After the scene is filmed/photographed then one can re-position and re-point a virtual camera and have it correctly render the scene. And do so with higher quality results than photogrammetry and NeRF techniques.