PROMPT RADAR / PROMPT平面设计
社交媒体帖子 - Jiangnan Night Hanfu Cinematic Photography
A professional prompt for generating high-fidelity cinematic photography of traditional Chinese Hanfu in a Jiangnan water town setting at night, featuring CCD camera aesthetics and realistic textures.
@drmrzhong 0 热度 0 浏览
ORIGINAL VISUAL / 来源见文末编辑先看
它为什么值得进入灵感库
这条内容值得留下的,不是表面的风格词,而是它提供了一条可以继续实验的画面路径。 先规定信息层级和文字安全区,再谈视觉风格;无法落版的漂亮画面不是完整设计。
01 / 原始方法
完整视觉 Prompt
Theme: A carved wooden corridor overlooking the water on the second floor of a Jiangnan song tower under a moonlit night. A clearly adult woman leans against the carved railing, turning her head to look at the camera after hearing a soft call. Like a cinematic shot, a natural and beautiful Ming Dynasty Jiangnan photoshoot. Subject: Facial identity strictly follows the attached reference image: consistently matching facial features, proportions, skin tone, and overall identity; hairstyle, hair accessories, and expressions follow the current era, scene, and activity, without copying the reference's hairstyle or neutral expression. Body Anchors: She has a petite Korean fashion model figure, about 163cm tall, elegant feminine proportions, naturally charming yet realistic hourglass silhouette. She has naturally full and well-proportioned breasts with realistic volume, soft lower fullness, authentic weight, natural gravity, and believable soft tissue. Her body features narrow ribs, a slim and toned waist, soft rounded hips, full upper thighs, firm feminine limbs, and realistic anatomy. Each piece of clothing fits her body naturally with believable fabric tension, realistic folds, subtle compression, and authentic garment physics. Hairstyle, makeup, and expression are determined by the theme. Character/Expression: Makeup: Apricot pink milk tea amber makeup—gradient apricot pink milk tea eyeshadow, slight rose brown at the eye corners, low-saturation peach blush spreading from under eyes to the center of the face; Expression: Lost in thought by the window, features: eyes not fully focused, lips relaxed, face calm; Hairstyle and accessories: Classical half-up half-down style with jade ornaments, long black hair partially tied up, bun positioned at middle-low height, long hair falling naturally, a few stray hairs framing the face. Clothing/Pose: Clothing: Classical long dress with a cinnabar red to ivory white gradient, complete opaque inner layer with a low-shoulder neckline, wide draping water sleeves, sheer outer gauze, restrained gold thread embroidery, and era-appropriate back ribbons, clear but not overly tight waistband, paired with gilded flower branch hairpins, pearl tassel earrings, and thin gold bracelets; Pose: Leaning against the carved railing, turning back to look at the camera after hearing a soft call, environmental contact points providing slight support, weight mainly on one leg or the seating surface, the other limb naturally relaxed; Material Physics: Silk, gauze, thin silk, crepe, or cotton/linen retain their respective textures, metal ornaments only provide local reflections without altering the physics of the main fabric. Background/Lighting: Second-floor carved corridor of a Jiangnan song tower by the water at night. The environment includes only carved wooden railings, vermillion gauze curtains, gilded palace lanterns, a guqin (ancient zither), celadon wine sets, a sandalwood censer, and a screen. Dancers and musicians downstairs are soft, out-of-focus environmental layers. Outside the corridor, the moonlit sky and light-colored gauze curtains form a wide, soft main light, continuously illuminating the entire face. The face receives only one primary flash direction, with highlights rolling off continuously, avoiding random discrete spots or local bleaching. Composition/Lens: 70mm two-thirds body environmental portrait, lens height near the railing level, the figure's turning line and wide sleeves forming a diagonal, depth of the tower and the performance below softly blurred; 9:16 vertical aspect ratio, the figure located slightly left of center and occupying about 60%-72% of the frame, leaving negative space in the direction of sight or movement. Texture/Style: Portable CCD direct flash: 2000s portable CCD digital camera, 28–35mm equivalent close-range wide-angle, on-camera flash illuminating the subject, background one stop lower, limited dynamic range with minor highlight clipping, slight digital sharpening, cool/warm color cast, realistic shadow noise, and upload compression. Skin shows realistic pores and fine facial hair, natural skin tone transitions, clear hair details, believable fabric textures, and subtle digital noise with natural high dynamic range. A small handwritten signature 'Zhong' is naturally added in the bottom right corner. Negative: Avoid random facial spots, local bleaching, oily shine, point light sources, multi-colored light fragmenting facial features; avoid extra fingers, missing fingers, extra arms, deformed gestures; avoid underage look, baby face, plastic skin, oily highlights, excessive blurring, doll-like skin texture; avoid gibberish text, real brand logos, or watermarks that distract from the subject.
02 / 复刻路径
不要一次改完所有变量
用 GPT Image 2 复刻时,先固定主体和构图,再逐次替换光线、色调或材质中的一个变量,这样更容易判断哪项描述真正起作用。
- 主体:A professional prompt for generating high-fidelity cinematic photography of traditional Chinese Hanfu in a Jiangnan water town setting at night, featuring CCD camera aesthetics and realistic textures.
- 视觉方向:社区实测摄影提示词;按原提示词的光线描述执行
- 镜头与构图:按原提示词的镜头与景别执行;按原提示词的构图与画幅执行
- 交付约束:auto,遵循原提示词中的避免项;不要水印、乱码、低清、畸形手、错误透视或主体漂移。
负向约束
遵循原提示词中的避免项;不要水印、乱码、低清、畸形手、错误透视或主体漂移。
模型建议
GPT Image 2 / Grok Imagine / Midjourney / Flux。先用原始 Prompt 建立基准结果,再根据模型对自然语言、镜头运动或文字排版的能力做局部改写。
来源与编辑原则