Try: “Vertical 9:16 cinematic fashion transformation video featuring the person in the reference photo. Scene one: warm bedroom lighting, chest-up camera, the person playfully adjusts their hair, mouths a short phrase, and raises two fingers like small devil horns. Scene two: two white door panels move inward from both sides, covering the frame in a clean transition. Scene three: reveal the same person on a windy city street at night, wearing an oversized black leather jacket, with softly illuminated buildings behind them. The person looks confidently at the camera as their hair moves naturally in the wind. Preserve facial identity, natural hands, consistent proportions, realistic lighting, no logos or text.” For a more dependable result, generate the indoor and outdoor scenes separately, then align the door transition with the music beat during editing.