OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image
雨中水坑倒影关系四阶段视觉故事
Four-Stage Visual Story of Rain Puddle Reflections
一个4-image视觉故事系列，每张图片单独生成、各为垂直9:16，不拼贴；以雨街水坑中自然不完美的倒影讲述激情、爱情、冲突与分离四个阶段。
提示词 / Prompt
中文提示词
创建一个4-image视觉故事系列，每张图片必须单独生成，不得组合成拼贴；所有图片均为垂直9:16、电影感、写实、有氛围。每张图的主要主体是雨天街道水坑里的倒影，真实人物只以匿名的局部鞋子、腿或衣角出现在画面边缘。倒影必须自然且不完美，带有波纹、雨滴、湿沥青纹理、水渍和轻微变形，水坑不是完美镜面；现实姿势与方向必须和倒影匹配。保持柔和电影级光线、可信城市环境、微妙叙事感和无文字。四张图使用相同的倒影摄影方法与视觉语言，但发生在不同时间和地点，形成真实关系进展，而不是同一分钟的四个瞬间。图1“激情”：夏日雨后下午街道，一对匿名情侣共用透明伞并靠得很近，倒影亲密温暖，水面相对平静。图2“爱情”：黄昏老街，情侣在雨中安静拥抱，暖光映入水坑，两个倒影几乎融合，情绪温柔平静。图3“冲突”：夜晚寒冷雨街，争吵后的两人分开，一个在伞下、另一个站在雨中，倒影彼此分离，波纹更多、色调更冷。图4“分离”：雨夜街道或车站，一人撑伞静止，另一人转身走远，倒影距离更远并被雨水部分打断，体现距离与失落。每张图片保持单一画面、无文字、无标志、无水印、无清晰可识别面孔。
English prompt
Create a 4-image visual story series; generate each image separately and never combine them into a collage. Every image is vertical 9:16, cinematic, realistic, and atmospheric. The main subject in each image is the reflection in a puddle on a rainy street; real people appear only as anonymous partial shoes, legs, or clothing at the frame edge. Reflections must feel natural and imperfect, with ripples, raindrops, wet asphalt texture, water stains, and slight distortion; the puddle must not look like a perfect mirror. The real-world pose and direction must match the reflected pose and direction. Keep soft cinematic lighting, a believable urban environment, subtle storytelling, and no text. Use the same reflection-photography method and visual language across all four images, but place each scene in a different time and location so the series shows a real relationship progression rather than four moments from the same minute. Image 1, “Passion”: a summer afternoon after rain, an anonymous couple walking very close beneath one transparent umbrella; their reflections are intimate and warm, with relatively calm water. Image 2, “Love”: the couple quietly embraces on an older street at dusk; warm light reflects in the puddle and their reflections almost merge, tender and calm. Image 3, “Conflict”: after an argument on a colder rainy city street at night; the two people are separated, one under the umbrella and the other in the rain, with divided reflections, stronger ripples, and a colder tone. Image 4, “Separation”: a rainy night street or station; one person remains still under an umbrella while the other turns away and walks off, their reflections farther apart and partly broken by rain, conveying distance and loss. Keep each image single-frame, text-free, logo-free, watermark-free, and without clear identifiable faces.
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
相关作品
- 超写实电影感大唐皇后
超写实电影感静帧，1girl，唐代皇后，威仪端庄的年轻女子，高耸双环望仙髻，凤冠缀珠翠与红宝石，额贴花钿，齐胸绛红织金抹胸襦裙，外罩大袖纱衣，金约臂钏与璎珞垂胸，9:16 竖屏构图，人物居中全身站立，背后朱红立柱与殿门向画面深处层层递进形成纵向纵深，两侧仪仗队列虚化，大明宫正殿，MetaHuman 级皮肤质感，次表面散射，可见毛孔与细小绒毛，头发梳理整齐服帖，发丝根根分明、层次清晰，贴合头型，鬓角收拢，无翘起碎发与凌乱飘散，路径追踪全局光照，殿内天光与冷调侧光，物理材质，ACES 色调映射，50mm 镜头，浅景深，轻微胶片颗粒，8K，超精细，焦点锐利
- 超写实电影感大唐美女
超写实电影感静帧，1girl，唐代贵妃，丰腴雍容的年轻女子，高髻盘发插金步摇与珍珠，额间贴花钿，鬓边斜红，齐胸抹胸式石榴红暗花绫罗襦裙，肩颈与锁骨线条清晰，织金纹样与缂丝暗纹细致可见，披帛缠绕臂间自然垂落，9:16 竖屏构图，人物居中偏上，头顶留少量空间，下方完整呈现披帛垂落与裙裾，沉香亭内，身后牡丹与半透轻纱帷幔纵向延伸，MetaHuman 级皮肤质感，次表面散射，可见毛孔与细小绒毛，头发梳理整齐服帖，发丝根根分明、层次清晰，贴合头型，鬓角收拢，无翘起碎发与凌乱飘散，路径追踪全局光照，烛火暖金侧逆光，物理材质，ACES 色调映射，85mm 镜头，浅景深，轻微胶片颗粒，8K，超精细，焦点锐利
- hyperrealistic cinematic girl
hyperrealistic cinematic still, 1girl, cosplay, [character: hair + costume + prop], intricate costume in real fabric, leather and metal hardware, [location], MetaHuman-grade skin with subsurface scattering, visible pores and vellus hair, individual hair strands, path-traced global illumination, phys
- cos girl
生成一张竖屏电影感角色人像：一位年轻亚洲女性 coser，扮演[角色]，全身/七分身站在画面中央，纵向构图，背景是[场景]的局部，只保留肩后的环境氛围。她穿着细节丰富的 cosplay 服装，面料有真实的织物纹理、缝线与金属扣件，配有[道具]。皮肤是真实人像质感，可见毛孔、细小绒毛和次表面散射的通透感。头发为[发色]的[发型]，梳理得整齐服帖，发丝顺滑有光泽、根根分明、层次清晰，贴合头型与脸颊，鬓角自然收到耳后，发梢干净利落，没有翘起的碎发、没有凌乱飘散，也不遮挡面部。镜头 85mm、f/1.8，浅景深，背景柔和虚化，光线为柔和侧逆光加轮廓光，画面带轻微胶片颗粒，8K 超精细，整体为电影静帧质
- 机能girl
cosplay 摄影，年轻亚洲女性，银白色短波波头，黑色机能风外套带蓝色发光电路纹，露指手套，夜晚霓虹街道，地面雨水反光，蓝紫色轮廓光，真实皮肤质感，85mm 镜头，电影感，超精细
- cosplay 摄影写真
cosplay 摄影写真，一位年轻亚洲女性，黑色长发配银质发簪，白色与浅蓝渐变的刺绣汉服，宽袖，手持细长长剑，雪山背景，空中飘雪，电影感柔和轮廓光，真实皮肤质感，浅景深，85mm 镜头，超精细，8K