Transcription
Have you ever wondered how those cinematic AI historical videos are created? Many people see the final result on YouTube, but they don't see the process behind it. The research, the script writing, the voice generation, the image creation, the animation, and the final editing.
In this video, I will show you my complete workflow of creating an AI historical video from a simple idea to a finished YouTube video. I will show you every step, every AI tool I use and how different technologies work together to create a final result. So let's start from the beginning.
The first step is always the story. Before creating any images or video, I need a strong script because the story controls everything that comes afterward. For this, I use Claude AI free plan. I provide Claude with the topic of a historic building I want to cover. For example, a famous landmark from New York, Paris, London or any other city. I asked Claude to create a documentary-style script. The script should include the history of the building, why it was built, who designed it, important events connected to it, interesting facts, a storytelling style that keeps the viewer engaged. So, uh, and I'll provide you the script in the in the comment as well. So let's let's write the script. Okay. So it's a uh the story should include this is the basic structure of the sto story or the script. But uh you have to mention it either you want it to have a shorts or you want it to have a long video. So according to that, you have to tell Claude to write the uh raw script. Okay, by writing this, you already uh you already told Claude that the script should not be exceeding more than 2 minutes. So let's see what it writes. Okay. So now it asks me, happy to write this, which NYC landmark is this script for? Let's say Empire Building. Now it's going to write me a raw script for that. See, now we have a raw script of the Empire State Building. Now we have the hook, we have the setup, we have the build and twist, and last payoff plus CTA.
Okay, now we will move forward to step two. So what we need to do is we need to create the voiceover in order to connect it with our story. So once the script is ready, we will use another website, a free tool which allows us to convert the text to speech into a uh audio file and to use it in our story. The website name is TTS Maker. This is a free tool which allows us to generate the audio using the text. The main thing uh in a video is uh is the audio because it creates the mood of the entire documentary. For historical videos, I usually prefer a calm storytelling style voice that feels similar to a documentary narrator. Let's download the Let's insert the first hook which is two lines. Now we have inserted it. Now what we need to do is we can select a language in which language you wanted to uh generate the audio. But by default, it's selected to English, and obviously, we need uh the voice in English. And here it's going to be voices. We have 148 free voices and uh you can prefer any of it, like if you select this.
>> TTS Maker is a text-to-speech tool, a professional AI voice generator dedicated to delivering.
>> TTS Maker is a free text-to-speech tool that provides speech synthesis services.
>> Temaker is a professional voice generator that uses artificial intelligence technology.
>> This is an example of a gentle motion. Imagine speaking to a child about their day and eye appointment.
>> TSMaker is a free text-to-speech tool that provides speech synthesis services.
Let's select this one. Now what we need to do, we need to type the capture and convert to speech. We'll verify it. Okay, now the conversion is started. Okay. In 1930, two rival businessmen were racing to build the tallest building on Earth.
Okay. So, we have already downloaded our 9 seconds audio. So now the in the same way you are going to download and you are sorry, you are going to convert all the uh text into the audio. For now, uh for our tutorial purpose, let's put it uh only for the hook.
So now we will move forward to step three. The step three is creating audio timing and timestamp because now we have the audio file. So now what we need to do, we need to convert the audio into the timestamp so that we can match with the timing and generate the images. So uh for this, I use another free tool which is TurboAI. You can generate three audio transcription transcription. Uh, you can generate three audio transcription daily. It's a free tool. So we go to open dashboard and now we'll add a transcript here. So, uh, let me check where our audio is. Sorry, I think I didn't download it.
>> TTS Maker is a free text-to-speech tool that provides speech synthesis services.
>> In 1930, two.
Okay. So we will download, we will upload the audio file. Okay. Now the uploading is done. We will go for whale, which is most accurate. We'll transcribe it. It's it's running right now. Okay, now our audio is converted into a timestamp. So as it is just a 9 seconds audio, we only have two timestamps. So what we will do is we will put this timestamp from the transcript and put it in the cloud again and ask Claude here is the timestamp from the audio. Please write me the image prompt for each timestamp. So what it will do, it will write me the image prompts for each scene. For now, it will only generate uh I think two image prompts because our uh audio is only 9 seconds. So here it is. Uh, as it's asking me that you only have these timestamps because I usually generate more than 2 minutes of audio. So that is why, so I said yes. Now, once the timestamps are generated, sorry, the image prompts are generated, now we'll move forward to our next step and which is the main step, which is step three. So now we have two image prompts with us. So what we will do is we will copy these image prompts.
Okay. So now the next step, which is step four, and which is creating AI prompts. Now, as I already copied it, and the website which I'm going to use is Google Flow. You can go for your first link. You click on new project. Now here is the Google Flow. We need to do some settings before before generating before starting uh generating the images. The model should be selected to Nano Banana. Uh, either you wanted to select, either you wanted to generate the images for a long video or for a short video. If it's for a long video, then the ratio should be 16 by 9, and if it's a short video, then it will be 9 by 16. Now, for example, just imagine that right now we only have two prompts, but for example, if you have more than 20 prompts, then if you manually add each prompt, it will take a lot of time and it will be hard to de uh it will be hard to generate all the uh images one by one in a prompt. So what we will do, we will use a bulk image generator in order to automate our process. For that, I use an extension called BulkyGen Flow Automation. You can easily find this uh you can easily find this extension in the Google Extension Google Chrome Extension. Just search it on uh web and you can type BG gen, sorry, Bulky, Bulky Gen Flow Automation. Here it is. When you click on that, it will appear right away. So, uh, it's already installed in my system. So, you just need to click on Add to Chrome, and it will be installed here.
Now, what I will do, I will paste the prompt here, and you need to select the generator. Right now, we are using Google Flow. So, now I'm inserting both my uh prompts. So the main thing is we need to uh we need to put a one-line space between the images. So uh we will just remove this and uh okay, so now these these are our two prompts. So as soon as I start generating, it will start generating and downloading it automatically. So it's the easiest one and the simplest uh Google extension for Google Flow bulk image generation. So as soon as I click on start generation, it starts the process and it automatically sends the image to Google Flow and it starts generating the image, and once it's downloaded, uh, once it's generated, it will download it automatically uh as you can see in the. So now it's almost generated, and as soon as it's generated, it will automatically download to my system, and one important thing, Nano Balanana 2 is completely free and it will not take any credit. If you look at here, generating will use zero credits. That means that image will not be taking any credits in order to generate it. So you can generate as many as you can. I have generated almost 35 images in a single day, and the uh credit did not get exhausted. So it's completely free. So now both the images have been downloaded and automatically uh available in my system. As you can see, both the images are showing up here.
Now we have the images, we have the uh, we have the uh, audio. Now what we have to do? We need to create the animation, which is the last step. Sorry, this is not the last step. This is the seventh step, and after that, it will be editing the video. So now what we need to do, once all the images are downloaded, I organize them into one zip folder. This makes the next step easier because I can upload all the images together instead of selecting them one by one. So let me close this bulk, and we are done with the Google Flow. So let's go here and copy these two images. Go to documents. Create a new folder, new video. Paste both the images here. Let's rename them one and two. Now I'm going to zip it. So that I'm going to compress the video. Now we have the v uh all the images. Now I'm going to paste this zip folder into the Claude again. And what I'm going to ask Claude is to write me a video prompt for these images. So what I'm going to uh write, write me the video prompts for these images according to the timestamp above. So now what Claude will do, it will write me the video prompts for each image. Before uh, before Meta AI blocked the video generation, there was an extension for Meta AI as well to generate the bulk video. But as Meta blocked the video option, they have shifted it to another website which allows us to generate the videos. But we need to do it manually. Hopefully, in a couple of days, we will get the extension for for that website. But for now, we need we have to use it manually.
So now we have the image prompts. So what we will do is we will go to our next step, which is step number eight, animating images into videos. For that, I'm going to I'm using the website which is wibes.ai. This is a Meta website and it allows us to generate the images into video and animate them into a video. So let's open it. I'll go with create new project. Now here it is. So now I'm going to upload both the images which we have created. Image one and image two. Upload it. So you can upload up to 12 images. Now we already have two uploaded media. We will go to our first prompt, which is this. We'll copy it here. Now this is our second image and this is our first image. So we will go to manual animate. And here is another option which you can use auto animate. So that that will help to animate it according to uh wibes.ai. AI uh model. I just put it here and as soon as I hit it, it starts animating the image and once it's done, it will generate the video and then we will download.
Okay. So as you can see, the video has been successfully animated. Sorry, the image has been successfully animated. So now what we can do, we just need to go here and download it. And the download started. Sorry. Okay. So the download started and now we have the audio, we have the video. Same, we will go with this one. The second im uh the second image. We will go here in the cloud. Copy the animation. Go to manual animate. Type it and hit it. It will take some time to animate the image. And once it's done, image conversion has been completed. Now the video is available to download. We'll go to the download button here and click on this. The download started. It will take two, three seconds to download the video. And once the download is done, then we will move to our final step, which is the final uh video editing.
So for that, I will use the free tool which is CapCut. CapCut provides free editing of the video. We will go here and click on edit video. It will take us to a module or interface from where we can download. We can import our media and work on the videos and combine them into a single video and upload it and make it ready for YouTube. So from here, I'm selecting the uh the images, uh the videos which I have just developed. Both of them have been imported. Now I'll upload the audio file which we generated using the TTS Maker. All all of them have been imported into the system. So I'll put the TTSMaker audio file here. And uh it didn't upload it properly. So I'll go and upload it again. This is the uploaded one. I'll drag and drop here the first one and the second one. And now here is my video.
In 1930, two rival businessmen were racing to build the tallest building on Earth. One of them cheated using a trick that's still hidden in this building's design today.
See, that's how you can develop the AI videos using the free tools available on the internet. The last step is to export it. Once you are done with it, you can play with it. You can add uh captions, you can add elements in it, you can add transcript, you can add effects in it. And once you are done with that, you just need to click on export and click on download, and your audio uh and your video will be downloaded into your local system, and from there you can upload it to YouTube. Thank you very much for watching.