PROMPT RADAR / PROMPT插画三维
个人资料 / 头像 - 3D Character Name Sculpture
A specialized prompt that creates a large 3D text object based on a character's design, incorporating the character into the scene.
@varts_works 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
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。 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
Please use the attached image as the primary reference.
The uploaded character themselves is the main protagonist.
Maintain their face, hairstyle, hair color, eyes, costume, body type, accessories, art style, and age feel.
Create a giant three-dimensional object of the {argument name="name" default="character name"}.
━━━━━━━━━━━━━━━━━━
[Letter Design]
━━━━━━━━━━━━━━━━━━
- Turn the entire name into one giant 3D object.
- Design the font according to the character's atmosphere.
- Cute, stylish, high-end 3D sculpture.
- Thick, presence-filled 3D letters.
- Automatically select materials best suited for the character, such as gloss, glass, matte, metal, candy, crystal, stuffed animal, cloud, flower, etc.
- Color is also automatically determined according to the character.
- The name must be easy to read.
━━━━━━━━━━━━━━━━━━
[Character]
━━━━━━━━━━━━━━━━━━
Characters should naturally pose on the text object.
Randomly select cute poses such as:
- Lying down
- Sitting
- Dangling legs
- Sprawling
- Jumping
- Standing
- Leaning against the letters
- Making a heart shape
- Waving
- Looking this way with a smile
Naturally match the size of the text with the size of the character.
━━━━━━━━━━━━━━━━━━
[Background]
━━━━━━━━━━━━━━━━━━
Automatically generate the background according to the character's world view.
Examples:
- Pastel
- Fantasy
- Space
- Flower field
- Neon
- Starry sky
- Gothic
- Land of sweets
- Ice world
- Japanese style
- Forest
- Underwater
- Above the clouds
Unify the background color, light, effects, decorations, accessories, and atmosphere according to the character.
━━━━━━━━━━━━━━━━━━
[Angle of View]
━━━━━━━━━━━━━━━━━━
- Landscape (4:3)
- Full body included
- The entire name is not cut off
- Both the character and the letters are in the center
- High quality enough to be used as an SNS icon or profile image
━━━━━━━━━━━━━━━━━━
[Image Quality]
━━━━━━━━━━━━━━━━━━
Ultra-high quality, Ultra-high definition, Cute, SNS-friendly, Beautiful lighting, Movie-level lighting, Sharp, Vivid, Detailed, Professional spec, No text (except for the name object)
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 specialized prompt that creates a large 3D text object based on a character's design, incorporating the character into the scene.
- 视觉方向:社区实测视觉提示词;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