PROMPT RADAR / PROMPT插画三维
社交媒体帖子 - Cinematic Sunset Train Interior Portrait
A highly detailed cinematic prompt for GPT Image 2 depicting a woman in a vintage train car during sunset.
@CyberTotal2026 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
它最有价值的部分,是把一个模糊想法推进到了可以被模型执行的程度。 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
Subject:
{argument name="subject" default="Vintage train car at sunset"}
Main Figure:
A portrait of a woman standing by a vintage train window, holding a vertical metal pole with her right hand and looking out. A thigh-up shot in portrait orientation, wearing a {argument name="outfit" default="natural white short dress"}, carrying a rattan bag, with other passengers and hanging straps visible in the background.
Character and Expression:
Body facing forward, face turned in profile toward the window on the left. A contemplative, quiet expression with gaze fixed far away and lips closed. A gentle oval face, small chin, almond eyes seen from the side, natural eyebrows, a thin bridge and rounded tip of the nose, and pale coral lips. Bright skin under the sunset glow with minimal makeup. Dark brown hair gathered in a low ponytail, with thin bangs and stray strands near the cheeks.
Attire and Pose:
A natural white dress with thin shoulder straps, a deep V-neck with lace trim and overlapping fabric, a wide elastic waist, and a short flared skirt with thin lace at the hem. A delicate necklace and bracelet, and a brown open-weave rattan bag on the right shoulder. Her left hand (right side of the frame) lightly grips a chest-high metal pole, while her right arm hangs by her side.
Background and Lighting:
On the left, a series of windows with vintage metal frames; in the center and right, silver vertical poles and white triangular hanging straps above. Passengers lined up on seats in the back right. Hard primary light from the low sunset streams through the left window, illuminating the face, neck, chest, and white dress in gold, while the interior falls into deep amber shadows. Large dark bokeh overlaps the left and right foreground.
Composition and Camera:
A 3:4 vertical environmental portrait at roughly eye level, capturing from head to mid-thigh. The person occupies much of the center, with the window on the left and poles/passengers extending to the right. Focus on the face, surrounded by foreground blur and a shallow depth of field as if peeking in.
Texture and Style:
Cinematic photorealistic photography. Highly detailed rendering of old metal, window glass, thin cotton fabric, rattan, and skin, unified by golden sunset light and deep brown shadows.
Negative:
Do not face the window-looking profile forward; do not omit background passengers or foreground bokeh.
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 highly detailed cinematic prompt for GPT Image 2 depicting a woman in a vintage train car during sunset.
- 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