PROMPT RADAR / IMAGE人像时尚
个人资料 / 头像 - Sunlit Cafe Coffee Portrait
A photorealistic vertical lifestyle portrait of a smiling young woman holding coffee in a calm plant-filled cafe.
@muku_sns 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。 先锁定人物身份与服装轮廓,再用光线、焦段和动作消除“普通写真感”。
01 / 原始方法
完整视觉 Prompt
Create a highly realistic lifestyle portrait of a {argument name="subject" default="young East Asian woman in her mid-20s"} sitting at a bright indoor cafe, smiling warmly at the camera while holding a small beige ceramic coffee cup near her face with her right hand. She has {argument name="hair color and style" default="long natural black hair, softly tousled and parted near the center"}, clear natural skin, subtle makeup, and a relaxed gentle expression. She wears a white textured V-neck camisole, a loose light beige linen cardigan draped off one shoulder, and high-waisted beige linen drawstring pants. Frame her from about the knees or waist up at a marble cafe table, with a small white saucer visible in the foreground. The setting is a calm sunlit modern cafe with large black-framed windows behind her, warm neutral concrete and wood surfaces, leafy green indoor plants on both sides, and soft outdoor greenery visible through the glass. Use {argument name="lighting" default="natural morning sunlight with dappled leaf shadows"}, shallow depth of field, soft bokeh background, realistic skin texture, 50mm portrait lens feel, vertical 2:3 composition, serene cozy atmosphere, muted cream and green color palette, no text, no watermark, no exaggerated beauty filter, no fantasy effects.
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 lifestyle portrait of a smiling young woman holding coffee in a calm plant-filled cafe.
- 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