PROMPT RADAR / PROMPT视频叙事
个人资料 / 头像 - Y2K Style Photo Booth Collage
A realistic prompt for generating a three-frame vertical photo booth collage of close-up selfies with natural expressions and Y2K aesthetic flash lighting.
@Naiknelofar788 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
先别急着照抄整段 Prompt:真正有用的是主体、空间和镜头之间的约束关系。 把一段描述拆成起始状态、动作变化、镜头响应和结束画面,连续性比形容词更重要。
01 / 原始方法
完整视觉 Prompt
Do not change the facial features or hair color. 100% likeness. Create a realistic image.\n\nA three-frame vertical collage (photo booth / photo booth grid) consisting of three close-up selfies of the same girl, showing different natural, lively emotions.\n\nCamera angle: Close-up face / portrait, vertical format.\n\nFrames & Emotions\n\nTop frame:\nThe girl has her eyes closed and is making a duck-face pout.\n\nMiddle frame:\nPlayfully bold expression — the girl slightly scrunches her nose and casually bites her finger while looking sideways.\n\nBottom frame:\nA carefree, confident expression — looking directly into the camera through a slightly falling strand of hair, with her lips slightly parted in a flirtatious expression. Around her neck is a delicate silver chain with a pendant. Well-groomed manicure.\n\nClothing, Background & Lighting\n\nClothing:\n{argument name="clothing item" default="Black (charcoal) oversized hoodie / zip-up hoodie with a red “RED FLAG" print."}\n\nBackground & Lighting:\nMinimalist white/light-colored wall.\n\nLighting is a harsh direct camera flash, creating sharp, defined shadows of the girl on the wall, in a Y2K and late-night party photography style.\n\nShot on an iPhone 17, with realistic mobile-phone processing, no excessive plastic-like skin smoothing, natural depth of field, subtle digital noise, and delicate film grain.
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 realistic prompt for generating a three-frame vertical photo booth collage of close-up selfies with natural expressions and Y2K aesthetic flash 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