated multiple results quickly, then it might show you that error. So just wait for some time and then re use flow, and it should work fine. So this is now with the sound. Let's play it. And as we can see, it is all a continue shot. Let me just play it so that we can hear do this. I'll not speak in between. Welcome back. So this is a nice command that you can use I specifically you are using Frames to Video. What I'll do is I'll attach all of these prompts with the resources section of this video so that you can directly try it. But I always prefer to just play around and be creative, use your own prompts to see how the videos turn out. Also let's move to one more example. Let's say that we haven't defined the end shot. So AI or Google Veo three will automatically detect how the end shot should be. And sometimes it's more or less same. It's only in the specific conditions when we want to define the ending as well. So let's define it like this. And the benefit of this is since these are not two frames, we don't need to define this last condition as well. It will be a continued shot automatically in most of the cases. And now let's see the result. Welcome back. So this is amazing. Let's download this. And if you want to upscale this video, what we need to do is just click on the three dots. Welcome back. Click on Download. Upscale and wait for it to upscale the video. Welcome back. But meanwhile, we can see that the results look amazing. And now the upscale video is downloaded. Let's see the results. Welcome back. And here is our upscaled video. As you can see, it looks very nice. Welcome back. And it's a continuous shot without any break. So that's about it. This is how you can use a Frames to Video function in Google Veo three. Just try this by yourself, and I will see you in the next one. 12. Ingredients to Video: Video, we are going to explore an other amazing functionality of Google Veo three, and that is ingredients to video. So basically what we do is just lets me just click on New Project. And what we can do is if we switch to settings, and actually, this one is the nano banana settings. As you can see, we are unable to switch to the video. That's because the agent is turned on. So always, when agent is highlighted in white, you need to uncheck this only then you will see all of these options like this. So we switch to ingredients, and basically what ingredient does is you can simply select three or more photos, let's say, a model, any product that you want her to promote or him to promote. And then any background or so on if you want that scene to be in the specific background. So this is a very powerful tool in Google Veo three. So let's actually, first of all, create a product because I already have created a model from Chat GBT. So what we can do is just we switch to image, and we want that product to be let's let's say three by four or one ratio one is also fine. NanobnanaP ocean blue perfume by SAD, let's say, this is our product. So this is our perfume name Ocean Blue by SAT on perfume bottle. I want my brand name to be written on the perfume. And just remember, guys, that it's not compulsory that you have to generate this AIE products right here in nanobanana. If you already have a product and you have a model, you can create a video for a model that is promoting that specific product. So that's the magic of Google Veo three. So let's actually create a new project or actually let's save this product to favorites and go pack, start a new project. I already have my model setup. We will turn off the agent, and first of all, let's upload an image. Now, if you have a model picture, whether it's real, whether it's AI, you can use that. We can upload it. And this is my model. So we will use any of the picture. Let's say this one. Let's click on Add to Prompt, and let's switch over to Video and actually ingredients to Video. 16 by nine is fine. Light version is fine because it uses the minimum credits, and we will use one more picture of our product. So let's switch it to backdates and this is my product. Let's add it. And if you want any specific scene, you can define that as well, whether you want to upload it from the gallery or however you want to define the scene. You can define it from here. So let's quickly generate it with Chat GPD this time. So I didn't mention a proper prompt. I just quickly mentioned a perfume store for women inside view, create an image. Okay, so this is let's say my scene. I want to add it to the ingredients as well. Let's download this and let me just upload it here. It should be in the downloads. Here it is. So let's add that image that we have recently uploaded. Click on at to Brompt and now we will write at we will select our first image of the woman at woman. So this is, as you can see, a kind of tagged caption is trying the perfume. Or let's say model is sitting on the sofa or let's say chair, trying the perfume. At the read, let's select the perfume. This one is the product. And in the and she is in at the read this perfume store. After she sprays the perfume on her wrist and smells it, she says, Wow. It's really amazing in Indian English style. Now, if you want to define how the camera shot should be like a white shot, a close shot or whatever, you can define it here, but let's just keep it simple. I just gave a random prompt nothing fancy. Let's rendit the results and let's see how it turns out. So we combine all the ingredients into a scene. Let's see the functionality of it. So now we can see that it has created a video for the product. She's sitting in this background, and this is our model. Let's see the video. Wow. It's really amazing. So this is very, very nice and looks very realistic. The only problem is she didn't spray it on her wrist. Where she wear bangles let's just define the exact position of where she should smell it. But as you can see how perfectly VEO three has combined all the ingredients into a scene and created an awesome video. So just imagine the power of this that you can use any model promoting any product in any scene you want. Now let's see one other result. This looks also very nice. Wow. It's really amazing. So this one is good as well. So it's your personal choice, whatever you like. This one shows a brand name as well, and this is a close up shot, so I prefer this one. Wow, it's really amazing. I like this one, personally. So if you want to upscale this, let's try doing it. Like on the street dots. Wow. Download upscale, wait for it to upscale. So our video is downloaded, let's see, as we can already notice very good quality. Wow, it's really amazing. So this is very, very nice. And as you can notice in the second prompt, she's also wearing bangles. So not only you can define one or two ingredients, you can define multiple, whatever you want your model to wear and wherever you want the place to be. Really yummy. Now, let's try something else. Let's say that I want to generate some of the images and products directly in Flow. So we will create a new project. And this time, we will create the model directly in Google Flow. So I've given a simple prompt and ultra realistic image of a charming man wearing a suit. He has a stylish beard and mustache wheel well groomed and looks like a Turkish man. Let's see and generate a result. And if it is able to create one ultra realistic model, we can use that in multiple prompts and for multiple products. Although the picture is quite realistic, but the man is too old, age, let's say, 30. But even this one looks very realistic. We can use this for our promotion if our product is especially for mature people. So this one is very nice. Although it's a close up shot, we want to promote a watch, a Rolex watch. Very stylish and modern. So I've given a very simple command, although it doesn't only showed the watch, it showed the hand as well. But that is more of a realistic scenario where we found a product, but that product is not separated from the scene or a person, but we want this person to promote this watch. Let's see, let's add both of them to favorites and let's actually create a new project. You can continue in the same project as well. Let's actually try to create it in the same project so that we can use this functionality as well. Click on video, click on ingredients. 16 by nine is fine. Light version is fine, and we just start uploading our images. So the first image is this one, and the second image is of watch. And we use at the rate this person promoting this watch. He's wearing this watch. And I want to say that let AI decide what he speaks for the promotion. At the read, this is the watch. You'll add it to the prompt and make him say something amazing about the watch. Because sometimes when you define too much, it doesn't produce nice results, but when you give simple commands and let AI decide what to promote and how you should promote it, it can turn out sometimes very good results. Let's generate a video, and here it is. Let's see this. This time piece is not just an accessory. It's a statement of elegance, featuring unmatched craftsmanship and a design that truly transcends. Okay, this is very good. This time piece is not just big. Okay, this is very good, but the watch looks a bit weird. It looks big, and he's not wearing this as well. So let's just alter a little bit. If you want to reuse the prompt, you can just simply click on this arrow sign where it says reuse the prompt, and it will be loaded right away in the bottom. So what I have done, I have just reinforced that he's wearing this watch on his wrist. This is the watch and promoting it like an ad. And also, but let's say, he is showing the watch. He's wearing on his wrist. So you can refine this prompt using hat DBT, but I'll just keep it simple right here. And this time, it's using 20 grades, and so let's see it has renertd with Omni flash. Experience the art of precision. This chronograph is more than a watch. It is a legacy on your wrist. Define your style today. So, again, it has done the same mistakes. Sometimes experiences the art when you use it in a new project, everything is automatically resolved. So let's try it in a new project. Again, I'll turn off this agent and use the video. Video should be set to ingredients. I just want to generate one result using Video light, and let's start uploading our assets. This one is the person, and this is the watch. Add the read. You might have to change the teat to find that recently used image because Google Veo will create some categories about the teats. So you might have to switch it to some of the teats when you created it. So anyways, let's add this person to the prompt. Let's write again, add the read, wearing this watch to the camera and saying something about it. Let's trend rate. And we have basically simply generated a simple scene with an option that three should automatically decide what to say and how the screen should look like. And this time, let's see, I think that this is what we exactly wanted. Lot of watches. But this one with the light blue dial, it's something special. It looks great everywhere. You know, I've had a lot of watches, but this one with the light blue dial, it's something special. It looks great everywhere. So this is perfect, guys. Let's upscale this. You know, I've had a lot of watches, but this one with the light blue dial. And as soon as the video is upscale, it will be downloaded automatically. You will see it here. Okay, so now our video is downloaded. Let's click on Show in folder, and here is our video. So as you can see, the quality is amazing and it looks ultra realistic. That's the kind of videos that we can use in our promotions on our websites or wherever we want. So as you can see, this is very nice. It's something special. Joe. Now, one more thing that is so amazing in Google Veo three in the ingredients function is you can add your favorite voice as well. So let's say that I don't want this default voice. I want to add something else. I want to hear it first. So we have different voices. When you click on Upload, you switch to voices and you have different voices here. Let's hear it one by one. Hi there. We're introducing a new feature where I can read this aloud in a natural human like voice. So this is nice. Hi there. We're introducing a new feature. Hi there. We're in Hi there. We're introducing a new feature where I can read. Hi there. We're introducing a new Hi there. We're introducing a new feature where I can read this. Hi there. We're introducing a new Hi there. We're introducing a new feature where I can Hi there. We're introducing a new Hi there. We're introducing a new feature. Hi there. We're introducing a new So, for example, I like this voice. I want to add it to the prompt and everywhere. And although the default voice looks very good, but still we want to use and change the voice. First of all, let's listen to the default voice that it created with the video. Well, it's something special. It looks great everywhere. So it is a Turkish kind of English accent, so that's better. But let's just upload a voice, and let's combine all the ingredients and let's generate the results again. And you will notice the difference. Now let's play it. Take a look at this. It's truly a masterpiece, crafted with precision. I'm incredibly impressed with the detail and the light blue dial. So although this voice looks fake, but that's how you can place the voice and basically dub it in any voice. You can take a video and you can dub. You know, I've had a lot of watches, but this one with the light blue dial. But this one is very realistic. It accurately picks up the accent of a person by his looks. So most of the times, you don't need to do this because VO three is getting advanced day by day. But this is what you can use. You can also explore different characters and functions and avatars here. So this is how you can use the ingredients function and add different products, different backgrounds, different persons, and combine them all into one scene to make any kind of promotional video, any kind of informative video or whatever you need. Different people have different requirements, but I don't know what is your main aim, but you can get creative with it. So just try this by yourself, and I will see you in the next one. 13. Creating AI Avatars: Video, we are going to explore this newly launched tool by Google. That is OmniFlash that is integrated in Flow and Gemini both. So you can explore this in Veo three. That is a very, very useful tool. So let's actually explore this in much detail. First of all, we need to create our avatar. So if we go to Gemini on our PC, you will notice that the interface is completely changed. It has a glow effect. With a blue background, that looks amazing. This is totally a new look in gemini. So what you can do is just head over to the corner where you see the settings icon. Click on that. And here you might see your Avatar option. So you will create your Avatar. You will just give your voice video sample so that it can record it in its memory. But if you don't get it here, you can also access Gemini using the mobile app. Let me show you how. So what we will do is just download the Gemini application like this one. You just open it up we agree here. And as you can see that we get the Gemini flash option that is Gemini Omni here. If you don't get this interface or you'll find an older version, what you can do is just head over to the Playstore and right here Gemini, uninstall the earlier version, and you will get the update option. You can update it from here. So once that is done, let's head over back to Gemini. And we need to click on this two lines. From here, if you're using Pro account, then you should be able to click on the settings and get your Avatar option. So if you scroll down a little bit, we can get this Avatar option. Just remember that this is only available for the Pro version. So click on that and it will capture your Selfie let's click on Get Started. Create an avatar that sounds and looks like. So click on I Agree. Allow camera and microphone access. Allow while visiting the site. Okay. You can also see the tips so that you can properly record your avatar. So let's start. So basically, it will ask for the quick verification. You need to speak out some numbers and you move your head towards right, towards left. It might ask you multiple times until it gets it right, and after that, your profile will be uploaded. So this is a complete procedure of it. So as you can see, before we spoke out some numbers, and now we are moving our head towards right, towards left, and that's how you can record the facial profile of your avatar into Gemini. So if you don't get this option on the PC, you can easily get it on the mobile application. Now you can close this window, we receive your recording. You can now close this window and return to Gemini. So let's close it from the top. And as we can see, our avatar is ready, and we can add ourselves to any frame you create by typing at the rate me or at the rate Sana them 201 in your prompt. So this is good. My Avatars saved. And if you see that your results are not very accurate, you can retake this and create your new Avatar. So for now, it's fine. Now, guys, I just found out that if you have created your avatar in the application, you can only create your videos based on your Avatar only in that application. It is not right now available in the app. I've typed at the read me or tried different things, but it's not working here. So let's try writing it here at the Rad. And as we can see, my avatar is shown at the top. And let's say that I will just continue writing my prompt. That as walking in New York City wearing stylish, white denim jacket, white T shirt, black jeans, anything you want to define, white shoes. And we can define the time as golden R. That is the evening time. And we will continue writing and says in happy tone. This is let's say that's me enjoying my life in New York, and you can too, and he laughs. Or let's say that's my avatar, enjoying my life in New York, and you can too, and he laughs. So let's click on the plus sign. And let's set it to videos. You can also make some settings that you want it to be landscape or portrait, whatever. So let's send it, and let's just wait for it to process. So we'll just wait for a few moments, and I'll just fast forward the process to show you the result once it completes. And our video is ready. Let's play it. Now, unfortunately, I didn't record the internal sound of the microphone, but you can see the visual appearance, how closely it matches with the avatar. The only thing is I gave it a cap sample, so you can give that the same. So look at this, guys. If you want him to wear a cap, you can add that thing because that's what we have in the sample. But already, we can see that it looks amazing. You can add some more conditions like ultra realistic and so on, because the background looks a little bit cartoonish, so you can make it look real or you can even give the city center actual footage so that it can mimic that. So I just added a prompt of adding a cap. Let's see if we need to rewrite our prompt completely, or we can just add something because that's the power of Omni flash. It should be able to add a cap. We don't need to redefine the whole scene. Let's see. So this is perfect, guys. This is how you can generate the video. Let's listen to this. So this is perfect. Look how realistic this looks. You can save the video from here. You can branch it in a new chat. You can like it. So if you want to upload different Avatars, you can do that, as well. You can just re launch your avatar. You can retake it and define it. But look how practically it has recreated this scene. It is also showing the spot right here on my face. So our video is downloaded. Let's see here it is in the videos. Let's play it. That's my avatar. Enjoying my life in New York. And you can, too. Let's see this. That's my avatar. Enjoying my life in New York. And you can, too. So look how perfect it looks. So that is how you can create your Avatar with Google Omni just by giving a sample, and you can recreate any scene. And you can just imagine how you can use it in different scenarios. You can now create a movie based scenes by using your avatar or cloning a scene. So that's the power of it. The potential is just unlimited. So this is how you do it. Destry this by yourself and evals you in the next one. 14. Working with Omni Avatars: This video, we are going to see how we can create our own Avatar in Google Flow. So if you are unable to use it in Google Gemini on the PC browser version, then you can use Google Flow for that. We've already seen how we can create this Avatar option in Gemini application on the mobile, but let's do it on the PC and let's see how we can create one on Google Flow. So what we do is just head over to Google Flow in Google. Right flow. You need to sign in with your same ID. Just remember that you must be using a pro version for that. So we click on the new project, disclose all of these options, and you can use this agent option as well. That is a very, very productive option. So first of all, let's explore this agent option. That is recently introduced in Google Flow, and that is a very interesting option. It makes AI video creation process so much easier. You just have to talk with an agent just like we normally do on chat. So make sure it is enabled. If it is not enabled, it will turn out black. When you enable it, it will just shine it a little bit, and you can use that. So for example, first of all, we will create a model or simple Indian simple, modern Indian girl aged 25 curly here. So if you want to just generate an image. We don't need to specify any other thing. It will automatically read it in the agent settings. As you can see, the agent settings is turned on here. So we click on Save, we send it, and as you can see, it's thinking. So once we give any command, it will start a session. So what is the benefit of the session? Whenever you want to add or remove something, you don't need to write the prompt all over again. You just mention remove this, add that, whatever. You just do that like you normally talk with your graphic designer. It will do it immediately for you. You don't have to mention or attach the images again and write the prompt again. You don't need to do all of that thing. So it says, Would you like to create this image union costing zero credits? Absolutely. So click on Approve. So this is our model Perfect. Let's now make a product that we want her to promote. So let's say that we want her to promote chocolate. Quoted dates product. Image with branding box, let's say. Again, it will only use zero credits because that's the power of Gemini. It doesn't use any credits for using nano Banana, especially if you're using P version, you should have no problem generating unlimited images for zero credits, or it only consumes a credit for the videos. So this image, we like this image. We'll add it to the favorites. And this is our chocolate filled dates. Let's save it to favorites. And now let's add the image. For example, we want this image to be used. So we add this to Prompt and we add the other one to the favorites as well. Click on add to Prompt. Now, we have referenced that both in the agent section. Now we will write here that we want a UGC video of this model. We will write add the rate, and it will open up the images. We will add this model because she's the one who will be promoting our product. So let's add to the prompt. And we will write here that she's holding give us space, holding what? We'll add our product. Again, at the rate, we will choose our product here, add to the prompt. And she is talking dietly into the camera seeing something in Indian style English accent. She says, first, I thought chocolate Deeds is a weird combo. But, honestly, I just love it. I'm soon going to buy one or my one box for my mom as well on her birthday. Now after that, you just open up the settings just to make sure that whatever you want to make, if you want to make a UTC video of nine by 16 for the reels, you can switch to nine by 16. Otherwise, 16 by nine is fine and confirm before generating always. Okay, that's fine. But this time we will not be generating the images, right? So we have to move to Omni flash, and Video generation results are all good. So you'll use the same Omni flash, and we'll use only one result for that. Click Save. And since we have mentioned UDC video already, it will directly create the video for us. But actually, this time, let's switch it to nine by 16 because we also need to explore how we can create some viral reels that should be used for Instagram or even Facebook ads or Instagram Reels ads and so on. So nine by 16 is also very popular. So let's click on Save. So let's send it and it will analyze our requirements first and will tell us the minimum credits that will be used for creating our UDC video. And as you can see here, it is asking us that this will generate 15 credits. So I think that we didn't check the setting. We want to generate 10 seconds of the video. But I believe OmniFlash by default generates 10 seconds of videos instead of eight like VEO three. So click on Save. Let's continue, send it again, and it will ask you to use the credits again. If we approve this, it will go forward. So let's approve this time, and let's wait and hope our UDC video is perfect. As far as these pictures are concerned, they are so good looks wonderful. And we have already analyzed so far that we don't need any fancy prompts nowadays. All of these AI agents are now smart enough to just pick up the simple prompt and they will create a video out of it. Okay, so if it fails, what you can do is just copy that, and you already see that as favorites, right? So what you can do is just open up the new project. Make sure to copy your text first. And then if you want to use this chat, we can use that, click on AD and we can add the images. But we need to explore some different dates for it. So this is the girl but we need to tag her as well. Add the read, add to the prompt. And she's holding she's holding what? Add the ad. We'll add our chocolates here and just check the settings again. We want to generate a reel. Let's continue with Omni flash Video tendon tool. And we need to change nine by 16 here actually. So click on Save. Click on Gendrate and we'll approve that. And now, hopefully it will generate our video soon. So let's wait for it to complete. And here is our video. Let's see. Let's. First, I thought chocolated dates is a weird com. It already looks very nice. Let me just make it full screen and play it. First, I thought chocolated dates is a weird combo, but, honestly, I just love it. I'm soon going to buy one box for my mom, as well on her birthday. So this is perfect, guys. If you want her to eat that chocolate. What you do is we will reuse this prompt here. She eats the chocolate date. And then she says, Let's just add this condition right here. Or what you can do is just go back. First, she should eat the chocolate. The box should be on the table. She is only holding the date. So as you can see, I just added a condition, and I didn't define the overall UDC ad video prompt. I didn't actually attach the images and sample and all the things. But it's again using 15 cars, so that's fine. But the quality of this thing you don't have to define it again. It will do it automatically using the agent. So that is an amazing function that is recently introduced by Omni. That makes your work so much easier. You just hat like you normally do with a chat model, and it'll do all the things for you. So this time, as you can see, the dates are placed on the table, and she eats first, and then she should give the review. So let's see this video. First, I thought chocolate dates is a weird combo, but honestly, I just love it. I'm soon going to buy one box for my mom, as well on her birthday. So that's amazing, guys. So this is super natural. This looks very nice. We can also notice the quality of this reel is just amazing. Let's download it. But first, I thought Chocolate dates is a weird compost, I thought. So we are going to upscale this video, and it will download automatically. So now our videos downloaded. Let's view it. Chocolate dates. So this is perfect, guys. So this is how you can create UGC videos directly in Google Flow using the Agent feature. Just try this for yourself, and I will see you in the next one. 15. Keeping Characters Consistent: This video, we are going to see how we can create a character in Google Omni inside Veo three. How can we assign a specific voice and how you can maintain the character consistency. So once you create a character and assign a voice for it, every time you just refer in the chat that we want that character to be in that specific scene. For example, we name our character as Mike. So every time we refer to Mike, it will automatically assign a voice that is similar to the previous one and also assign a character that is exactly the same as the previous one. With this simple technique. So we have this feature right here, Characters on the left hand side. We click on that. And from here, you can build and reuse characters for the consistent videos. You can use a sample below or create one of your own. So let's describe our character. So what we did is we have taken the support of Chat JBD also, we can do it directly in Google Flow as well. But I've just randomly defined that I want an Egyptian graceful man with a charming personality and character, build a prom for it. So it has generated two prompts, basically. Let's use one, paste it here. And actually, we have seen in the past that the simple prompts works best. So we can compare it side by side. Let me just paste my simple command. We will use nanobnana to create our character, click on next. And this is our character. You can re roll this or try again if you don't like this. Okay, so this is nice. If you want to change the character, we don't like this character, we can click on try again. But otherwise, let's continue with this character. Re role is to change the camera angles, change the sight or something like that. So we click on download this character, and we can actually name this character. Suggest Egyptian man names. So for example, let's name it as Omer. We can select a voice for it. Let's play it. Hi there. We're introduced? No, we need more manly voice. Hi there. We're introduced. So this is good. Let's check the next one. Hi there. We're Hi there. So let's say we want to use this sound and assign this to the character. So we click on Add, and now every single time we refer to Omer, it will automatically pick up this voice and this character, as well. We can define some character in fo, as well. Let's say how he should act in every single scene. What is his personality traits? For example, he's confident, shy, aggressive, or whatever. You can define that to craft your character. So for example, I will mention Omer is shy, although it's not compulsory, it's optional as we can see it here. Character info is optional. Or we can say that he's funny and very confident. Let's mention it like this. And we can define some other characteristics and he runs fingers through his here while talking. Let's say we defined that characteristics of this character. And now we will click on the top right hand corner, click on Done, and now, and our character is successfully created, named Omar. So if we go to All Media, or we go back, let's say, we click on the New project. Hi there. Okay, so I have actually missed saving that personality. So click on Done, and now it is saved properly, our character, here it is. Whenever you want to refer to that person, you just mention that as a reference. I don't know why it didn't save it the last time. So in the characters, you will find your character here, or under all media as well, you will find this character tagged and also it will be a character, so it will be with this sign. So anyways, let's move to our prompt media, or you can set it to all media. So we will just mention add the read and we will write Omer, make sure the spellings are correct. So we will mention that person. Click on that or click on AT Prompt is walking in the housing society and says, This is our new project. You can get a flat here for as low as 1% installment per month, and that's it. No other requirements. Let's try to create an ad or kind of a simple video for that. Now, under the settings, we will choose a 16 by nine, and we'll actually use Omniflash because we want to generate a video like on save let's generate and it says, Would you like to kick off this one Video generation costing 15 credits? Click on Approve. Click Approve again, and it will generate a video for you. So just remember the traits of his personality, he's funny, always smiling, confident, and he runs fingers through his hears while talking. Let's see if he's able to maintain those characters. Six is our new project. You can get a flat for as low as 1% installment per month. And that's it. No other requirements. This is our new project. You can get a flat for as low as 1% installment per month, and that's it. No other requirements. So as we can see, very cool personality, and we can define it separately in the prompt, as well. And let's try to give a command that runs fingers through his here while talking ultra realistic scene, it should not look AI. Click on Approve. This is our new project. Okay, so this is the new video? Let's listen this. This is our new project. You can get a flat for as low as 1% installment per month. And that's it. No other requirements. Let's actually watch project. This is our new project. Let's actually back this and watch it here. This is our newbject. You can get a flat for as low as 1% installment per month. And that's it. No other requirements. So this is pretty nice. You can try different variations of the prompts, and you should be able to get the best result. Now, let's put this character in a separate chat. Let's go back, click on Start a New project, and let's say that I just want to create this character or use that character reference. So we can start a chat or you can go back, actually. And if you scroll down, you should be able to see all of your characters here. Now, we'll just go to the prompt on the right hand side, and you should be able to use that character. But just remember that to use that specific characters, you need to open the specific character chat as well. So click on it, click on Add to Prompt. Okay, now, let's try to define a new scene. I just mentioned my character. He's stuck in the desert, runs to the camera and is crying and worried. Voice, he says, we will die here if we don't see our ship, and then he shouts and says angrily. You people think it's funny. Just wait until the water runs out. Let's see if Google Veo is able to maintain the character consistency, the voice, and also is it able to pick the whole scene or not. Now, let's see this video. We will die here if we don't you can see here, the character consistency is very good. You can see the sample down here and my character as well. We will die here if we don't see our ship. You people think it's funny. Just wait until the water runs out. So it is amazing. Just imagine the power of this. You can use it in making you can use it in filmmaking, ad making or any kind of videos that is being vialed across social media nowadays. All right. So we have tested our character consistency. Let's try to put him in a new scene. I just close the prompt from here and we can add a new video or create a new scene. At the rate, we will choose our character at to prompt. So let's try a little bit singing. Omer is enjoying on a beach, playing guitar with friends. They are all listening his memorizing voice, and he's singing Titanic song every time in my dreams. Let's see the results. Now, to be honest, I'm not very confident about how this singing part will be done in VEO three, especially if you are maintaining that same specific voice that we have defined with this character. And it says it will be generated using Omni flash model. We have already selected Omniflash. I don't know what's the problem. Let's try to change it a little bit. Let's actually not make the mention of that specific song, maybe because of the copyright problem. It is not able to generate that. So let's click on Aproof now, as you can see. Now, let's see this. He only has two friends sitting near him on the beach. As the sun goes down, as the sun goes down, we are sitting by the sea, feeling so much love my good friends and me. And suddenly the third friend pops up as well. So like I said, he didn't sing in a very good voice, so is not very good in singing, I guess. Let me just add make him sing like a professional singer. Let's try a couple of variants. So some prompts works, some prompts doesn't. But as you can see, the character consistency is just amazing. Mm hm. So if you compare this image with this one and this one, so all of them are pretty consistent in all of the scenes, whether he's laughing, dancing, worried, and so on. So that is how you can maintain the character consistency. I just creating your character in Google Veo three. And not only you have to make the character altogether in view, you can also upload the image. So let's click on the new project. And actually, when we click on the characters, we can upload the specific character we want as well. We can make our own avatar which we have already seen in the previous videos. But basically, we can use any character by just uploading the image as well. So just try this by yourself, and I will see you in the next one. 16. Working with the AI Agent: Video, we are going to see how we can use the agent inside the flow so we can have a better video creation. Basically, it will aid you in the video creation. So basically, agent works as an assistant besides you. You just instruct person this AI person, and it will generate the results based on that. Also it will wait for your permissions. For example, if we click on the settings, confirm before generating. So whenever any results are being generated, it will always ask you for the confirmation or you can get different settings. Image generation, for video generation, and so on and so forth, which model it should use. So this is all about the agent. If you want to turn it on, you just toggle it on once and it should turn white. This means the agent is on, but if you want to turn it off, you just click on it again and it should turn black. And also, if you want the results to be in the specific style, or you want to dictate some of the instructions to this assistant, you will click on Agent Instructions, and then you can add those instructions here. You can mention the reference images as well, how the image or video should be generated. Create guidelines for the agent, and then it will work accordingly. Very useful feature. So let's ask something Agent. You can also define some task, allocate some tasks or also you can ask for some things. Like, let's say that I want to ask how do I make portrait videos instead of landscape. Now Agent is thinking, and it will answer your query here. So it says you need to click on the settings. In the prompt box, you will see the video preferences and select nine by 16 instead of 16 by nine in the video tenderon default. As you can see, I followed that instructions, and I am able to change the Video generation orientation. Now, for example, I want to generate a scene. Let's say that I want to create four shots video, and it should be about hero with other people stuck on island. And their ship is gone. Write shots and sequences with dialogues. So this agent should be able to guide you completely on how we can generate one. Now, it's thinking, and you should probably get a window on our right when it generates the result. But if for some reason, you don't see it, you can expand it from here, and then you should be able to see this window. So let's close it. It's still drafting the island narrative. And if you want to see the result, you click here. And as you can see, the four short cinematic story board, the frame one, the Ash of our past. This is how it should sound like the narration, the S takes what it wants, leaving us with nothing but ash of our past. And then the second scene. Then the third scene, then four scene. So you can use that to create scenes as well. Also, if you want guidance about anything within w three, it should be able to guide you as an agent as well. Like hat JBD, it is also asking, would you like to refine the dialogue or shots before we begin generating? And let's say that no just generate it. Our hero is let's let's say myself, added to the prompt. So it is using the 15 credits. Let's click on a proof. Basically, for your Avatar, it should be able to generate it wherever you want. You shouldn't be in that specific chat of the character. And now, as you can see, a beautiful scene has generated. Let's see this We will survive this. We need to find shelter before nightfall. Follow me. So, as you can see, a beautiful scene, it should be ultra realistic life like video. Video generation should be 16 by nine. And you can also attach any reference image if you want to create a real life like scene. So for example, I want the island to look like this, let's say, we can use that reference image here. So let's just upload the media. And we add it to the prompt like reference image. Andscape video. Let's mention that as well. Click on generate, click on a proof. And as you can see, now it's generating the landscape video. Okay, so it has generated a new scene. Let's play it. We're actually stranded. We will make it. We have to go in. So pretty nice. We're actually stranded. At this time, the scenes look better. It has improved a little bit. But the more prompts you give, you should be able to get the ultra realistic scenes as well. Make it. We have to go in. So as you can see, it has closely matched this image with my Avatar, which is my own picture just like this. So just imagine the power of the avatar creation and then using it using the agent function into VEO three. Very interesting features. So after that, if you want to define something else here, some other scenes merge in, you can ask Agent to do it for you. So that is a little bit of introduction on how we can use the Agent feature on Google Flow or Google Veo three. Obviously, we will explore it in much detail when we will work on a couple of projects in Google Flow once our course is complete. So just write this by yourself, explore all of these features, and I will see you in the next one. 17. Extending Video Duration: Do, we are going to see how we can extend the VEO three videos duration beyond 8 seconds. And yes, you can easily do it in Google Veo three. Although this is not a direct feature, by default, it only allows you 8 seconds, but you can add different scenes and you can easily add extensions to your current scene or video. So for the demonstration purposes, I have created an ad with the support of Chat GBD, and I asked to write a script for an ad of civic car modifications, review by a person talking about auto gallery services, he's mentioning his experience. I put a note here. I want to extend the clip beyond 8 seconds to demonstrate how it works. So just write it accordingly, the first clip, and then the extension of it. So it has created clip one and then the continuation of it after asking a couple of questions. As you can see, the first 8 seconds customer speaks to the camera. Honestly, I am really happy with the modification, and then you can actually see the results and it shows about the car. Then the close up shot. So everything should fit in within a single ad. How do we do it? It should land beyond 8 seconds. Let's see how it works on Google Flow. So first of all, what we'll do is we will open up a new project, and from here on the left, we will use the scenes. Click on that. And now on the right hand side, if you're using Agent, you will get this option here, or otherwise, if you click on the settings down here, or you just click on the back area, you will get to this section. So it's your choice if you want to use the agent or not. Right now I'm not using the agent. Let's past our first scene. I'll just mention this. And let's mention this together along with the headings. And let's write an Indian customer speaking to the camera. You'll use Omni flash for it, 10 seconds. That is fine. We only want to generate one result. If you want to generate two, it will cost you more. So let's generate it. And then we will see how we can extend that. Okay, so our clip is ready. Let's listen to this and see how it looks. Honestly, I'm really happy with how my civic turned out after getting it modified by Auto Gallery. The finishing is really good, and the car looks completely different now. So this is all about car modification, right? But this is the first scene. If you want to extend the scene, here's how you can do it for the second scene. Okay, for the second scene, if we open this up, it will be opened in an editor. But if you click on plus sign, we need to add a clip. That means we need to create the clip first, and then you can add the extension to it and it will be added to the scene. There's one way of doing this as well. So what you can also do is you can just use the Agent services. Like I said before in the previous video that we will explore Agent feature in much more detail. So down here, if you go back, you have this agent feature. You click on it to turn it on. Make sure that it is shown in white. And after that, if you click on this little expand sign, it will open up the agent. After that, what we'll do is we'll just click on the plus sign and use the first video. Click on Add to Prompt. And after that, what you do is we will copy the complete, clip two reference or you can say dialogs and you can paste it here. And we can ask merge the two clips and expand time duration, whatever. So you click on it, hit Enter and after some time, it might show you this message that it's unable to do that. So that's not a problem. We can just remove this, hit Enter again, and it'll ask you for the credit utilization. So it sees that 15 credits will be used. So click on apof to continue. And like you see here, it immediately created a new scene. So we just have to see whether it extends the clip or creates a new one. But in both cases, it's very easy to combine them both. Flawless incredible attention to detail. So in scene two, although we have provided the reference, but still it didn't use it. So that's why this shot is completely different than the earlier one, as you can see. Here, he had different civic, but now it's demonstrated in the new model. So what you can do is you can simply add a reference shot here, you can use the save frame. Now, how we can use that save frame, it's very simple. First of all, we'll open up the EDR clip and we'll go to the corner and click on Save to frame, over over it, click on Save to frame, and now we can use that. You can either create a new scene or you can just simply use the agent, click on plus sign, and then you use the saved frame, click on add to prompt, and it will create an extension following the same scene. We'll approve this for the cret approval, click on Approve, and it will create a second video for us. And now let's listen to this second clip. Because we have reused the reference image, you can see the results yourself. So let's actually listen to this clip. You can actually see the difference yourself. Let me show you the details. The fit and finishing are really clean. Everything looks properly. I'm really satisfied. I'd recommend Auto Gallery services. Great work by Auto Gallery. So you can see that it has created a perfect extension to the clip. Now, the question is, how do we combine those clips so that we can get an extended version of it? The first thing is that you need to first locate the earlier clip and then add to that clip, or you can also use the second clip and add the clip then we can also rearrange that. I'm going to show you the second method where we will open this clip and then we will rearrange the videos so that you can get the idea of how you can rearrange the clips as well. Now to merge those clips, what you do is we'll open up the second clip, and we can add it to the scene or you can move it further if you want to. Actually, let's click on Plus Sign and click on Air Clip and let's use the first shot, click on Add to scene. And after that, it will create a scene like this, and it will be opened in the editor. After that, you can simply move it by hovering over, as you can see. If we try to hold that clip and over towards the left, it will be shifted to the first clip, and then you can add a couple of portions in that clip. For example, the speech is still continued, so it shouldn't show the scene that he's leaving. So we can, you know, reduce that point. We can get rid of that point. And now we can also look at the scenes again. We need to cut a little bit more. And now the second scene looks great. Let's listen to this finally. And we can already see that it's almost of 18 seconds of the clip. Let's play it. Honestly, I'm really happy with how my civic turned out after getting it modified by Auto Gallery. The finishing is really good and the car looks completely different now. You can actually see the difference yourself. Let me show you the details. The fit and finishing are really clean. Everything looks properly done. I'm really satisfied. I'd recommend Auto Gallery services. Great work by Auto Gallery. Honestly. So this is good. Obviously, we need more blending in future so that we can create more perfect ads. But if we download it this way, let's see how it exports the overall scene. So here is my clip. Honestly, I'm really happy with how my civic turned out after getting it modified by Auto Gallery. The finishing is really good, and the car looks completely different now. You can actually see the difference yourself. Let me show you the details. The fit and finishing are really clean. Everything looks properly. I'm really satisfied. I'd recommend Auto Gallery services. Great work by Auto Gallery. So, as you can see, it's almost of 18 seconds, and this is how you can double the duration or actually extend the Video three videos duration beyond 8 seconds. Just try this by yourself, and I will see you in the next one. 18. Exploring Flow Tools: This video, we are going to explore about the tools that Google Flow or Google Veo three offers. If you click on, let's say, a new project from here, on the left, we can get the option of tools. So these are very effective ways if you want to automate your different tasks. For example, if you want to create especially the cinematic car videos, you can create with this tool, you can generate subtitles automatically by using this tool. So we have a couple of them, and they can be really, really helpful. And we will demonstrate some of them here as well. And also, if you explore it on the top, you can create your own tools as well. So click on that. You can submit your own tool, and that will be featured and used by community if you'd like to. And also, it offers premade templates as well. So for example, you want the drawing to stylize or scene explorer mockup and all the tools if you want to use that. So instead of creating your own, you can use the templates as well if you want quick results. So these are some very, very interesting tools. So let's actually explore some of these, and these are very, very interesting. Like I said. Let's say that we want to explore simple sketch. All we have to do is just open it up, and this allows you to draw anything and it will be converted into an image based on you prompt. So for example, I just made a very bad drawing. One person right here, the other person right here. One is happy, while the other is sad. And let's write a deep message here. So I just wrote a message. One old woman happy with her life while the other is sad. They both had same luxuries of life. It's only about the mindset. You can give additional introductions. You can upload a gallery for the reference image as well. But let's just continue with this prompt. Hit Enter and it'll start generating on her, right? So as you can see, they took the reference image and they created image of two different women. Actually, they look the same. So let's mention it two different, two different women. And let's enter again. So it depends on what kind of sketch do you make, and it will create into something amazing. So as you can see, one is sad, even though she's younger, she had more luxuries and all, and one is happy. So this is how you can use these kind of tools. Obviously, this is a simple, very simple demonstration of the tools. Another very interesting tool is the mockup, and you can use it for your branding, whether you want to show your brand on a mobile on a billboard, on a bag. Here is how you can easily do it. First of all, let's just create one image. So for example, I'll just mention it here. Create an image of AI energy drink. Ultra realistic. I'll just keep it simple. And in the tools, we'll choose nano banana. Nanobnanas already selected click on Save, and let's strand it. And right away, it started creating two ultra realistic images. So this is amazing. Let's use this one. This one looks pretty good. So what we'll do is we will like it, market as favorites. And also, if you want to download it, you can download it into original size. Let's say one resolution. Okay, so it just downloaded. Now let's go to the Tools. Let's say that we want the same image to be displayed on the billboard. We will use mockup for that, open the tool, import the image. We can import it, as you can see, with the d one, or you can upload it from the media. So you can do anything, and you can also have the option, even if we didn't like the other one, still we have this option displayed right here. Click on Add to Image and where do we want to display it? Let's say that we want the billboard advertisement for that. It's generating a mockup. Let's see. And, guys, just look at it. It looks so amazing, realistic. And just imagine the power of it. You can display it in a poster. As soon as we click on the poster, it'll start mocking up the other image. You can display it in a book, on a pen, shirt, laptop, wherever. So let's see that. It also looks very good. On a bag, let's try this. And you can see now it's on a bag. So you can try different commands with it, and also you can create your own tools. It's a pretty interesting area. Also, if anything interesting comes to my mind, you can check it out in the end of the course in the frequently asked questions or the other useful stuff. You can get the use of different functionalities apart from this course topics. So pretty useful tool. You can explore different tools by using these simple guides as well. It also demonstrates how it works. So that's all about how we can use tool if you might be thinking that what it is all about in the new project that you can explore, you can try using different tools and see by yourself how it should be useful in your work. So that's about it. Just try this by yourself and I will see you in the next. 19. Using the Extend Feature: This video, we are going to see how we can extend any video clip beyond eight second using the Extend feature. So before we have already seen how we can use the ad clip feature, for that, you want some reference image. But now we are going to explore how we can extend those videos beyond eight second using the Extend feature of the clip. So for example, let me just show you what is Extend feature. So before we had these kind of clips, right? So once we click on this plus sign, we only had the add clip option and not the extend, but this time, we are going to see how we can use the Extend feature and how it appears. So we click on the new project, and let's take the support of Chat JPT. Mm. So I've written one prompt here, create a horror scene in the city. People were living their life normally. Let's make it life. But they felt something strange. Suddenly, there's earthquake building starts collapsing. Then a person like a hero movie character came here with his superpowers. He rescued them, divide scenes into three parts each within 8 seconds limit allowed by VEO three. So as you can see, it has divided it into three parts. Let's use the first prompt, and we are already in the flow. We'll not use the agent. Instead. What you'll do is we will use VEO 3.1 instead of Omniflash. If you want to use less credits, you can use the light version as well. But let me just use this one. Okay, let's just view it. Soon as I can. I'll be there as soon as I Okay, although it's not ultra realistic, but let's just continue with the scene. We'll use the plus sign, and this time it should show the extend option as well. Like I said, since we were using the view 3.1, in this case, it's using the Extend feature as well. It allows Extend feature. So click on Extend, and the prompt to extend, we will just simply copy the second scene and paste it in here. And it will be very easy to extend the clip. You don't have to use the reference again. And now just look at this clip and observe how seamlessly it has merged those clips or extended that scene just by extending the prompt. Let me just play it. Listen to this carefully, and view the visuals as well, how seamlessly it has extended the video. I'll be there as soon as I can, and traffic is just terrible. Hold on. What's that noise? Did you see the birds? Something's wrong. Run, run. Run. It's coming down. I'll be there. So this is very nice. Let's extend it further. Click on the plus sign, click on Extend. And just we will merge it with the third scene. And this is the next one. Let's wait for it to complete. And then we will see the overall result of the clip. So our third clip is also generated. Let's see the overall result. And right now, we just learning the extension of the clips. We are not learning how we can create the ultra realistic scenes. That is for the later part. But let's just see how seamlessly the extension is happening here. Let's play it. Soon as I can, and traffic is just terrible. Hold on. What's that noise? Did you see the birds? Something's wrong. Run, run. Run. It's coming down. Get back. Whoa. Look at that shield. I'll be there as soon as I can. So as you can see, it has merged all of these three clips perfectly. If we try to download it, it will be exported as a single scene and as a single video. So very important if you want to publish the clips as YouTube shots or even for your longer videos, you can do it seamlessly here. Now let's view the scene. I'll be there as soon as I can, and traffic is just terrible. Hold on. What's that noise? Did you see the birds? Something's wrong. Run, run. Run. It's coming down. Get back. Whoa. Look at that shield. Back. Down. Get back. It's coming down. Get back. So although it's not perfect, like I said, the one man that was running in the public certainly became a superhero. As you can see here. But the clips are extending pretty fine. So that's how you do it. With Extend feature, just try this by yourself, and I will see you in the next one. 20. Static Images to Video: Video, we are going to see how we can convert any static picture into a small video using the animate function. So for example, let me just take an example of this image right here. We open that up and you can actually use the three dots on the right hand corner and click on Animate, or otherwise, you can upload a new image and then animate that. So as you can see, we have a couple of images here. If you want to, let's say, animate this girl, you can see already that it's a static image, but once you want to animate that, you just click on the three dots, click on Animate, and you can define the first and the end scene. But in this case, I'll just use the first scene, and I'll write something like Let's say that I've written that I want to animate that she's shy with the camera, laughs a little bit lightly, and then focuses again on the camera. Let's see how it animates that. Okay, our video is generated. Let's view it. Can I? Can I? Can I. So you can see that it has applied some music with it as well. But if you just want to use the simple clip, you can remove the music afterwards. And also let's try it in a new project. Actually, I've created one clip that a man is eating his lunch in a restaurant and a yellow taxi is passing by. I've generated this clip with Chat GPD, but let's just try that directly in Google Flow, as well. Click on the New Project, and then we will animate that. So first of all, let's generate an image. I'll just simply switch to image 16 by nine, it will be Nano Banana Pro, and you might use one result or two results. Let's actually use one. And you can see this results is looking quite good. What we can do is just click on these three dots and we can quickly animate that. Eating food, looking at the window and enjoying city activity. I don't want him to particularly look at the taxi, see the overall scene out of the window. So this is a static image, and we want to convert it into a short video clip. By just animating it, you will see the quick result of it. And I believe this result is much better with the nano baanana instead of this. Firstly, the face of the person is not visible and he's looking weirdly on this side because the window is on an unusual side. It's like in the slant facing. And in this one, it looks nice, but the car looks too big, I guess. So this looks a cinematic scene, especially how you see it in the movies. So I believe this is better. So after waiting for some time, this is the result we got. Pretty nice and pretty realistic. So that is how you can animate different images from static to a video using the animate functions. So it can be used as a small filter clip. If you want to add some of the scenes or some of the contributory scenes into your videos, this is how you can use it. Just try this by yourself, and I will see you in the next one. 21. Extending with JSON Prompts: Studio, we are going to see a practical demonstration on how you can use the Jason Tmting to extend your clips. And that also seamlessly. As you can see here, this clip is around 43 seconds. And let's try to play it a little bit, and then I'll show you the final video result, as well. We talk about enough in the UK is how expensive getting around can be. Public transport is really convenient, but the prices are getting crazy. Sometimes I check the price of a train ticket and genuinely think, wait. That's how much. So you might not even feel where the scene is changing. And let's see the final result of it as well, and then I'll show you how we can use it in the scene Builder using VEO three. Honestly, one thing people don't talk about enough in the UK is how expensive getting around can be. Public transport is really convenient, but the prices are getting crazy. Honestly, one thing people don't talk about enough in the UK is how expensive getting around can be. Public transport is really convenient, but the prices are getting crazy. Sometimes I check the price of a train ticket and genuinely think, Wait. That's how much. And if you travel regularly for work, those costs add up really quickly. Then you've got rent, food, bills, and everything else. Sometimes it honestly feels like you're working just to survive, not actually live. And that's the frustrating though. Of course, public transport has its benefits, and you can get around without owning a car. But with rising fares, rent, food, and bills, it's becoming harder to keep up. Sometimes it feels like you're just trying to get through the month. So you can see the awesome results. You are not noticing any kind of cuts, and that is the power of Jason prompting. We haven't used anything different, but we just tried to create the clip with Jason prompting. So let's see how we can create that clip from scratch. First of all, I just mentioned this right here, a simple prompt, a 25-year-old modern woman talking about her cost of transport in UK and its benefits and the cons and high cost survival issues this way with high cost. And just write a script that she's almost talking about 24 seconds, walking on the street, include script seen. It should be ultra realistic. I just mentioned the random text, as you can see. She's logging and talking directly into the camera. She's Indian but has a UK accent because she lived all her life in UK. So it has generated the initial results just like so, but as you can see, the dialogue is quite short. So I requested it to make it for 24 seconds still. The script is short. So let's I just mentioned 2 minutes long dialogue, and it has created some of the dialogues. This looks quite nice. And then finally, once I got enough dialogues to divide it into the scenes, I just requested it to convert it into JCNFmat. So here is the JCN format. You can see. But still, the dialogue looks quite concise. So I asked him to write all dialogues and divide it into three scenes. Each portion should be of 10 seconds. So again, it has generated some results. This is the scene one. It has defined the scene and the dialogue. Scene two, define the scene and the dialogue and scene three. But again, I wanted it in Jason format, so I tried to convert it into Jason format. So it has divided it into scene one, scene two, and scene three. So what I will do here is I will just share this whole chat with you so that you can get a better idea or how you can refine your results with hat GPT. Okay, so I've divided it into different scenes and then extended the scene as I like. Let's actually try this all in Google Flow. I'll just go to the new project, go back. One step back as well. I'm on the main screen now, click on New Project, and let's actually try to paste that Jason format. So here's scene one. Scene one is ending right here, and from here, scene two is starting. So let's copy and paste that. Let's see how it creates the results. You can use the agent feature, but I'm using simple version of it here. So I'll just uncheck that. Video is fine, ingredients. You can use the frames as well, but ingredients are fine, and 16 by nine, all good. We are using VO 3.1 it although it will take some time to generate the results, but it will use minimum credits to generate the results for you. I just want one video. Let's try to create. Honestly, one Okay, so this is the results. Let's play it. The thing people don't talk about enough in the UK is how expensive getting around can be. Public transport is really convenient, but the prices are getting crazy. Honestly, one thing Okay, so I noticed one thing that there are natural scene cuts here. So if you want to eliminate that, you can ask it mention it should be without scene cuts. Start altogether with scene one. Let's try it again. I'll copy from here. We don't Try it in the new project, paste it, and generate it. Okay, so our result is gendertd. Let's play it this time. Honestly, one thing people don't talk about enough in the UK is how expensive getting around can be. Public transport is really convenient, but the prices are getting crazy. Okay, so first part is good. Let's move to scene two and click on the plus sign, click on Extend with Vo 3.1 lit, paste it here. And since it is GSM prompting, you will see the result will be quite natural. Honestly, one thing people don't talk now, at first, you might see that nothing has happened, and it just loads it again and the other one is showing in gray, but that's not what it is. You have to wait for some time in order to let the clip process. So just wait until it processes, just like so. Okay, so you can see that it has renated the second clip after processing a little bit. Let's see how it works. In the UK is how expensive getting around can be. Public transport is really convenient, but the prices are getting crazy. Sometimes I check the price of a train ticket and genuinely think. Sometimes I check the price of a train ticket. So as you can see that we can feel the cut here. Crazy. So in those cases, what you can do? As you can see that it ends with prices are getting crazy. So we will simply say here, write the second scene script and continue with the talk that the prices are getting crazy. Now let's try again. Click on Extend witho, paste it, and let's wait for it to process. Okay, I still didn't get the right results. So what you can do is just generate a complete decent prompt and make it and divide it into the scenes that it should seamlessly continue in the next scene without any cuts. So this time it has generated quite a good amount of script. We have to try it again, and that's how the practical project works. So let's start a new project, actually, and we will copy this complete script up to scene one. Or what you can do is first of all, just copy the complete scene and then extend through clips. Generate the video. I obviously will generate only 8 seconds of a clip, and then we will extend to the scenes. If it failed for the first time, we'll just regenerate that. Okay, so I have tried a couple of times, but it didn't work. So let's actually continue with our initial idea that was this thing up to scene one. Try it here. Generate a new result. Honestly, one thing people don't talk about Okay, so this is our first clip? Honestly, one thing people don't talk about enough in the UK is how expensive getting around can be. Okay, so she took a little bit of pause, and that's a nice thing. Let's continue with scene two. Just make sure to copy it with the bracket opening and bracket closing. That makes the results more even. Click on the plus sign, click on Extend clip, and we'll just mention our second reference and add it. Honestly, one thing people don't talk about enough in the. Okay, so this is a result. Let's look at this. The prices are getting crazy. Public transport is really convenien. And if you want to reduce the pause, you can just move it backwards, and it looks like this now. The price is again. But this way, it might show the jump a little bit. So instead, if we extend it. The price is again. The price is The prices. You can see that it looks much more natural. So we will continue with this. Let's extend it to the third clip. I'll just paste the third clip here and wait for the results. So this is all the practical demonstration. That's why I didn't cut any part, whether it goes right, whether it goes wrong, so that you can get a better idea on how you can get the best results out of it. So if you want to prevent jump cuts, this is the way for it. Okay, so let's continue with the third clip and see whether it continues without a jump cut or not. And genuinely think, wait, that's how much. And if you. So you can see for yourself that we didn't feel any jump cut here. So this is perfect. Genuinely think, wait, that's how much. And if you try. Just listen this again. Wait, that's how much. Perfect. So I've shared this complete chat with you. So you can actually practice this without the jump cuts. Let's continue to our scene what? Scene number four. So extend the clip. Let me just paste the fourth one. And again, we will start with the curling brackets opening and the curly brackets closing should be our closing for scene four. Let's simply paste it here. Dund that, and it will automatically take the reference and do it without the jump cut. So the main prompt is this one that works best for us. So I'll just piece that in the word document where you will get the reference of the chat. But this is the code, I believe, for prevention of jump cuts. So I just mentioned it here, although it didn't paste perfectly. Let's try it with the text. Okay, this is fine. So you can use that as well. Okay, so this is our fourth clip. Should join seamlessly without any jump cut. Let's try this. Then you've got rent, food, bills and every shade. Sometimes it honestly feels like so what? Perfect. Food bills and every shape. Move to fourth clip. Just the last one. Let's try the last one. I just already clicked on the Extend clip, paste it here, generate the results. And sometimes if the conversation is starting from other point, you can refine that and ask At JBD to continue the conversation from that point. So let's try the final clip. Okay, so next clip, let's listen this. Live. And that's the frustrating part, but with rising fat Now, you can notice that it doesn't continue properly, although it has no jump cuts, but the conversation looks quite weird. And that's the frustrating part, but with rising she quickly says, but with rising fears, continue the last scene with the conversation, and that's the frustrating part because just continue with that. You can also dictate it easily using your mic. You can use this clip. With me? I'll just delete this. Because it ended with B. So extend the clip, paste it here, and I'll just mention it here because the transport is something you generally need. Okay, so I believe that should be it. Let's listen to the last clip. Straight part because something you genuinely need. But with vibe? Because it's something you genuinely need. It only miss by some supporting words you can say. But that is how you can combine all of the clips into one clip without any jump cuts. Stry this by yourself, and I will see you in the next one. 22. Two-Person Dialogue Clips: This video, we are going to see how we can make a dialogue of the multi characters in a single video, and I'm going to show you how we can fix the VEO three multi character dialogue errors because you might have noticed that sometimes if you combine both characters or you create one single picture of it for the reference image, it might jumble up dialogues. So let's say person one is also speaking person two dialogues, and that's messed up. So if you want to rendert a video like this, let's say like this, as you can see, this is a long video. The person one is frustrated, is talking about why I'm not finding a job, and I've earned every skill, but still experience is required. And then the second person joins in, and he's telling about a solution and how he has overcomed that thing, and what are the benefits of joining those courses, let's say, and he's telling all about it. Let's see after that what happens. And now you will see that the first person joins in and he's talking on the other side, and he's answering the second person is answering for his questions, and this is how the scene ends. But as you can see the working, let me show you if I show you my project of this 900 offer, let's say, you will notice that I have used a lot of credits. So this will cost you a lot of credits if you generate the results like this because every time you have to extend the scene, you have to cut different scenes. Maybe it will not be as you like, or someone else is speaking someone else's dialogue. So if you want to overcome this, you need one additional step. What I've observed, if we separate the dialogue of the two, then conversation will happen in a much better style. So I've also tried this thing that I've generated one person image and the other person image, and I've combined that for the character reference. I just added it to the prompt and then tried to speak the dialogue, but now it speaks something like this. Let me just show you. So the two persons are in the same frame. I don't want that to happen. So I want separate conversation by each person. Also with the minimal credits, I don't want to waste my credits. So this is what I have found. What we do is we just go to Chat GPT, and let's ask it to separate all person one dialogues. Okay. And then we will say separate all person two dialogues. No divan in parts so that we can continue the clip. No divan or person one dialogues. Two parts. And let's try it now. So if we start a new project, let's say this is my project. Let's use the reference image. Let me just upload one. So this is the person searching for a job. We will create a complete dialogue for that person. Add the to the prompt. Conversation. Without credits waste. Okay. Let's try it. So what we do is just copy and paste that first command and make sure that it should be only like 14 to 15 words per conversation. You will see for yourself how much can be done in a single stretch. If you use Veo 3.1 it, you'll use the minimum credit that is ten. Just trying to save the credits. And this way, you will notice that you will end up using very less credits as opposite to the previous scenario. Okay, so although we did get the result and it has completed all of these into a single conversation. So we can identify that as well. Let me just do that. Count number of words. So you can do a little research on it. It has 29 words. So this means that at a stingle stretch, using only ten credits, you can generate almost 30 words. So this way, you will not use more credits. You can try 35 then 36, and whatever works best, you can use that. But I forgot to mention the scenario. That's why it turns up like this. So he's not actually talking on the call. I didn't mention that. Let's actually mention that this time. Actually, this time, let's mention three lines to check that whether it covers it in a single conversation. But we have to define the scene as well. Otherwise, it will not pick it right. So here is the scene. We can adjust the scene as well. Scene. Let's make this setting and dialogue. Fresh graduate is on street after being frustrated, rejected from a job, he looks frustrated and exhausted while repeatedly researching and speaking to his friend on phone. Let's try to generate the results like that. And meanwhile, let's count the words, as well. If it fits in, that's great. How many words are these 45 words? Now, let's listen to this. Now, you might have noticed that it has covered all the dialogues, but it has made the piece faster. So if you want it to be in the minimum piece or the understandable piece, you have to reduce the words. Otherwise, you can keep it as it is, as well. You can use that as well if you want to save credits. Now, let's extend it. So as you can see that this is person two dialogues, who was searching for jobs. Let's extend that. I'll just click on plus sign, and you will see how seamlessly it will extend the clip. Without any error, without any consistency problems, we can simply extend and make the conversation for person one first or person two first. So this means we can cover almost three lines. Okay. So let's use that and it will extend the scene. Now, what we will do is we will first extend the whole scene for person one and then we will create it for person two, and then we will use CapCut to divide those scenes. But this way, you will see that we can cover most of the conversation in a very simple way using very less credits and it even takes lesser time than you think. Okay, in the previous parts, the internal audio was silent. So let's listen to this clip again. Bro, I'm seriously fed up. I've been applying for jobs everywhere. Every interview ends the same way. They say I don't have experience. Exactly. Where will I get experience if nobody gives me an opportunity? So you might have noticed that it has covered the maximum words, which is almost 45 words in a single stretch using only ten grades, but it has maze the conversation speed faster. So if you want the conversation to be in the moderate speed, then you need to reduce the number of words. Otherwise, it can continue the conversation this way. And simply we have extended the clip, and you will see the consistency goes on and we can cover all of the dialogues for one single person. But I have the degree, but employers want practical skills and real world experience. I've thought about it, but honestly, I don't know where to start. Bro, I'm seriously fed up. Where to start. Now, the second clip almost ended on where to start. So if we extend the clip from here, we will mention account tech. Where did you learn from these courses? Okay. So we click on the Extend, and we will continue now from here, okay? So we'll continue from here, mention all of these conversations. And this way, you will use and utilize the credits in a much better way. I've been app. Now, let's continue. Now where to start. What did you learn from their courses? Really? That sounds like exactly what I need right now. That could actually help me become more confident during interview. Yes. Okay, so this is what it has already covered. Let's extend it even further. But if you want to keep it simple, the editing, you want to keep it simple, then you have to divide it with chat BT using in such a way that one person finished the conversation and the next person joins in, and each one is using maximum words of 40 every time so that you can download every single clip separately, and then you have to combine it in CapCut. But since we are trying to cover the maximum credits and the dialogues are between the person one and then person two joins in then person one, but this is all in a single clip, so you have to separate it yourself. But that's not a problem. I'll just show you right now. Let's click on the plus sign, click on Extend. And this is a complete practical project that I'm showing you right here. Let's continue views. Wow, that's great, man. So you really recommend those courses? Thanks, Bro. Hopefully, this can finally help me get my first job, Bro. So let's download it, although last clip didn't turn out that well because a girl appears in front as well, and he's not also talking on the phone, and he's holding the resume in a weird style. So you can modify that. But I'm just here to show you how the two parts of the conversation works. Okay, so Bro, I'm seriously fed up. Yeah. Bro, I'm seriously fed up. I've been applying for jobs everyday. Now, if for some reason it has failed to export, you can download individual clips as well. People conversation. Let me just say that. One, graduate frustrated. So you'll just mention it too. It's very easy to manage. So this one is two. You also do have the sequence available like 74, six, seven, 47, so you should have no problem in arranging the clips. Let's make it three. So these are only three time credits that we use. One, two, three, and four. Okay. Let's just download that. This is the fourth one. Okay, so this is person one. Let's make a folder of Person two as well. Okay, and this time we will start a new project, okay? So what we'll do is click on New Project. You can actually upload a character. I didn't see if that character. So this is my character, too. We can download that in the original size. And now let's use it in the new project. Upload the image, add it to the prompt, and then we will start with person two dialog or person one Dialogue in this case. So this way, you can easily make the two sided conversation very easily with using minimum credits. Let me just tell you how we can combine those clips as well. But just use the three dialogues. Let's read the video. And the next time when you want to combine those clips, you can also define that's a quick code that each conversation can contain almost 20 to 25 words safely. So you can try using a 25 words conversation from each side, and then you should be able to download and combine all of those in CapCut very easily. Really? What's happening? You sound extremely frustrated. But you're a fresh graduate, right? What exactly are they expecting? Yeah, that's remake this. Let's say, I want to remake this. This didn't work nice. The later part was, that's almost fine, but I wanted a better result. Really? What's happening? You sound extremely frustrated. But you're a fresh graduate, right? What exactly are they expecting? Yeah. Okay, for the sake of example, let's use that. Let's actually extend the clip. Or what can we do? Okay, we forgot to define the scene that this is the scene, make it a heading. And this is the dialogue, okay? Talking on phone call. Rather than receiving, let's make it. He's already talking. Now, let's play it. Really? What's happening? You sound extremely frustrated. But you're a fresh graduate, right? What exactly are they expecting? Yeah, that's honestly one of the biggest problems fresh graduates face. So this is much better. Let's see the conversation. Phase. Up till phase, it has covered that successfully, and now we have the extend option as well. Let's extend that, and let's use three more conversations, okay? So we'll use it like that. And make sure to use the full stop as well so that we can easily define those clips as well. It is not a continuous conversation, but it'll take a pause here. So three more dialogues. Really? What's happening? You should start building Okay, so let's listen this. Biggest problems fresh graduates face. Then you should start building practical skills alongside your degree. Actually, I recently joined some courses from account tech training. Practical skills alongside your degree. Okay, so the later part didn't work well, so we can cut out that part and use and continue with actually seeing he continues explaining his friend on phone. Let's make it a dialogue, okay? Actually, we want to extend that. So plus extend. And then we mentioned the prompt here because it was using Omniflash, so we didn't want that. Tied your degree. Actually, I recently joined some courses from account tech training. So that's the frustrating part that it doesn't pick it quite well. I don't know why. But let's actually save this for now. And this is how you can extend the conversations and then mix those clips. Let me just show you in the next video how we can merge those clips or start editing it in CapCut. Very easy process. So that's how it is done. Just write by yourself, and I will see you in the next one. 23. Combining Clips in CapCut: Video, we are going to learn how we can combine multiple clips that we generated from Google Veo three into CapCut, and that's a very easy process. Now, what's the benefit of it and why we are not doing it directly in Google Veo three or Omni. That's because it is using too much credits, and what we have come up with that if we have two persons dialogue, what we can do is we can generate the dialogues for each person separately, all dialogues of person one, and then all dialogues of person two. Now, since this is a conversation between the two, firstly, per person one speaks, then person two, just like so, and we will divide and cut the clips in CapCut separately so that not only it will use very less credits, also the videos will be very consistent as you can see the conversation. It didn't change the face. Character consistency is awesome. And this way, you can actually combine a bigger clip with a much simpler method. So let's open up CapCut. You can download it for your PC, and let's actually import the clips. Go to the downloads, and here is MyPe conversation. Let's import all of that on open and it didn't open up for the second one. Okay. So now we want to use that. Let's use the first one. Bro, I'm seriously fed up. I've been applying for jobs everywhere. Every interview ends the same way. They say I don't have experience. Exactly. We Okay, so exactly, you need to cut that here. So actually, we will split it because we want to use it later. Let's move it up or down wherever you want. And then let's use the second scene. Let's listen to that. Really? What's happening? Seriously fed up. I've been applying for jobs everywhere. Every interview ends Seriously fed up. I've been applying for jobs everywhere. Every interview ends the same way. They say I don't have experience. Bro, I'm seriously fed up. I've been Okay, so Bro, I'm seriously fed up. So, Bro, I'm seriously fed up. Then we will use the second clip Rey, What's happening? Bro, I'm seriously fed up. Move it up. Really? I've been Okay. Really? What's happening? Okay, let's cut it. Let's join it with this clip now. I've been applying for jobs everywhere. Every interview ends the same way. They say I don't have experience. Okay, let's use that. Here now you can see the result. No, I'm seriously fed up. Really? What's happening? I've been applying. You can extend the part a little bit if you feel that it's cutting too early. Bro, I'm seriously fed up. Billy? What's happening? You fed up. Really? What's happening? It's happening. I've been applying for jobs. So like, you can see that we extended both clips, and now it is much better. Bro, I'm seriously fed up. Really? What's happening? I've been applying for jobs everywhere. Every interview ends the same way. They say I don't have experience. You sound extremely frustrated. But you're a fresh graduate, right? What exactly are they expecting? Okay, so we end up the conversation there. And now the first person. Exactly. Where will I get experience if nobody gives me an opportunity? Let's pace it like that. So I'm just quickly trying to show you how the dialogues will be arranged. Seriously fed up. Really? What's happening? I've been applying for jobs everywhere. Every interview ends the same way. They say I don't have experience. You sound extremely frustrated. But you're a fresh graduate, right? What exactly are they expecting? Exactly. Where will I get experience if nobody gives me an opportunity? Yeah, that's honestly one of the biggest problems fresh graduates face. Then you should start building practice Okay, Edgewood's face. Okay, so we will cut the next part here, and let's join it with the second conversation. Like this one. Okay, so although it's not perfect, but let me just show you the results and how you can combine those clips. Bro, I'm seriously fed up. Really? What's happening? I've been applying for jobs everywhere. Every interview ends the same way. They say I don't have experience. You sound extremely frustrated. But you're a fresh graduate, right? What exactly are they expecting? Exactly. Where will I get experience if nobody gives me an opportunity? So this is how you can join those clips into CapCut very easily just by swapping through different areas. You can use the cut clip. So for example, one conversation is ended here, let's say, what we'll do is just split that. You will keep it at the bottom, or you can move it around wherever you want. Then the second clip, let's say we use that, we will cut it from here and wherever we want to use that part, we can adjust it accordingly in the timeline. After that, you can just export that and there you have the two people conversation clips created fully with Google Veo three using the List credits and simply with the help of Chat GBT, you can achieve that. Try this by yourself, and I will see you in the next one. 24. Maintaining Character Consistency in VEO 3: Let's talk about the character consistency now. So in this video, we are going to bring all of the stuff that we have learned together, whether it's ingredients, whether it's image generation, whether it's video making. We are going to see and use all of these tools to maintain the character consistency, okay? So, actually, let's start a new project from the main screen. So what we do is, let's say that we open up the settings right here and just notice that the agent is turned off only then you will see the settings. So I'll just set it to video, and we will choose ingredients as well. And just to use lesser credits, I'll just make sure that only one result should be published. And for the reference, you can select your own avatar or you can use any character. I'm going to use let's upload actually one character. So actually, what we'll do is we'll open up any of the previous project and we download our character. If you're unable to reference that, you can upload it directly from here. Let's upload that. This is my character. Click on Add to Prompt. We click on Add to Prompt, and after that, it is loaded in the new project. And I'll just define what camera angle I want. Let's say I want a wide shot First of all, just mentioned the short type. And right, this man is walking in Tobi mall. Or let's take him outside is walking in a park on a bright sunny day. And he suddenly stops in middle and take out his wallet to check how much money he have. So just a random scene that I defined, along with a couple of descriptions, how the short type should be what our character is doing on any specific location. How is the day and what is the scene or character doing. Now here, since we have also added ingredients, we could mention every single ingredients that he'll be holding, what kind of watch he's wearing, what kind of wallet he takes out. We can mention all of that, but I'll just keep it as it is just to see its own creativity and how the results are coming out on the basis of its own creativity, right? So let's process that Now, sometimes it might give you this failed message that this prompt might violate our policies about generating prominent people. Please try different prompt or send a feedback. You have not been charged for this gendon. So it is detecting that it is a famous person. However, we have created this AI individual. So I just mentioned here, this person attach is AI created by me, and let's process that. Okay, it again got failed prompts. What I've done is either you can attach some other model which is not similar or not a look alike of any celebrity. I guess this person is. So I have used my own Avatar white shirt. This man is walking in the park on a bright sunny day, and he suddenly stops and checks his wallet. Let's see. Let's see what I've got left. I definitely need to find an ATM. So this is a result. As you can see, you can use your own Avatar, as well. And that's pretty nice. Let's see what I've got left. I definitely need to find an ATM. Perfect. Now, let's actually try an other avatar, as well. Let's use. Let me just upload that one more avatar, upload media. So let's actually choose this person now. So this time, I just mentioned and attach the person and mention that make this AI person walking shot. Let's process that. And now let's look at the results. So this is nice, very nice shot. And now the key of character consistency is, let's say that we want a backshot of this person. From behind, I'll just mention that you will see that as long as the character is attached in the chat, it will maintain that character consistency. This is the result. Let's see, as you can see, the person is similar, it has successfully maintained the character consistency. Now what we can do is we can add both of these to our scene Builder I'll just choose that and click on the three dots, add to scene, and let's create a new scene and add to scene. The scene which is created, I'll just choose that here for the second shot as well. As you can see, both of these clips are added to my scene. Now let's go inside the Scene Builder, and let's see that the scene should start from behind. So I can rearrange that in the scene Builder. I'll just only take a few angles just like so. Please this one first. Just going to take this much of the scene. Let's reduce that to this point. And then I'll just pass this much of the scene from here, reduce that. It will both combine. And you can expand that according to your liking, wherever you see the transition accurate. So you can choose that. So whatever you do, as you can see, you can adjust the transitions according to your liking. And let's say that we want to add a scene, a very close up shot in his wallet. I'll just add it here and add a dialogue Where's my credit card? Oh, it's right here, something like that. So we want to use that same reference image every single time. So we just click on this plus sign within the scene builder, and we'll choose that man. And we can also take the reference of this frame. So we can save that frame, and we can now attach that frame here. So in this scene, we can only attach the frame as well. A very close up shot in his wallet. And he tries to find credit card and says in a wording tone, where is my credit card? Oh, it's right here. And then he finds it. I'll just add that and says, Oh, it's right here. And you can also put the dialogue in the quotation marks. So it delivers that exactly in the same way. So once you want to continue a story in the scene Builder, you can just add the last frame. You can save that frame whenever you pause the video, you can save that frame. You can use that as a reference. But one thing you can keep in mind that whenever you want to use that as a reference, although you can add it like this, but you have to click on this plus sign and add a clip, and then add it to the Prompt so that it will extend it further. And let's mention the details again. So I mentioned the scene again. Close up shot in his wallet, he says in a worrying tone, Oh my God, here's my ATM card, and then he finds it and takes a calm breath and says, Oh, it's right here. And let's look at the result. Basically what it does it since you have used the previous frame reference, it has combined it into the second frame, and that is very, very nice. So let's actually play it. Scene changes. He stops and says. So very nice, very nice. So this is how you can maintain the character consistency. Now, let's try something crazy adding to this scene. Let's say that we want to use this saved frame. We want to save this frame, and we want to magically transform this man into a woman wearing the same clothes. Can we do that? Let's see. Frame is saved. We can use that in the reference. Let's use that. This is the frame. Let's say, certainly, I'm bright. Magical light comes out of his wallet. He's worried. What's happening. It takes the shape of white dust, white, magical dust, let's say. Covering his whole body. And then he transforms into a woman wearing same clothes. And we'll remove the rest of details. Let's see what happens. Now, one thing I forgot to do here, you must click on the plus sign to create a new scene or the extension of it. Otherwise, it might replace the previous clip as well. So in this case, it might replace that. Alright, so like I said that it replaced the earlier clip, and it is trying to replace that as you can see. So what is even better is whatever project you're working on where you have all of these clips, let me just show you from starting whatever project you have opened. Basically, this is the Man walking project. Now, because we have used this image as first, that's why it's showing as a title image, so don't be confused. So what we can do is we can actually take the reference. Simply from here and add it to the prompt, and it will maintain the character consistency. And we just write the prompt again. Now, this time, it will keep it totally separate, so you can easily arrange or rearrange that in the scene Builder. So let's actually, what we'll do is we'll delete this scene altogether. The second one because I think that has not generated any of the scene. Okay, so here is the result, and you can see that it has generated that separately, very nice. But the earlier scene was having some problems. So what I want to do is we want to add this one to the scene. So we just open that up and we'll add the rest of the clips behind that, we can do that, as well. So click on Add clip and we will choose the first scene, Add to Media, and you will see that it will automatically create a new scene. It will switch to the Scene Builder where you can rearrange all the clips. So now two clips are added. We'll first add this one. And when this is added, you can add the second one, click on the plus sign, add this two media, and we'll place it here. You can reduce that. It will come up like this. And before even closing the wallet, he sees something strange. You can see like this that is nice. And then he weirdly transforms into a girl. And he's very happy about it. So this is how you can maintain the character consistency, join multiple clips and maintain it in whatever angle you want. So you can create a full story like that, like you see nowadays on YouTube. So this is how you can maintain the character consistency just by using the frame referencing on every single extend, and you can build all of that into the scene. So whatever project you are in, you can open up any clip and basically add the clips to the later part. Also, what we can do is we can just use the frames as well. So let's say that we have captured the girl picture. We just take that as a reference saving to the frame. And now we have used that before as well, but let's take this. Instead of ingredients, we'll switch to frame. We'll choose a starting frame, and we'll choose the ending frame. But so we have generated the scene. We have made the starting scene, the frame reference, we'll choose that and we take the girl as a frame and magical light comes out of his wallet and suddenly white dust particles covers his whole body and transforms into a woman. He's worried about this transition, and he transformed into a woman, worriedly says what just happened. So this is another sample of the two clips, if you want this kind of transition clips or any other frame references joining into the second. You can use the frames as well. But the main part is the frame referencing. This is the only code to maintain the character consistency throughout, whether you want it in the scene Builder, whether you want it into individual clips or whether you do it in the frames. All over the place, it will maintain the character consistency, and you will just see it in a moment. Let's see that. As you can see, it has perfectly maintained the reference. Ish opened. And he's still very happy about how he transformed. So anyways, these are the little things that you can even cut from your clip to create your story. But as you can see, it's working perfectly fine. And you can create awesome stories from that. So this is all about character consistency. Just try this by yourself, and I will see you in the next one. 25. Make Interior Design videos in VEO 3: In this video, we are going to learn how we can create the room transformations or interior design transformations, time laps videos that you see on YouTube. That looks something like this. We have a blank hall, very raw, something like this, and the workers comes in and they design the room just like this. So let's splay it. So you will see the workers coming in. They're working on it. And then they converted it into a finish form. Complete Interior Design videos. And the final design might look something like this. It starts from the raw hall, and then they construct it. They put the ceiling, tile design, create a sitting area. And you can also see that I have extended the clip beyond eight to 10 seconds. So let's see how we can create that in Google Veo three. But before even starting, let's analyze and see how much views these videos are getting. So this is one example of our channel structural aesthetics. It only has published 89 videos, and within 89 videos, they have gained 388 K subscribers. And if we search by popular, these videos are getting a crazy amount of views, as you can see, 23 million, 30,000,060 3 million 249 million even. So you can create these kind of videos easily by just using the simple prompt. And we are going to see some more examples on how we can improvise that as well. So let's see this oldest video that he has published. So if you basically check out the description, the oldest videos was just published like a few months ago. So it's not even one year, and he's getting a lot of views. So I'm going to guide you complete steps on how we can create similar kind of videos. So let's go to Google Flow and let's actually start a new project. I'll just go to home, and from here, you can start a new project. Click on that. And on the other side, let's disclose that. We will open Gemini, or you can open hat GBD as well if you like. So to create these kind of amazing videos, I have compiled a very detailed prompt for that, and one prompt will do the job. You can easily create these kind of videos just by using this one prompt. So you can get this in the description or resources section of this video. So let's actually copy that. And we will just go to whatever we want to use. Like I said, hat GBT, grog, Gemini, whatever you want to use, you can just paste it there. Now, I'm using the Pro version, but you can use that or easily create from the free version as well. Just make sure to copy this from the start to end. So first of all, what it will do is it will give us some of the ideas that we can create. Luxury interior design concepts, living room design, master bedroom design, smart kitchen design, bathroom design, executive office design. We have the dining room design. Closet design cinema theater design, sun room design, basement, and we have a lot of other options as well. So you just have to choose, please select one design option number. You just have to type that number, and that's it. So for example, I want to design a master bedroom. I can choose two. And we'll start creating images for that. So these are all the steps. If you want it to be in more detail, we can write make it five steps. So this was a room and the structure starts building. Then this is the third and fourth step and fifth step. And basically what you get is this kind of grid image. If we check that, these are all the pictures on a single frame, but we don't want that. So you try make it five steps. It still combiles it into a grid. But I just wrote a couple of times, and then I defined five steps individual images. It says, here are five individual representing images of the complete transformation and the labels you requested. So this is the initial state. This is the construction has started, the base and structure and all the things, labor starts working. They have put on the padding. And now they have finished the room. So these are all the individual images that you can download. Let's quickly download that individual images. And once all the images are downloaded, you can keep it into a project because all of these appear like that. So I can just cut that and I can save this into a project folder, paste all the images here. Let's make it full screen. Alright. So after that, we will go to Google Flow, and we will select this settings. We'll select video, you'll select frame, and we will select nine by 16 or 16 by nine, whatever you like for your content. So I can choose that. You can choose 8 seconds or 10 seconds. I'll just choose 10 seconds for here. And if you're unable to select those options, you just have to turn off the agent, make sure that it is turned gray and not white. And then you can start selecting the frames. So I'll just name that as room transformation, okay? So this is our project. We'll select the start frame. Actually, let's click on the upload media and let's upload all of these together. If you want convenience, you can choose and define that what is the step one or what is step two, and so on. So this is one. This is two, three, four, and this is five. All right? You can upload like that as well. So however you like, once the upload is complete, what you'll do is just click on the start button. Make sure you have selected frames and video, and also 10 seconds and 720 GB. This is the complete settings, okay? So click on Start and we will choose the first clip. This is the initial state, and this is the infrastructural and structural work. Done. And by the way, for this step, I think 6 seconds would be fine as well. So it will take us lesser credits so you can demonstrate in 6 seconds as well. Now what we can do is we can define the step one and step two, and also we can ask Gemini to define it to define those steps. First of all, let's try it like this, and let's read the video. And I'll just fast forward the process and show you the final results because it's just processing. Okay, I've just checked the process, but it is not what we need because it shows us the work in progress, but no labor is doing that. So instead, what you can do is look at our initial prompt. So this is the transition that you can pick or you can ask to show that as well. So this is a renovation process prompt. Transition Image one to Image two add workers in the fast time lapse motion, surface preparing and structural work, begin installing luxury materials, no visible paces, realistic construction logic. Okay? So what we do is since we have turned this into five steps, And we'll also see all steps attached in pictures so that they can define the steps properly as well. Let's take the reference, and we'll also write that you have to write the prompt Okay. Now, let's send it and you will get some detailed prompts that you can use. So this is step one to step two. I'll just copy that and paste it here. Send it. And you can see that it turns out great just the way we want it. Perfect. But the problem is they have included the music as well. So we can turn off the music sound effects. No music. And look at the second result now, just the way we want it, it is only with the sound effects and no music. So this is the video that we wanted. After that, let's do the second one. And actually, what we will do is we will compile that frame by frame. So this is the second prompt, and we will again start the first frame or let's say, second, and we will select the third frame. And again, 6 seconds. I'll just read this clip. And this is the second clip. As you can see, all the sound effects are there, but with no music, that's exactly what we wanted. And let's continue to the third scene. So we have used that. Step three to step four, copy that. I'll just choose step three, add to the prompt, step four, add to the prompt. And I'll just keep this all working sounds, no music. And on the top of it, I'll just paste this prompt. Just make sure you select the right one. This is the third step. Okay. Let's process that. And again, this is a 6 seconds clip on a seven NU B. Now, this is our third clip. Let's quickly extend it to the fourth clip. But let's just see the progress, and it's all looking great. They're installing the paddings and so on. And then we have this one. Step four to step five. And basically, what you notice is the finished form has curtains, whereas in this one, it has placed some paddings here. So you can edit that part. And now let's look at this. So this is the finished form. I like that. So you can cut off this part and use the other one. After that, what you can do is we have the master part as well, continuous end to end five step video. Actually, we don't want that. So what we'll do is we'll pick up the steps one by one. So this was the first one. What we can do is just click on this and we will start bedroom design video. This was the first one. We can click on Add clip, add the clip, and you have to check and identify the second step. I think it's that one. So as soon as you add the second clip, I've just chosen the second one. It will create a scene, and then you can define the scene. Let's say, bedroom scene. Okay. Now what we can also do is we can go Flow, go to any of our existing project. And for example, this was the clip I want to add into the scene. So we can click on three dots, click on Add to scene, and click on bedroom scene. So it will add it there. Quickly switch to Scene Builder after that. And we see the clip added right here. So we can remove the earlier clip because it didn't have the sound, and now it works like this. So this is the first scene. And you can remove the extra part where it is starting so that it can match the other portion. Let's say I want to cut of this part. You can do that as well. Let's actually continue and add all the clips first. Click on the plus sign again. This is the third one. Now, this one is the next clip, so I'll just click on three dots, click on add to the bedroom scene. And when it's added, we'll click on Swish to Scene Builder. And as we can see it is added. Let's go to the flow again. And this is the final step, so I'll use that add to the scene, and bedroom scene. Switch to Scene Builder, and we'll see the extension appear here. So this is the final video. You can see that they start cleaning up. They start building the structure, and then they continue to build the structure, and you can cut off this part, actually. So you can reduce that actually like this, and now it looks fine. Also, let's reduce this part, like I said, that once you add all the clips, you should be able to tend and cut some of the clips. So like this, we can do their finishing, put curtains, put the furniture, and this is the final form. What we can also do is you just go do Gemini. Or what we can do is we can save this to frame, and then we can use that frame. So for example, I go back, and this time what we'll do is we'll not actually the frames to video. We'll do the simple ingredients, and we can just add that frame. And we want to just show the complete walk around of the room. So it's nine by 16, 6 seconds and all. Let's render that and we will finally add this one to the scene as well. And what we do is we will close the project, whatever is open, and we'll go to this final look, click on the three dots, click on add to scene and click on bedroom scene. And now it is added. Let's click on Switch to Scene Builder. And this is our final results. Let's look at it. I hope it is good. So it started cleaning, building the structure, continue the work, like the padding and all, and then they install the tiles, curtains. This is the bed, and this is the final look. So this is amazing. This is how you can create all of these viral videos. After that, you can click on Download, and it will start exporting scene. Now, once your scene is exported, it will look something like this. We can click on that, and this is the final result. So this looks quite amazing. Continue the work, pairing and all, install the tiles, put the curtains, install the bed, and this is the final finished form. So this is just a quick example. I'm sure that when you work on it professionally, you can make much better videos than this one. But this is how you can make this room transformation viral videos or interior viral videos. In Google Veo three, try this by yourself, and I will see you in the next one. 26. Create Construction TimeLapse Video and Conclusion: This video, we are going to see how we can build these kind of construction AI videos that you see on the screen. It is very simple to do, and all you need is a simple single picture of a house or a building or a property. And that's it. You need simple prompts, and all of these prompts, I'll attach with the resources section of this video, so you can check that out. So let's get started, and let's see how we can build these kind of amazing videos. Now to create these kind of viral videos, we need two tools. One is Google Flow and the other is nano Banana. Now, if you are using Google Flow, you can access nano banana inside that as well. It has a built in function. If we go to Flow, you click on the new project and you can see that you can generate the images and the videos as well. So when you click on Images, it is automatically switched to nano Banana and you can choose any version of that. As you can see, we have Pro, we have nanobnana two and nanobnana Light. Also we can generate a lot of videos with Google Flow. So let's actually begin and let's see how we can get our first image to construct or make those kind of videos. Now to search these kind of images, either I can go to Pinterest or I can also go to Google Search simply, and you will see a lot of pictures that you can use. So let's say that I want to build for a Villa. I'll just write here, and we will get lots of amazing videos. Just make sure that you get the front view so that you can show the entire property construction steps. So let's say I want to use this one, I'll just click on the three dots and click on Download Image and my image is downloaded right away. And by the way, guys, we can also create abandoned houses videos that's being viral. Nowadays on YouTube, you might have seen the abundant Villa like these ones, so you can get these kind of pictures with Google Flow by entering the simple commands, and you will get the couple of images. So you can show the restoration process as well. But right now we are just creating the Construction TimeLapse videos. So it's your choice if you want to convert this into an abandoned place like before and after, you can show this one as after and this one as before, and you show all the restoration process. Okay, so we will continue with our example and let's open up a new project again, and we will use this image, add it to the prompt, add to the prompt like this. And first of all, we'll convert it to 1920 by 1080. So that we can create a vertical video. And as we can see, I got the result in 1920 by ten ATP. Let's go the second step of the construction. And first of all, what we'll do is from the finished picture, we will work backwards. So this is the final result. We will start working backwards. So the last step that you do in your house is the landscaping, like putting grass plants and all the things, parking your car and so on. So I need to make this all scene raw. Also, we will turn on the lighting at the end. So we want a prompt for that. So I have created a prompt that you can use. You will get this with the description of this video. So just check that out. First of all, I'll just reference this image. This prompt, it says a single building on an active construction site surrounded by raw barren dirt. The building exterior should look slightly messy and unfinished, with all windows unlit and dark, complete removal of all landscaping plant, grass pathways in vehicles, pretty realistic construction environment. Also you have the choice that if you want to convert it into nano Banana Pro to get even better results, you can do that as well. Let's try that. This is one step back and we'll start working towards backwards, and then we will compile all of that into a video. So this is perfect, guys. As you can see, this is the finished one, and this is the previous one. Very realistic looking, very nice. Now let's continue working backwards. And now we already have the reference image right here. Otherwise, what you can do is you can use that in the prompts. Just click on the three dots and click on Reuse in the prompt and you will get that. So after that, we'll move to our second prompt, and that is this, we want to convert it into a pure scaffolding site where the construction is happening. We'll piece this one and send it. Let's wait for the result. Again, the result look quite amazing. As you can see, it has converted it to a pure construction site with the same structure, same everything. You can try this again and it will give you a different results. So whatever you like, you can use that. And this time, you can see that we got a different result. This is also nice. So whichever you want to use, you can use that. I think this one looks more realistic, so we'll use that. Whatever is the extra prompts, you can just delete those images so that we don't have any problem in arranging all of these images, right? So that is all done, and now let's move forward or actually backwards to work on the foundation process. And for that, again, we will remove all of this prompt and make sure that we have the first reference image intact, and you can start working backwards. You can also use the most recent step or actually download this image. You can upscale that as well. And once it is downloaded, you can upload it again using plus sign upload and you add it to the prompt. After that, one step back is the foundation stage where we want to show this building in its foundation stage where the trenches are dug out, beam is laid and concrete columns with rods are visible. Do not change the view. Okay? So let's process that, and you will get this prompt as well with the resources section of this video. You can check all of these prompts in the Word file. So again, the results are pretty amazing. You can see that the foundation is laid out like this. And let's see that we want to reduce our work. What you can use is the first prompt that we used earlier, and let's try with that prompt as well. Whatever is getting the better result, we'll use that. So as you can see, the neighborhood looks the same in both of the pictures. This house right here appears in the second picture as well, and the house on the right is also visible. I believe this one is much better than the earlier one. So we can move this through trash. And after that, let's use this image again, click on the plus sign, the construction side, and now we want to convert this into a plain land or plot. And now with reference to the last image which is this one, I want to convert it into just the plot site. So we want to remove everything. So we attach basically the image from here which we have downloaded and then re uploaded. And we use this prompt. Remove all the construction mud, footing, trenches, columns, and show this as a plain flat, empty side with sparks uneven grass and weed. Let's process that so look at this guys. This looks perfect, just the way we want it. So we can use that as the first reference image. Now, let's actually start downloading that. Meanwhile, we'll upscale this to two K as well. And let's start putting it into the folder. House construction. This is one. Let's download the second one, and I'll quickly rename all of those and save this into a folder. So this is one, two, then we have three. You can use any one of them. Then we have four, and then we have five. I'll cut it from here, pieced it here, name it as two. Then we have three. This is four. And actually, you can see in fourth image, the site looks a little bit different. So we'll download that and we'll take the reference of this image. And we'll ask make the neighborhood angle, and visibility of the second image like the first one. So if that works, that is great. Otherwise, we'll continue with that as well. And look at this. This is much better and much more realistic than the earlier one. Let's download that. And this one is stage four, and then Image five, we will turn that as well like this. So you can download that first and upload it again to make it identical. Okay, so I'll reattach that, click on Upload and upload this site once again. This is the finished side. This is basically the fifth step because we are working from backwards in the video. And make neighborhood angle and visibility of the second image like the first one. Let's continue. And after a couple of re tries, this angle works. As you can see in some of the prompts, it didn't change anything. It didn't make the neighborhood similar. And after a couple of tries, it has generated this result, but this time it changed the overall design or color theme. So I reprocess, and I got that. So this is the finished prompt that I will mention in the resources section again, or I'll even mention the improve one. Okay, so let's download this final image in two K. Save it, and there comes our fifth picture. So once we got all of our pictures or frames, it's time to animate them into a video. And for that, we will just use a simple prompt that will continue throughout all of that frames. And even if it is a change, it will be a very simple prompt. I'll mention all of those prompts in the resources as well. So let's proceed with the most interesting part, and that is animating all of these pictures into a time lapse. So we will basically go to the home tab and open up a new project on Google Flow. And from here at the bottom, you will see this nano baanana option or some other option. Make sure the agent is turned off, otherwise, when you click on this specific area, it will open up the chat box like this. So let's turn it off, and let's just choose videos from here and we will choose frames to videos. Now it's your choice if you want to create 16 by nine or nine by 16, but all of our frames are in the horizontal form, right? So we will use this 16 by nine ratio. 720 P is fine. And we can choose 6 seconds, 8 seconds, and so on. But let's just choose 6 seconds. I'll just generate one result. It will only cost us ten credits. Now I'm using the pro version, but when you sign up with the new account, you will also get 150 credits for free, and that should be enough to create this kind of viral videos. But obviously, after that, you have to purchase a plan. So let's select the start image. Actually, let's upload that first. I'll just choose the images, and let's upload all of these into the project. Go to the house construction. I'll choose all of these, and it will be added right here. And after uploading all of the frames, we will use it one by one. First of all, we will start with the plain land, right? So we will click on the video and we'll choose frames. We'll choose the starting frame of site one, that is this one, and we will choose the ending frame from one to second portion where the Foundation is being laid. So we'll add it to the prompt. And like I said, I'll use very simple prompt, and I'll give you that as well. So this is the prompt, realistic progression time lapse of an empty ground into progress construction. Workers machinery carrying materials move into the frame, no cuts, no jumps, real time, motion. And we can create 6 seconds of the clip, 8 seconds of the clip or however you want. So let's actually generate 8 seconds of the clip, which is maximum. And you can use the fast version light version or quality version, whatever you like. So let's choose fast because this is better. You'll choose 16 by nine, and it is of 8 seconds. Let's read the clip. And actually, I have tried that, but it doesn't look very realistic, as you can see, it has constructed some other thing and then finished the clip onto this side, which is not right. So what we do is we'll choose video frames again and we'll choose Omni flash because it gives us much more realistic results. And let's make it 6 seconds of the clip because we have to attach five more clips to it. And as you can see, the video turned out really great. You can download this to check that. It will also upscale that video. And you can see the video looks quite nice. We can use that. I'll make this video one. And let's quickly generate more videos like this using the same formula. Now this time, I'll just remove the frames from start and end. Now the second frame which we used, which is the foundation site will become the starting point and the ending point will be the next step after that. So next step is this one. I'll click on Add to Prompt. And remember that you might have to generate the video multiple times until you get your results, and you can download that as well to check if that meets your requirements. Okay and after that, we will just use simple prompts like this one. Workers and machinery working to complete the second stage or up to the second stage. We'll just write simple prompts, and this time, let's use Frames and OmniFlash 720 DP. Let's do that because it is using lesser credits. So let's generate that. Let's look at the results. Let's play it. Okay. That looks quite nice, and that is exactly what we wanted. Let's download that upscale 210 ATP. I've generated one other result as well. So whatever you like, you can use that. Let's download this one as well. And meanwhile, let's read our other clips. Now, this gray structure becomes the start and the end point will be this one, okay? And we will use the same command and process that. Same settings, go ahead. Meanwhile, let's use it again. We will use the fourth picture and we'll complete it to the final stage. And this is our result for the earlier one, which is now partially done. We want to finalize everything, so I just mentioned. This simple command, workers finalize everything, including lawn preparation, plantation, lighting. Different workers are working on different tasks. And finally, car parks in working to complete up to second stage and closing scene with evening time. And finally, a car parks in working to complete up to the second stage and closing scene with evening time in house lights turns on and a peaceful environment is created. No music, only sound effects. Okay. Let's read that. As you can see that I have just used my random wordings. I'll just attach that as well. So you can use that. Meanwhile, we'll download this clip. And this is our final result, as you can see that almost everything is finalized, and finally, the car parks in in a peaceful environment. So we can download that as well. Let's quickly rearrange the clips. First was this one. Second is this one. I'll place it here. Then we have this one. I'll just name it as three. And finally, the fourth one, which is right here. So this is the final step for and we can even make some changes finally owner car parks in to see everything. Let's render that. Okay, so these are all the clips. Now is the time to stitch those clips into one. And for that, you can basically use any editing software. I'll use CapCut. Let's create a new project and we will simply import all of those clips. You can rearrange those clips one by one. You can drag and drop it from the top as well. And actually, it's ideal to cut off the last part of each clip so that the transition is smooth. And whatever part you see as not matching, you can cut off all of that part. Now to export this, we will click on the Export button on the top right hand corner. You will choose your location and you will choose the resolution and click Export. I've already exported that, so let's look at the final result. So let's play it, and it looks quite interesting. Look at this from plot to construction to base and everything is done step by step. This is how you can make these kind of amazing construction videos, looks very nice. And the final steps are going on. This is it. Amazing, right? So this is how you can make all of these kind of construction or interior designing videos. And I've shown you a couple of examples of what can be done with Google Veo three, and that brings us to the end of the course. Thank you so much for joining in. I hope you have learned something valuable. And if you did, just remember to leave on our course. It will be very, very helpful. So take care and good luck with your learning journey.