TLDR: Keyframes are done! I’ll be working on and releasing audio/dialogue next! This animation should be around 4 minutes long so it will be the most detailed transformation I’ve tried so far!
I know I sometimes post keyframes, but it feels a bit spoilery idk. I think it would be cool if the first time you see them is in the final animation. I hope that’s okay! I will say, I think I’m going to add the BE tag to this one lol.
This is little test I wanted to try before going any further. Please let me know if you have any feedback! I was testing:
Renat’s voice (ChatGPT Advanced Voice Mode -> Eleven voice changer) - Partial success - Not particularly happy with the voice but I think I just need to roll more voices in Eleven until I get one I like. Conceptually, I think ChatGPT AVM to Eleven is going to be fine. Eleven picked up the accent well!
Runway-ML’s new Act One feature which is similar to Live Portrait and basically lets me transfer video of my face onto the character - Success! - This seems way better than Live Portrait to me. Much more realistic and picks up head movement better. It's not cheap, but worth it I think. Now I just need to work on my acting/lip-syncing lol!
Just generally how Kling would handle the keyframes - Partial success - Kling does not want to hold the camera still. Could be a skill issue but very annoying to have constant camera movement. Hopefully, it doesn't detract too much from the transformation.
So I think I’m going to approach this animation a bit differently since it’s very dialogue-heavy. I’m going to generate the dialogue first and then decide how much animation is needed. I may only need like 30 seconds of Renat sitting there, and then I can just do different facial animations on top of the base animation. Hailey’s animation will, of course, be trickier since the dialogue has to fit properly with the transformation.
I will release the dialogue as an audio skit when it’s done! As always thank you for your patience and your support!