[Duration, e.g. 10]-second scene. Reference image 1 is [character A]; reference image 2 is [character B]; voice reference 1 is A's voice, voice reference 2 is B's voice. Setting: [location]. Medium two-shot, slow push-in. A says: [line 1]. B pauses, then replies: [line 2]. Ambient sound: [ambience]. No music. Natural lip sync, steady handheld camera.