Why doesn't MiniMax H3 understand simple, direct instructions in prompts?
retained text
So I've been trying to gen a simple video where a subject is simply standing toward the camera, and I want them to be framed so I can see them from head to upper thighs. If I prompt something like "<Subject 1> is standing and facing the camera, and you can see them from head to thighs." their head is cropped out of the video, and you can only see them from the chest to legs. But if I change the prompt to something like "<Subject 1> is standing and facing the camera, and you can see them from head to thighs. They are looking directly at the viewer and smiling." THEN their head is included in the shot. ...why do I have to prompt it where the body part has to be doing something/showing something before it's included in the gen?
document fetched 2026-11-02 00:29 · canonical https://reddit.com/r/StableDiffusion/comments/1xlaj6u/why_doesnt_minimax_h3_understand_simple_direct