PROMPT RADAR / IMAGE平面设计

个人资料 / 头像 - Eight Distinct Studio Portraits Grid

Generates a photorealistic 4-by-2 grid of eight similar-styled but visibly distinct studio headshots for face variation comparison.

@earlyfield8612 0 热度 0 浏览
个人资料 / 头像 - Eight Distinct Studio Portraits GridORIGINAL 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。 先规定信息层级和文字安全区,再谈视觉风格;无法落版的漂亮画面不是完整设计。

完整视觉 Prompt

Goal: Create a photorealistic comparison grid of eight distinct young East Asian women’s ID-style headshots, emphasizing clearly different facial features while keeping the same general styling. Canvas: Wide horizontal image in a 16:9 aspect ratio, clean white studio background, divided into exactly 8 equal portrait panels arranged in a 4-column by 2-row grid. Use thin light-gray divider lines between panels, no text, no logos, no watermark. Subject details: Show exactly 8 separate head-and-shoulders portraits, each facing directly forward and looking into the camera with a calm neutral expression. All are young adult East Asian women with fair skin, long dark brown to black wavy hair, soft see-through bangs, natural makeup, a white knit sweater, and a delicate small necklace. Make each face visibly unique: 1) oval face with balanced eyes and soft lips, 2) slightly rounder face with wider eyes and gentler cheeks, 3) narrower face with straighter brows and smaller mouth, 4) brighter wide-eyed face with slightly shorter bangs, 5) longer face with heavier waves and more serious gaze, 6) petite heart-shaped face with softer jaw and subtle pout, 7) round face with fuller cheeks and larger eyes, 8) slim face with longer hair framing the cheeks and a composed expression. Visual style: High-resolution realistic studio photography, soft even front lighting, minimal shadows, natural skin texture, clean beauty portrait retouching, sharp focus on eyes, subtle hair detail, consistent camera distance and crop across all panels. Customizable elements: Use {argument name="number of portraits" default="8"} portraits in a {argument name="grid layout" default="4 columns by 2 rows"} grid, featuring {argument name="subject type" default="young East Asian women"} with {argument name="hairstyle" default="long dark wavy hair with see-through bangs"} wearing {argument name="outfit" default="white knit sweaters with delicate small necklaces"}. Constraints: Do not make the faces identical; each portrait must have a distinct face shape, eye shape, brow style, nose, mouth, and overall facial balance. Keep the clothing, background, lighting, and pose consistent so the facial differences are the main comparison point.
查看原作者内容

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。

  1. 主体:Generates a photorealistic 4-by-2 grid of eight similar-styled but visibly distinct studio headshots for face variation comparison.
  2. 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
  3. 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
  4. 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

保留作者,也保留判断。

本页收录公开案例并提供中文索引、结构化拆解和复刻建议,不冒充原作者作品,也不替代原帖上下文。

作者:Yas⚡️BizDev-Engineer · 收录于 2026年8月10日

访问原始来源

沿着这条线继续看