PROMPT RADAR / PROMPT视频叙事
社交媒体帖子 - Cinematic Asian Woman Meadow Portrait
A peaceful, minimalist portrait prompt of an East Asian woman in a wildflower meadow under a blue sky.
@ArianaSafa4 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
Cinematic minimalist portrait of a {argument name="subject" default="beautiful young East Asian woman"} standing alone in an {argument name="setting" default="open wildflower meadow"} beneath an endless blue sky, gazing upward with closed or half-open eyes as if embracing the wind. Long, naturally tousled black hair flowing freely across her face, soft porcelain skin, delicate features, subtle natural makeup, calm introspective expression. Wearing {argument name="clothing" default="layered autumn fashion"}—a dark oversized wool coat over a knit sweater, vintage blouse, or a soft floral dress in muted earthy tones. Golden dry grass and wildflowers surround her, creating a peaceful countryside atmosphere. Captured from a low-angle perspective with abundant negative space and a vast sky filling most of the frame. Soft cinematic daylight, gentle breeze moving the hair and clothing naturally, dreamy color grading, warm earthy tones mixed with cool blue skies, subtle film grain, shallow depth of field, creamy background bokeh, Kodak Portra 400 aesthetic, emotional editorial fashion photography, Scandinavian minimalism, Japanese cinematic mood, natural lighting, 85mm lens, f/1.8, ultra-realistic skin texture, fine hair details, high dynamic range, magazine-quality composition, quiet melancholic atmosphere, award-winning photography, 8K, ultra-detailed, no text, no watermark.
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 peaceful, minimalist portrait prompt of an East Asian woman in a wildflower meadow under a blue sky.
- 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