PROMPT RADAR / IMAGE插画三维
个人资料 / 头像 - Girl in Blue Floral Kimono
A photorealistic vertical portrait of a serene young East Asian girl in a pale blue floral kimono, ideal for elegant character or fashion imagery.
@akanechan_f3 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。 轮廓语言、材质规则和世界观要互相支持,避免把互不相干的风格词堆在一起。
01 / 原始方法
完整视觉 Prompt
Create a highly detailed photorealistic vertical portrait of a young East Asian girl wearing a pale blue floral kimono, seated at a white table outdoors in soft daylight. She has {argument name="hair color" default="dark brown"} shoulder-length softly wavy hair parted to the side, smooth porcelain skin, delicate youthful features, subtle pink makeup, and a calm introspective expression as she looks downward slightly. Her left cheek rests gently on her left hand, with her elbow on the table; her other arm is folded in front of her inside the kimono sleeve. Add exactly 2 visible gold-and-pearl accessories: an ornate floral hairpin with dangling chains on the right side of her hair, and a matching dangling floral earring on her right ear. The kimono should be {argument name="kimono color" default="dusty pale blue"} with a white inner collar and an elegant pattern of cream and blush cherry blossoms with fine gold branches across the sleeves and bodice. Use a shallow depth of field with a softly blurred green-and-white outdoor background, creamy bokeh, natural window-like sunlight, gentle highlights on the hair and fabric, realistic skin texture, crisp facial detail, and a serene refined atmosphere. Frame as a close medium portrait from chest to head, vertical 2:3 composition, centered subject, no text, no watermark, no extra people.
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 photorealistic vertical portrait of a serene young East Asian girl in a pale blue floral kimono, ideal for elegant character or fashion imagery.
- 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