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

窄巷酒馆木凳三人肖像

Portrait of Three on a Wooden Bench in a Narrow Alley Tavern

明确成年的日本女性与两位年长成年男性坐在狭窄的复古酒吧卡座中，门框构图、琥珀色灯泡与细腻胶片颗粒构成安静的纪实肖像。

窄巷酒馆木凳三人肖像 — AI image creation
窄巷酒馆木凳三人肖像

提示词 / Prompt

中文提示词

创作一张3:4竖幅复古纪实肖像，场景依照来源提示词主体内容设定为狭窄的日式旧酒吧。画面中央是一位身份参考 ref-10a-v1 中明确成年的日本女性，栗色直发扎成马尾并留齐刘海，妆容低调；她穿着得体的浅蓝色翻领衬衫、深绿色窄领带、藏蓝色过膝裙、白手套和简洁黑鞋。她端坐在窄木凳上，双膝并拢，戴手套的双手轻轻交叠在膝间小玻璃杯上方，下巴略低，平静地看向镜头。她两侧各坐着一位年长成年男性配角，穿无标志、无徽章的复古风格日常夹克或衬衫；左侧人物从门边稍稍向内探身，右侧人物安静地交叠双臂。由入口向内拍摄，让磨损的木门框形成画中框。场景包含狭窄木墙、磨损的清漆木梁、浅架上的酒瓶与玻璃器皿、少量模糊不可读的手写纸条和价签，以及头顶的线缆与小物件。一两个裸露白炽灯泡投下局部琥珀色光池，光线迅速衰减到暗部；木面有细微反光，空气略显潮湿。以约1至1.5米的亲密平视距离拍摄，呈现35毫米镜头观感，紧凑地将三位成年人和逼仄卡座置于画面中央。采用暖色低饱和色调、轻微高光晕染、细腻胶片颗粒和浅景深，营造怀旧抓拍氛围但不过度模糊。避免可识别警徽或执法标志、模仿特定摄影师、可读文字、品牌、水印和错误手部结构。

English prompt

Create a 3:4 vertical vintage documentary portrait in a cramped Japanese bar, following the consistent scene in the source prompt body. At center is the clearly adult Japanese woman in configured identity reference ref-10a-v1, with chestnut-brown straight hair in a ponytail and blunt bangs, understated makeup, wearing a modest light-blue collared blouse, dark green narrow tie, navy below-knee skirt, white gloves and simple black shoes. She sits composed on a narrow wooden bench, knees together, gloved hands lightly overlapped above a small tumbler between her knees, chin slightly lowered as she looks calmly toward the camera. An older adult male supporting character sits on either side in unmarked period-inspired everyday jackets or shirts, without badges or insignia; the left figure leans slightly inward from the doorway, while the right quietly folds his arms. Shoot inward from the entrance so worn wooden jambs create a frame within the frame. Include narrow wood walls, scuffed varnished beams, bottles and glassware on shallow shelves, a few indistinct handwritten slips and price tags, and cords and small objects overhead. One or two bare incandescent bulbs cast localized amber pools that fade quickly into shadow; the wood catches faint reflections and the air feels slightly humid. Use an intimate eye-level view from about 1–1.5 m with a 35 mm lens character, tightly centering all three adults and the confined booth. Apply warm low-saturation color, gentle highlight bloom, fine film grain and shallow depth of field for a nostalgic candid mood without excessive blur. Avoid identifiable police or law-enforcement insignia, photographer imitation, legible writing, brands, watermark and malformed hands.

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