MiniMax H3 提示词 - 门后观察到的红白装扮角色扮演者

Reference image 1 defines the adult cosplayer’s identity, facial features, high ponytail hairstyle, red-and-white Mai Shiranui-inspired costume, black wrist guards, gold-trimmed sash, fabric textures, proportions, and overall candid photography style. Create a staged, consent-based hidden-camera-style cosplay video in a busy indoor public setting, filmed as if the camera operator is discreetly observing from behind a doorway and nearby foreground objects. Preserve the character’s appearance and costume throughout. 0–2 seconds｜Medium full shot｜Eye-level angle from behind a doorway｜85mm telephoto｜Slow lateral slide: the adult cosplayer stands off-center, partially obscured by a blurred doorframe and a dark foreground object, looking down at her smartphone while casually adjusting her ponytail with her other hand. Natural window light, shallow depth of field, realistic skin and fabric texture, slightly imperfect handheld framing. 2–4 seconds｜Medium close-up｜Slight side angle｜135mm telephoto｜Gentle push-in: she finishes adjusting her hair, lowers her hand, and briefly glances toward the phone screen. The background remains compressed and softly blurred, with subtle reflections in nearby glass and a passerby moving indistinctly behind her. 4–6 seconds｜Close-up｜Over-the-foreground-obstruction angle｜135mm telephoto｜Tiny handheld reframing movement: she notices the camera’s presence, turns her eyes and then her face slightly toward the lens with a subtle surprised-but-playful expression, holding the phone naturally at chest level. End on her recognizable face and upper costume, still partially hidden behind the foreground obstruction. Continuous visual continuity: the same adult cosplayer, hairstyle, red-and-white costume, accessories, phone, lighting direction, and background environment remain consistent. Candid staged cosplay photography, spontaneous observational timing, natural body movement, realistic motion blur, mild phone-camera noise, soft autofocus breathing, slight exposure fluctuation, imperfect off-center composition, no dramatic posing. Ambient indoor sounds only: distant footsteps, quiet crowd murmur, faint phone handling sounds, no dialogue, no music, no subtitles or logos. Negative constraints: no minors, no sexualized camera emphasis, no voyeuristic framing, no nudity, no wardrobe malfunction, no anatomy distortion, no extra fingers, no duplicated phone, no changing costume colors, no hairstyle drift, no face distortion, no sudden camera cuts, no teleporting objects, no exaggerated acting, no text overlays, no watermark.