Here are my findings so far Pika (watermarked tron video) pros, good quality, easy interface cons, only 30 free credits a day, videos can't be longer than 3 seconds, doesn't handle certain types of movement well Stable Diffusion+AE/Lockdown/Mocha tracking (unmarked tron clip) pros, very flexible, as powerful as your skills with AE are cons, tedious, takes a long time to make even a short clip ComfyUI video inpainting (alita clip) https://github.com/bruefire/ComfyUI-SeqImageLoader Currently experimenting. So far it doesn't handle longer videos very well, it looks like its trying to interpolate between random images. If I can solve that I'll have to figure out a way to isolate the image2image area so I can inpaint larger videos without sending the entire thing to stable diffusion or find a workflow that can upscale+blend the morph with the original video Has anyone else done any experiments?