PROMPT RADAR / PROMPT人像时尚
个人资料 / 头像 - Casual Backlit Knitwear Portrait
A prompt for generating a detailed, snapshot-style portrait of a woman in knitwear, emphasizing natural skin texture and a serene backlit atmosphere.
@AIVideoHub_ 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
Generate real-life photos, presenting portrait photography in a daily snapshot style. Use a realistic photography visual style to show a beautiful woman with porcelain-white, even, clear, and delicate skin. She is wearing {argument name="clothing" default="an off-the-shoulder sleeveless halter knit top"}, with shoulders exposed and exquisite collarbones, wearing finely detailed hair accessories and earrings, making a playful and cute pose with a cute expression. Hair strands at the temples fall naturally, quirky yet lazy. The background is {argument name="background" default="an ordinary room"}. Use backlighting and a low-angle upward composition. Natural soft light outlines her silhouette, making the hair glow and creating a quiet, relaxed atmosphere. Use a half-body shot and capture super details with a macro lens. Original ratio. Regenerate for me, only one person, not two, 9:16.
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 prompt for generating a detailed, snapshot-style portrait of a woman in knitwear, emphasizing natural skin texture and a serene backlit atmosphere.
- 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