← 返回提示词图库

PROMPT RECORD图像记录

将 image1 完全用作主体身份锚点 生成图像:所附图像中同一主体的超写实屋顶奢华肖像,高精度保留精确面部特征与自然身体比例、面部不对称性。 表情/姿势: 静谧奢华屋顶氛围,主体随意坐在现代...

Nano Banana DiamondZPetSpa Fri May 15 17:52:31 +0000 2026
查看来源

中文说明

将 image1 完全用作主体身份锚点 生成图像:所附图像中同一主体的超写实屋顶奢华肖像,高精度保留精确面部特征与自然身体比例、面部不对称性。 表情/姿势: 静谧奢华屋顶氛围,主体随意坐在现代屋顶咖啡馆椅子上,轻触脸侧一缕发丝,放松半眯眼看向镜头。姿势给人自然停在对话中途的感觉,不要像时装模特一样僵硬。 相机设置: 柔和午后光线下用旧CCD数码相机拍摄,可见数码颗粒,对比度柔和,金饰、天际线倒影和毛皮纹理上有轻微高光过曝,真实便携相机美学。 妆容: 柔棕色华丽妆容,光泽感裸色嘴唇,微妙的细眼线,玫瑰暖调水光皮肤,可见毛孔。 发型: 分层深色长发,柔软纤细纹理与自然蓬松度,在屋顶风中轻轻飘动。 服装: 巧克力棕色垂褶镂空长款连衣裙,腰部布料收褶, oversized棕色仿毛皮大衣滑落至肩下,分层金饰,黑色绗缝奢侈...

原始 Prompt

use image1 exactly as the main identity anchor

Create image: A hyper-photorealistic rooftop luxury portrait of the same subject in the attached image, preserve exact facial features and natural body proportions, facial asymmetry with high precision.

Expression / Pose:
Quiet luxury rooftop mood, subject seated casually on modern rooftop café chair while lightly touching a strand of hair beside her face, relaxed half-lidded gaze toward camera. Pose should feel naturally paused mid-conversation, not fashion-model stiff.

Camera:
Shot on an old CCD digital camera during soft afternoon light, visible digital grain, soft contrast, slight highlight clipping on gold jewelry, skyline reflections, and fur texture, realistic compact-camera aesthetic.

Makeup:
Soft brown glam makeup, glossy nude lips, subtle winged eyeliner, rosy warm dewy skin with visible pores.

Hair:
Long layered dark-brown hair with soft wispy texture and natural volume moving lightly in rooftop wind.

Outfit:
Chocolate-brown draped cutout maxi dress with gathered fabric around waist, oversized brown faux-fur coat falling off shoulders, layered gold jewelry, black quilted luxury handbag with gold chain accents.

Background:
Luxury rooftop café overlooking historic European architecture mixed with modern skyline towers beside the river.

Composition:
4:5 framing with subject placed slightly off-center, skyline and tower positioned upper-right depth, table objects creating foreground realism, minimal 20% upper negative space.

Negative prompt:
AI-stiff seated posture, distorted architecture, mannequin hands, plastic skin, fake fur texture, blurry face.