For pinpoint camera and action precision beyond what a text prompt alone can specify, record yourself performing the desired action and transfer that motion onto your character, with scene control turned on. This is a performance-capture-style alternative to describing motion in words — useful when a prompt like "she picks up the cup and turns" can't reliably produce the exact timing or physicality wanted.
Complements Who/Where/What/Camera/Mood Prompt Structure: where that framework covers what to specify in a text prompt, motion control replaces the "what" and "camera" elements with a literal recorded performance instead of a description.