PROMPT RADAR / IMAGE插画三维

社交媒体帖子 - Hamster Taiko Festival

A lively anime-style Japanese matsuri scene with cute hamster drummers, manga sound effects, and a dramatic festival crowd.

@hmst_yyyy 0 热度 0 浏览
社交媒体帖子 - Hamster Taiko FestivalORIGINAL 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。 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。

完整视觉 Prompt

Create a high-energy Japanese festival anime illustration of {argument name="animal performers" default="cute hamsters"} performing taiko drums at night. Use a square 1:1 canvas, dynamic manga perspective, warm golden stage lighting, explosive radial speed lines from the top center, falling multicolored confetti, and a packed matsuri crowd in the background. In the foreground, show exactly 3 main hamster drummers: left hamster in a blue happi coat winking while holding one drumstick, center hamster largest in a red happi coat joyfully raising one drumstick and striking with the other, and right hamster in a dark navy happi coat smiling with eyes closed while holding two drumsticks. Place exactly 3 large wooden taiko drums across the bottom foreground, each with rope lacing, black studs, and a black tomoe-style crest on the drum body. Behind them, fill the scene with a dense crowd of smaller hamster drummers wearing headbands and happi coats, carrying sticks and cheering; include festival lanterns on both sides and two vertical festival banners labeled 「祭」. Add exactly 5 oversized red-and-black Japanese manga sound-effect clusters reading 「ド」 or 「ドン」 across the upper half, with thick white outlines and black drop shadows. In the top-right corner, add a white rectangular narration box with a black border containing vertical Japanese text: {argument name="narration text" default="その宴は三日三晩続いた"}. Visual style: polished colorful anime, cute mascot character design, glossy eyes, rosy cheeks, expressive open mouths, detailed wooden drums, dramatic comic-book impact lines, saturated orange-yellow festival atmosphere, crisp line art, no photorealism, no watermark.
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:A lively anime-style Japanese matsuri scene with cute hamster drummers, manga sound effects, and a dramatic festival crowd.
  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:🐹🐹 · 收录于 2026年8月11日

访问原始来源

沿着这条线继续看