Готовый промт для кинематографичного ролика на Невском: камера спускается от глобуса Дома Зингера к переходу, повторяя движение исходного видео, а в кадре — ты, в своём образе и с живой улыбкой.
Шесть референсов. Загружайте строго по порядку — в промте они называются @ref1…@ref6.
Только камера и движение тела: ракурсы, переходы, тайминг, скорость, ритм шага. Внешность, мимика, одежда, телефон и локация не берутся.
Лицо и главная эмоция: узнаваемая улыбка, форма губ, живые глаза. Одежда и фон с фото не копируются.
Ещё один ракурс лица — черты и пропорции держатся во всех ракурсах камеры.
Весь образ: костюм, верхняя одежда, обувь, сумка и длина волос — по виду спереди и сзади.
Уровень улицы: переход через Невский, здания, канал и мост — расстановка всей сцены.
Стартовый кадр сверху: башня Дома Зингера с глобусом на переднем плане, за ней Казанский собор.
Create a photorealistic cinematic video by transferring the camera choreography and subject movement from @ref1, replacing the man with the woman in @ref2 and @ref3, dressing her in @ref4, and replacing the entire environment with the Saint Petersburg location in @ref5 and @ref6. REFERENCE ASSIGNMENTS @ref1 — CAMERA AND BODY MOVEMENT ONLY. Use its complete camera sequence, framing, transitions, timing, speed changes, walking rhythm, footwork and body movement. Do not copy the man's appearance, facial expression, clothing, phone or the original location. @ref2 and @ref3 — THE WOMAN'S REAL IDENTITY AND EXPRESSION. Use these photographs for her face, natural proportions, recognizable smile and lively expression. Do not copy the red clothing or indoor backgrounds. @ref4 — OUTFIT, ACCESSORIES AND HAIR LENGTH. Use the front and rear clothing views. Preserve all garment details, footwear, handbag and very long hair. @ref5 — STREET-LEVEL LOCATION AND SPATIAL LAYOUT. Use the crossing, street, buildings, canal and bridge arrangement shown here. @ref6 — OPENING AERIAL VIEW. Use the Singer House tower with its glass globe in the foreground and Kazan Cathedral beyond it. COMPLETE LOCATION REPLACEMENT FROM FRAME ONE Every frame must take place in the same Saint Petersburg location, beginning with the very first aerial frame. Completely replace the original video's buildings, skyline, street canyon, road layout and location-specific vehicles. Do not begin in the original city. Do not gradually introduce Saint Petersburg. Do not morph or transition between two cities. Do not retain the original city's scenery during fast camera movements or close-ups. The location is Nevsky Prospekt beside Singer House, also known as the House of Books, and Kazan Cathedral, near the Griboyedov Canal. The cathedral is KAZAN CATHEDRAL with its curved colonnade, as shown in @ref5 and @ref6. Do not substitute Saint Isaac's Cathedral or combine different landmarks. OPENING CAMERA DESCENT Begin high beside the glass globe on the ornate Singer House rooftop tower. Establish Kazan Cathedral beyond it, matching the spatial relationships in @ref6. Apply the opening camera descent from @ref1 to this setting: preserve its duration, acceleration, downward movement and approach to street level. Descend through open air beside the tower toward the pedestrian crossing across Nevsky Prospekt shown in @ref5. Reach the woman at the corresponding moment in @ref1 as she walks across the crossing. Do not add a separate orbit around the globe or an extended architectural introduction. The globe establishes the opening position; the source video determines the pace. PRESERVE THE COMPLETE CAMERA SEQUENCE After the opening descent, transfer the camera choreography from @ref1 relative to the woman throughout the rest of the video. Preserve every distinct movement, viewing angle and transition in its original order and at its corresponding time: the lateral full-body tracking; the approach to the face; the low tracking beside the feet and clothing; the overhead top-down view; the return toward the face and frontal portrait; the wider view of the moving subject; the final rising pullback. Follow the actual source video whenever its camera movement is more specific than this description. Preserve changes of camera height, distance, viewing side, speed and acceleration. Preserve the source framing and the subject's position within the frame. Do not omit or reorder transitions. Do not replace the sequence with prolonged side tracking. Do not simplify it into a generic orbit. Do not add static beauty shots. Preserve the source video's transition style and continuity. TRANSFER THE SUBJECT'S MOVEMENT Replace the original man completely with the woman from @ref2 and @ref3. Transfer his walking rhythm, stride timing, footwork, weight shifts, torso movement and progression through the crossing to her, adapted naturally to her own proportions and clothing. She follows one continuous believable route along the zebra crossing. She does not reverse direction or rotate her body simply because the camera moves around her. The handbag remains in her LEFT hand throughout. Adapt the left arm naturally to carrying it. Preserve the remaining movement rhythm without copying phone-handling gestures. Her feet make believable contact with the road. No sliding, floating or teleportation. FACIAL IDENTITY Use @ref2 and @ref3 as the only facial identity references. Preserve her recognizable facial structure and natural age: facial length-to-width ratio, forehead, eye shape and spacing, nose projection, cheeks, lips, jawline and chin. Do not widen, flatten or vertically compress her face. Do not redesign her features or replace her with a generic fashion model. Do not transfer facial features from the man in @ref1. Maintain her likeness through frontal, three-quarter, profile, overhead and low-angle views. Preserve natural facial depth and perspective during camera movement. Preserve her natural body proportions without artificial slimming, stretching or changing the head-to-body ratio. FACIAL ACTING — WARM, HAPPY AND ALIVE The video's camera and body movement come from @ref1, but the woman's facial acting comes from @ref2 and @ref3. She is relaxed, happy and confident. She enjoys being filmed. Use her real smile in @ref2 as the main expression reference, including the characteristic shape of her lips and the natural movement of her cheeks and eyes. While walking, she has a gentle, spontaneous smile. When the camera approaches her face, she briefly meets the lens with a warmer smile, then comfortably looks ahead again. Include natural blinking, small eye movements, breathing and subtle changes of expression. Do not copy the man's serious expression, fixed gaze or facial acting. No frozen face, blank stare, rigid mannequin behavior or permanent exaggerated grin. No talking, singing or lip-sync. She does not continuously turn her head to follow the camera around her. OUTFIT AND HAIR Reproduce the complete outfit from @ref4: the dark olive avant-garde suit; the pronounced lapels and elongated asymmetric jacket panels; the sculptural trousers with deep folds; the long pale olive shaggy fur coat; the warm suede boots with fluffy toe details; the black-and-cream houndstooth handbag. Keep all clothing colors, shapes, construction and accessories consistent throughout. Her brown hair is loose, very thick and wavy, reaching the top of her buttocks, as shown in the rear view of @ref4. No ponytail, bun, hat or glasses. No red top. No phone. Hair, coat and fabric move naturally with her steps and a light breeze. The handbag stays in the same hand and remains visible whenever the framing allows. FIXED GEOGRAPHY Build one coherent environment from @ref5 and @ref6. Maintain the same physical relationships between Singer House, Kazan Cathedral, Nevsky Prospekt, the crossing, pavements, canal and bridge throughout. Buildings and road markings remain fixed in world space. Camera movement changes their perspective and visibility, never their location. Show landmarks only when the camera is actually facing them. Do not move a cathedral or tower behind the woman to keep it visible. No duplicated buildings, mirrored street layout, moving zebra crossing, warped bridge or changing road width. Keep the camera in open space without passing through buildings, vehicles, street furniture or the woman. LIGHTING AND REALISM Use consistent soft daylight throughout the entire take. The location references provide architecture and layout, not a sequence of changing weather conditions. Maintain natural skin texture, realistic facial volume, detailed fabric and fur, coherent shadows and believable street scale. Background pedestrians move naturally. Vehicles yield while she crosses and remain physically separate from her route. OUTPUT Match the duration, aspect ratio and overall pacing of @ref1. No captions, text overlays, logos or watermarks. The final result must retain the full camera choreography and subject movement of @ref1 while showing this specific smiling woman in the outfit from @ref4, within the Saint Petersburg location from @ref5 and @ref6, from the first frame to the last.
Открой бота, загрузи референсы в указанном порядке и вставь промпт.