PROMPT RADAR / PROMPT视频叙事
个人资料 / 头像 - Melancholic Luxury Candid Portrait
A prompt for a spontaneous, smartphone-style candid portrait set in a misty morning field. It captures a melancholic mood with high detail on facial features and natural environmental elements.
@Kashberg_0 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。 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
Create a close-up portrait without altering the facial features or appearance. A melancholic luxury candid, as if it were a spontaneous photo taken on a smartphone. Aspect ratio: 9:16.
The girl is standing in the middle of {argument name="setting" default="tall grass surrounded by dense morning fog"}. The frame is cropped from the chest up, with a few blades of grass softly blurred in the foreground, while the background gradually fades into the mist.
She has long, thick hair, flawless makeup, sun-kissed skin, and dusty rose lips. Her body is turned at a 3/4 angle, with a calm, direct gaze into the camera. Her arms are gently wrapped around herself in a relaxed pose.
She is wearing an {argument name="clothing" default="oversized white shirt made of lightweight fabric"}. A strong wind naturally blows through her hair, shirt, and the surrounding grass.
The image captures a spontaneous moment in motion. Her face and eyes remain perfectly sharp and highly detailed, with clearly visible pores, eyelashes, and individual strands of hair. The moving hair and grass have subtle, natural motion blur, creating the feeling of a genuine photograph. High natural micro-contrast, optical sharpness, and no artificial over-sharpening.
Photorealistic, realistic skin texture, subtle film grain, natural digital noise, RAW smartphone aesthetic, deep perspective, no beauty filter, no plastic skin, no CGI.
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 prompt for a spontaneous, smartphone-style candid portrait set in a misty morning field. It captures a melancholic mood with high detail on facial features and natural environmental elements.
- 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