PROMPT RADAR / IMAGE插画三维
个人资料 / 头像 - Hyper-realistic Elevator Mirror Portrait
Detailed generation instructions for a sophisticated full-length mirror selfie inside a sleek contemporary elevator, focusing on texture and realistic lighting.
@mehvishs25 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 a hyper-realistic full-length mirror selfie of a clearly adult woman, maintaining her exact facial identity, facial structure, proportions, skin tone, and natural asymmetry. She is standing inside a sleek contemporary elevator with brushed-metal walls and subtle cool reflections. Her pose is relaxed and sophisticated: gently leaning against the elevator wall, one leg extended naturally while the other is bent with her foot resting lightly against the wall. She holds an iPhone close beside her cheek, while her free hand carries a small bouquet of pink roses together with a white shopping bag.
She wears a fitted black mini dress beneath an oversized black blazer, elegant black strappy heels, a delicate necklace, and small hoop earrings. Keep makeup light and realistic. Use subdued overhead elevator lighting with cool metallic highlights and believable shadows. Capture authentic iPhone 15 Pro Max photography characteristics, including natural skin pores, subtle texture, realistic fabric detail, mild smartphone grain, accurate mirror reflections, and natural perspective. Vertical full-body composition, 9:16, sharp facial detail and clothing.
Avoid identity changes, face beautification, artificial skin smoothing, extra fingers, malformed hands, duplicated people, duplicate reflections, distorted phone geometry, warped limbs, unrealistic reflections, CGI appearance, or plastic-looking skin.
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:Detailed generation instructions for a sophisticated full-length mirror selfie inside a sleek contemporary elevator, focusing on texture and realistic lighting.
- 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