プロンプト文
The camera follows the man in black as he flees quickly, followed by a group of people chasing him. The camera switches to side tracking. The character panics and knocks down a fruit stand on the roadside, gets up and continues to run away. The crowd makes panicked sounds.
参照入力
1 バリエーション
シネマティック動画 のすべてのプロンプト
すべて表示 →What this shot looks like
A man in a black coat sprints from a chasing crowd. The camera tracks behind him, then switches to a side-tracking move as he knocks over a roadside fruit stand mid-stride, picks himself up, and runs on. Panicked voices and the clatter of spilled produce fill the soundscape.
It's a compact action-film grammar exercise: protagonist flight, environmental destruction, and a camera-angle switch — all in one prompt.
When to use it
This is a strong evaluation prompt for AI video models:
- Chase-sequence test — does the model hold subject continuity (same coat, same silhouette) across the camera cut?
- Physical-object interaction — knocking over a fruit stand requires fruit, stand, momentum, and a recovery beat
- Directed sound test — many models still under-deliver on prompted ambience; this one explicitly asks for crowd panic noise on top of the action
Tips for customizing
- Swap "black coat" with another wardrobe call (red hoodie, white shirt) to test how rigidly the model honours costume detail
- Replace "fruit stand" with a different obstacle (newsstand, bicycle, café table) — same beat, different physics
- Add a city tag ("Hong Kong wet market", "Tokyo back alley") to test cultural set-dressing
Related
- More cinematic video prompts
- See all video models we've tested