Creating Characters and Environments
Axel starts by creating a character sheet for all his characters. He uses a t-pose as a reference image, which he uploads to Discord and then adds the rest of the character descriptions. He then creates a close-up of the face and uses a similar prompt structure to generate a medium shot of the character blended with the environment. To direct the characters, he slightly changes the prompt.
To keep the look of the environments consistent, he first creates a main reference image and then uses the pan and zoom tool to create new views from it. He also shows how to create entirely new shots in the same location by using a similar prompt structure and using the original image as an image prompt.
Animating Characters
Axel then brings everything together by merging his characters with the environment. He uses the zoom out or pan function to change the framing of the shot and adds emotions or other actions by slightly changing the prompt. To make the characters move, he uses Runway Gen 2’s new feature to generate videos out of image prompts.
To animate the characters’ faces, Axel uses 11 labs to generate voices and then imports the voice file into Adobe Premiere. He then uses the Wave2Lip Google Colab app to generate lip movement by uploading a video and a voice audio file. He also shows how to fix broken AI faces with GFP Gan and Code Former.
Enhancing Shots
To bring his shots to the next level, Axel adds AI-generated stock footage and flowery colored lens flares. He also shows how to use DaVinci Resolve’s D Flicker tool and Repair Image Node to make the footage look smoother.
Main Takeaways
In this video, AI expert Axel showed us how to use AI tools to generate entire movies in a surprisingly simple and fast way. He showed us how to create characters and environments, how to animate characters and how to enhance shots. He also gave us some tips and tricks to help us get the best results. So if you’re ready to make your own movie, this video is a great place to start!