PROMPT RADAR / PROMPT插画三维
个人资料 / 头像 - Multi-Pose Wavy Dark Hair Portrait
A prompt for generating three vertically stacked photos of a woman with long dark-brown wavy hair on a bed, using a reference image for identity.
@MissDelulu9 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
它最有价值的部分,是把一个模糊想法推进到了可以被模型执行的程度。 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 an ultra-realistic fashion portrait using the reference image as the exact identity source. Preserve facial features, proportions, skin tone, asymmetry, and natural skin texture without retouching or beautification.
{argument name="poses" default="Three vertically stacked smartphone-photo frames"} of a young woman with {argument name="hair style" default="long, voluminous dark-brown wavy hair"} resting on a plush white bed:
Top: Chin resting on one hand, looking at camera with a gentle smile.
Middle: Soft profile, looking away and smiling, hair flowing across face and shoulder.
Bottom: Cheek on hand, playful wink with an affectionate smile.
{argument name="outfit" default="Minimal black thin-strap top"}, natural makeup, soft brows, pink lips, flushed cheeks. Bright minimalist bedroom, white bedding/walls, sheer curtain, dark-grey headboard. Soft natural morning side-light, realistic smartphone exposure, wide-angle selfie perspective, crisp facial focus, photorealistic, unedited, natural pores.
Negative: CGI, 3D, cartoon, anime, deformed eyes, extra limbs, blur, heavy makeup, plastic/smooth skin, oversaturation, filters, harsh shadows, low resolution, bad anatomy.
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 three vertically stacked photos of a woman with long dark-brown wavy hair on a bed, using a reference image for identity.
- 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