PROMPT RADAR / IMAGE平面设计
社交媒体帖子 - Elderly Couple Diner Breakfast
A photorealistic warm diner breakfast scene featuring an elderly couple sharing a nostalgic morning meal.
@maestroalvarez 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。 先规定信息层级和文字安全区,再谈视觉风格;无法落版的漂亮画面不是完整设计。
01 / 原始方法
完整视觉 Prompt
Create a warm photorealistic lifestyle photograph of {argument name="elderly couple" default="an elderly white couple"} sitting side by side in a cozy old-fashioned diner booth during breakfast. The scene contains exactly 2 people: an elderly man on the left with short white hair wearing a brown-gray plaid flannel shirt, looking down with a gentle focused expression as his left hand reaches toward the food; and an elderly woman on the right with short curly light-brown-gray hair, thin eyeglasses, a dark green sweater, a gold wristwatch and small necklace, resting her chin on her hand while smiling affectionately at him. In the foreground on the table, show exactly 1 white breakfast plate with exactly 4 visible food components: 1 round golden waffle with 1 pat of melting butter, 1 portion of scrambled eggs, and 4 strips of crispy bacon. Add exactly 1 glass of orange juice in the lower-left foreground, slightly out of focus. Set the diner interior with warm sunlight streaming through venetian blinds on the left, wooden window frames, teal booth seating, amber pendant lights, a softly blurred customer in the background, and a vintage wall sign on the right. Use {argument name="lighting style" default="golden morning window light"}, shallow depth of field, natural skin texture, candid composition, documentary realism, 35mm lens look, warm nostalgic color grading, and no visible artificial posing. Avoid text overlays, logos, extra people in the foreground, or cartoon styling.
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 photorealistic warm diner breakfast scene featuring an elderly couple sharing a nostalgic morning meal.
- 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