OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

同一人物的九幕电影感时尚摄影

Nine Cinematic Fashion Portraits of the Same Person

以附图为同一人物与世界观的唯一参考，生成九张彼此独立且构图各异的实拍电影感照片，并排组成3×3叙事网格；继承人物与影像质感，但不复制附图镜头。

同一人物的九幕电影感时尚摄影 — AI image creation
同一人物的九幕电影感时尚摄影

提示词 / Prompt

中文提示词

将配置参考ref-10a-v1作为同一人物与同一摄影世界观的唯一参考。忠实继承人物面容、发型、体型、服装、皮肤质感、光线质感、色调、胶片感、空气感及照片所呈现的世界观；但不再现或复制参考图的构图、姿势、相机位置、角度、拍摄距离、裁剪或画面编排。将人物与世界重构为仿佛存在于一部长片之中，不只是时尚照片变体。创作从电影中截取的九个瞬间，排列为3×3网格。每格都是独立完整的照片，同时整体如同同一部电影的片段。九个瞬间呈现明确不同的情境，例如等待、凝视、移动、静止或远处存在；不要固定具体地点、小道具或事件，而应在参考图可解读的世界观范围内自然展开。九格构图必须全部不同：有意识改变拍摄距离、人物大小与画面占比、相机高度和角度、身体裁剪方式及人物与空间的关系。混合眼睛、脸、手、手指、脚、腿的局部超特写；身体局部裁剪；背影；中距离肖像；全身照；位于边缘或角落的小人物；广阔空间中的微小人物；以空间或景观为主的画面；以及人物仅占很小部分的电影式广角。明确对比人物充满画面与人物仅占极小面积的镜头。结合平视、低角度、高角度、远距离、人物背后及透过空间观察人物等视角。多用侧面和背影，不重复正面镜头；部分画面可以只见身体、让脸离开画框，或只让人物一部分入镜。绝不重复相似面部特写、姿势、拍摄距离、相机角度或身体裁剪方式；不得只改变脸朝向或视线方向来制作近似镜头。每一格单独观看都应是有力量的作品。将整组导演成相机偶然目击的电影场景，而非摆姿势的人物照片：通过视线、身体朝向、光影、空间与距离暗示前后剧情，用留白和沉默进行电影式表达。交替安静与戏剧性瞬间、强对比画面与安静细腻画面。采用电影照明而非均匀唯美照明：大胆使用强光与深影、局部逆光、剪影、仅照亮人物一部分的瞬间、人物从暗部浮现、远处射入的光线及表现时段变化的光线。让光影传递紧张、宁静、余韵、孤独、期待、不安与解放感，不依赖说明文字。最终呈现真实实拍照片，兼具高级时尚编辑品质与电影场景感；使用细腻35mm胶片颗粒、自然人体、真实皮肤、微妙光线阶调、电影对比度、浅景深、立体阴影和统一色彩分级。避免人工化电影海报风、CG感与过度处理。不要添加说明文字、标签、水印或logo。

English prompt

Use configured reference ref-10a-v1 as the sole reference for the same character and the same photographic world. Faithfully inherit the character's facial features, hairstyle, body type, clothing, skin texture, lighting texture, color tone, film feel, atmosphere and photographic worldview. Do not reproduce or copy the reference image's composition, pose, camera position, angle, shooting distance, crop or frame arrangement. Reconstruct the character and world as if they exist in a feature film, not merely as a fashion-photo variation. Create nine moments captured from a movie, arranged as a 3×3 grid. Each panel must be an independently complete photograph, while the whole set feels like fragments of one film. Let the nine moments suggest distinct situations—waiting, looking, moving, resting or existing at a distance—without fixing specific locations, small props or events; let them unfold naturally within the reference's interpretable worldview. Make all nine compositions different: deliberately vary shooting distance, character scale/occupancy, camera height and angle, body crop, and relationship to space. Mix extreme partial close-ups of eyes/face/hands/fingers/feet/legs; partial-body crops; back views; mid-range portraits; full-body views; small figures at frame edges/corners; tiny figures in vast space; landscape-dominant views; and cinematic wide angles where the character occupies very little of frame. Clearly contrast frame-filling views with tiny-subject views. Combine eye-level, low, high, distant, behind-the-character and through-space viewpoints. Favor side profiles and backs over repeated frontal views; some frames may show only the body, exclude the face, or include only part of the character. Never repeat a similar facial close-up, pose, shooting distance, camera angle or body-cropping method; do not make near-duplicates that differ only in face orientation or gaze. Each panel must stand as a powerful work alone. Direct the set as scenes accidentally witnessed by a camera, not posed character photos: use gaze, body orientation, light, shadow, space and distance to suggest what happened before and after, with cinematic direction through blank space and silence. Mix quiet and dramatic moments, strong contrast and delicate images. Use cinematic lighting rather than uniform beauty illumination: bold strong light and deep shadows, partial backlighting, silhouettes, light catching only part of the character, the character emerging from shadow, distant shafts of light and cues of time-of-day change. Let light and shadow carry tension, tranquility, lingering resonance, loneliness, anticipation, anxiety and liberation without explanatory text. Deliver authentic live-action photographs with high-fashion editorial quality, like a sophisticated story between fashion editorial and film scenes; fine 35mm grain, natural human bodies, real skin, subtle light gradations, cinematic contrast, shallow depth of field, dimensional shadows and unified color grading. Avoid artificial movie-poster styling, CG appearance and excessive processing. No added captions, labels, watermark or logo.

查看 API 模型

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作