← sources

Suggestions for a ComfyUI Minimax H3 Workflow for 45 clips - 3 minutes

source: reddit (reddit:StableDiffusion)published 2026-10-23 05:23ingested 2026-10-24 02:38

retained text

So I’ve been rendering a 3-minute conversation manually in 3-5 second clips, prompting and rendering each clip separately and using da Vinci resolve to edit.

But the clips are the same scene of 2 people talking, from different camera angles: wide, zoomed, over the shoulder, etc. the image character references are the same but with different audio clips per video.

Any comfyui workflows where I can put 45 shots in 1 giant prompt which lets comfyui or local AI to slice it up into smaller shots with separate references per clip and create the 45 output videos which I’ll throw into davinci?

document fetched 2026-10-24 00:32 · canonical https://reddit.com/r/StableDiffusion/comments/1xd5ah5/suggestions_for_a_comfyui_minimax_h3_workflow_for