Loading page
Shot library · Group
Video frames — walking, turning, gesturing, a camera move.
Motion is where prompt-only personas die. A model interpolating between frames will re-imagine a face it was never confident about, so this group assumes a trained route or a locked first frame.
0 stills
0 niche-bound
180 min setup
4 recipes
Full body walking toward a static camera at chest height, eight metres to two
Chest up, starting in profile, camera static on an 85mm
Chest up, static camera, hands entering the lower third
Medium wide, static camera, chair in the centre
Questions
Motion is where prompt-only personas die. A model interpolating between frames will re-imagine a face it was never confident about, so this group assumes a trained route or a locked first frame.
At least the trained model route. The only route that holds up for video and for anything shot from behind, in profile, or in bad light. It costs a training run and a storage decision, and it locks you to the model family you trained on — a LoRA built for one base model is not portable to another.