subject_definitions: <Video 1> is the source footage whose… | Deepfake AI видео

Промпт (текст в видео), по которому сгенерирован этот ролик

subject_definitions: <Video 1> is the source footage whose environment, camera movement, action timing, background, lighting and live-action photographic look must be preserved exactly. <Picture 1> is the reference image defining the identity, facial features, hairstyle, costume, horns and rendering style of <Subject 1>. <Subject 1> must be rendered in exactly the same artistic style as <Picture 1>. <Subject 1> is the chibi-proportioned anime devil girl from the first pass (design from <Picture 1>), rendered in the chibi Japanese anime cel-shaded style of <Picture 1> with thick black outlines and flat cel shading: oversized head about one third of her total height, small compact body, short limbs, pale skin, large red eyes, silver-white chin-length bob with a thin red streak in the fringe, a pair of red devil horns with dark bands at their base, white cropped futuristic jacket with black trim and shoulder armor, black armored chest piece over a black high collar, black armored gloves, black tactical belt with gold buckle, black thigh-holster straps, white trousers with knee pads, red chunky ankle boots. summary: [video editing + reference generation, refinement pass] Refine the first-pass result of <Video 1> with <Subject 1>. Identity, chibi proportions, costume, horns, camera, framing, motion, timing, background, lighting and the live-action photographic look must remain exactly as already established. retention_analysis: <Video 1>: fully_preserved - environment, camera, motion, timing, background, lighting and live-action photographic look, identical to the first pass. <Subject 1>: fully_preserved - chibi proportions, facial features, red horns, hair, costume and chibi Japanese anime style, identical to the first pass; only fine rendering detail is refined. detailed_description: [Style opening] <Subject 1> keeps the chibi Japanese anime cel-shaded style of <Picture 1>; the environment, background, lighting and photographic live-action look keep the original style of <Video 1>. [Shot 1] In this refinement pass, <Subject 1> continues the same dance from the same low upward camera angle and framing, with identical choreography, rhythm, timing and direction. Do not change her identity, her super-deformed proportions, her costume, her horns, the background, the doorway, the shelving, the lighting or any other person in the frame. Only linework, cel-shading edges and fine costume detail are refined. overall_soundscape: N/A non_diegetic_music: N/A