Galerie de prompts IA OpenZhenZhen · gpt-image-2.5 · image

Instantané de rue flou sous le flash nocturne de Tokyo

Blurred Street Snapshot Under Tokyo Night Flash

Instantané de rue flou sous le flash nocturne de Tokyo — découvrez cette création avec gpt-image-2.5 sur OpenZhenZhen, son prompt original et un exemple de génération à réutiliser.

Instantané de rue flou sous le flash nocturne de Tokyo — AI image creation
Instantané de rue flou sous le flash nocturne de Tokyo

Prompt

Prompt chinois

以用户上传的参考图作为主要可见人物的面部身份依据。在不与身份参考冲突的情况下，保留来源的场景、造型、服装、姿势、构图、光线和视觉风格。 创作一张9:16竖幅、超写实的智能手机抓拍照片：夜晚的东京安静人行道旁有一栋白色瓷砖墙建筑。一位明确成年的女性以半侧身从镜头前走过，正处于迈步中，像是在匆忙偶然的瞬间被拍下。她本能地对拍摄作出反应：边走边惊讶地回头，一只手抬到镜头附近部分遮挡镜头；身体继续向前，略微移出画面，构图刻意不完美。她神情转瞬即逝，略显尴尬又俏皮，羞涩克制地忍住笑意，绝非摆拍。 使用固定参考ref-10a-v1保持女性的面部身份。如果用户提供可选的视觉灵感附件，只将其用于宽泛、非身份性的氛围启发，不复制或匹配其中的脸、身份或身体。若没有该附件，也可依据此处文字场景直接生成；它不是必须生成的素材。 使用手持不稳、匆忙捕捉的现代智能手机拍摄，呈现强烈且不受控制的相机抖动。抬起的手、头发和脸部带有强烈方向性运动模糊；面部细节可因运动而拖影、条带化但仍隐约可感，身体周围可有轻微重影。这是摄影运动造成的模糊，不是增加或缺失身体结构。仅使用运动中触发的闪光灯：皮肤和手上出现刺眼闪光高光，周围深暗，曝光不均，高光过曝并形成硬阴影。白色瓷砖墙被拉成条纹光带，街景则逐渐沉入黑暗、几乎没有可读细节。画面噪点极多、严重模糊、原始而不完美，像偶然留下的短暂手机记忆；不要电影感、编辑感、精致感或摆拍感。避免动漫、插画、绘画、风格化渲染、影棚或美颜布光、柔和肖像光、清晰对焦、干净锐利的脸、磨皮、时尚摆拍、专业或三脚架摄影、身份复制、面部重塑、放大眼睛、夸张身体曲线、额外或缺失手指、扭曲手臂肩膀或颈部、文字和水印。

Prompt anglais

Use the uploaded reference image as the facial identity reference for the main visible person. Keep the source scene, styling, clothing, pose, composition, lighting and visual style unless they conflict with the identity reference. Create an ultra-realistic, candid 9:16 vertical smartphone photograph on a quiet Tokyo city sidewalk at night, beside a building with white ceramic-tile walls. Show one clearly adult woman walking past the camera in partial profile, captured mid-step in a rushed, accidental moment. She reacts instinctively to being photographed: turning her head back while still walking, as if surprised, and raising one hand close to the lens to partially block it. Her body continues forward and carries slightly out of frame, leaving imperfect framing. Her expression is fleetingly embarrassed and playful, with a shy restrained smile as she suppresses laughter; she is never posed. Preserve the woman's facial identity with fixed reference ref-10a-v1. If the user supplies an optional visual-inspiration attachment, use it only for broad non-identifying atmosphere and never to copy or match a face, identity or body. If absent, proceed from this written scene; the attachment is not a required generated asset. Capture on a modern smartphone held in a rushed, unstable manner, with strong uncontrolled camera shake. Apply heavy directional motion blur across the raised hand, hair and face; facial details may smear and streak with movement, remaining faintly perceptible, and the moving body may show slight ghosting. The blur is photographic motion, not extra or missing anatomy. Use only a flash firing mid-motion: harsh flash highlights on skin and hand, deep surrounding darkness, uneven exposure, blown highlights and hard shadows. Stretch the white tile wall into streaked light bands while street surroundings fade into darkness with minimal readable detail. Make the image extremely noisy and heavily blurred, raw and imperfect like a fleeting accidental phone memory—not cinematic, editorial, polished or staged. Avoid anime, illustration, painting, stylized rendering, studio or beauty lighting, soft portrait lighting, sharp focus, a clean sharp face, airbrushed skin, fashion posing, professional or tripod photography, identity copying, face reshaping, enlarged eyes, exaggerated body curves, extra or missing fingers, distorted arms, shoulders or neck, text and watermark.

Voir le modèle API

Découvrir plus de prompts image et vidéo · Générer des images · Générer des vidéos · Créer sur le canevas IA