In this video tutorial, we will explore how to create a short film, trailer, or music video using only Leonardo AI. We will cover image generation, prompt generation tips, image references, motion straight control, and much more.
Leonard is the tool for image generation, so it's not just a text-to-video generator. First, you have to create the image, and then you can animate it. The process is pretty straightforward, so you don't need to have technical skills. Let's see how we can do it.
Register in Leonardo AI and get free 150 credits daily.
Video editing services https://www.liromfilms.com/services-1
Apps and Promos. Check the AI for video, music and V.O generation.
First, go to image generation and type in the basic prompt.
After the image is generated, click on motion and then choose the motion straight.
Basically, it ranges from 1 to 10; one is less motion and ten is more motion for the image. It's hard to tell which level to choose, so the best way is to experiment. As you gain experience, it will become clearer how to choose the motion straight. There aren't many controls, and the video outcome is only 4 seconds long. Maybe it will be improved in the future. So, this is the basic workflow. Then, we can move to the advanced workflow.
The advanced workflow is basically upscaling the image for better quality. After upscaling the image, generate the video from the upscaled image. Keep in mind that the process requires a lot of tokens, so use them wisely.
I suggest reading the tutorial of Leonardo AI about prompt generation. I don't see much sense in repeating all this tutorial that you can read, and it's very well described.
One very cool feature that Leonardo has is the prompt generator. You can input the basic prompt, and then you can generate from it some advanced prompt. This is very useful when you run out of ideas or when you want to experiment or you don't have a clear idea what you want to achieve.. Even if you are good at prompt engineering, sometimes artificial intelligence creates much better prompts than you can imagine.
Also, they have another cool feature when you totally run out of ideas. You can click on The prompt idea generator, click on “IDEATE”", and then it gives you different kinds of ideas, up to eight, that you can use. I'm not sure if it's available in the free version, but in the paid version, it's totally available and it's a very cool feature.
Another cool feature is the image prompt input. You can use another image as guidance for your generations. It's very cool. There are different parameters, but I would like to show you one called "pose to image". You can take one image with, let's say, a character in this image, and then guide the artificial intelligence, in this case, Leonardo Ai, to generate the image with the character in the same pose. It's very, very useful.
The only thing you have to consider is that you probably cannot use any image from the internet because due to copyright restrictions, some stock agencies clearly state that you cannot use their stock images for this kind of thing. Anyway, I don't know if it really may have some copyright issues if you're doing this, but anyway, just be careful. So, don't try to use images from the internet. But anyway, that is not the issue. What you can do is generate the image and use this image as an image prompt, or you can use an image of yourself.
I needed a picture of a character with a gun, so I generated a picture of a soldier and used it as an image prompt for my generation.
The next part is character continuity. The idea is to have the same character across different generations. Currently, Leonardo AI doesn't have a feature where you can click a button and say, "Okay, give me the same character in posterior generations." In this case, what you can do is give a name to the character and use this name in future generations. The issue is that it probably won't give you the exact same character, but at least it will have a similar face or a similar type of character.
Artificial intelligence still has a lot of limitations. For example, I couldn't achieve characters interacting in fighting scenes or with gunshot effects. This aspect still needs a lot of improvement. For now, we can consider that artificial intelligence, at least for now, works well with cinematic shots featuring one character, such as close-ups or medium shots. It also works well with drone shots or master shots for environmental scenes. However, when it comes to interactions between two characters, it still has many limitations.