OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image
绯红丝绒墙前的露背缎面裙质感回眸照
Satin Backless Dress Glance Before a Crimson Velvet Wall
A 4:5 photoreal editorial flash portrait of an adult woman in a deep-red backless satin dress, looking back over her shoulder before a paneled crimson wall. Preserve ref-10a-v1 identity; any style reference is optional.
提示词 / Prompt
中文提示词
制作一张4:5竖幅、超写实的时尚闪光肖像，画面为成年女性，并使用 ref-10a-v1 作为必需的面部身份参考。保持同一张可辨识的脸及其比例、皮肤纹理、痣、雀斑或疤痕，以及眼睛、鼻子、嘴唇和下颌形状；不要混合身份、美化、磨皮或改变年龄。她在深绯红色墙壁前摆姿势，墙面带有垂直面板和小金属铆钉。采用四分之三背身姿势，肩膀倾斜，一只手臂自然垂在身侧，头部转向镜头，回眸微笑，神情温暖而自信。她穿深红色露背缎面礼服或缎面吊带裙，可为挂脖或细肩带，面料光滑并呈现逼真的光泽和褶皱，露背设计得体且符合 PG-13；长而深色的光泽波浪发垂在一侧背部。妆容为光泽感晚妆，眉形清晰、柔和高光、红色或酒红色缎面唇膏和淡淡腮红；仅佩戴小巧优雅的耳环，不出现可见品牌。整体呈现干净的高端时尚造型和电影般的夜生活氛围，细节清晰自然，胶片颗粒极轻，动态范围自然（非 HDR），保留真实未修饰的皮肤纹理。不要裸露或露骨性内容，不要标志、水印或可读文字。若提供可选风格参考图，只将其作为色彩、氛围、光线和构图的宽松参考；未提供时按此文字描述创作。
English prompt
Create a 4:5 vertical ultra-photorealistic editorial flash portrait of an adult woman, using ref-10a-v1 as the required face identity reference. Preserve the same recognizable facial features and proportions, skin texture, moles, freckles or scars, eye, nose, lip and jaw shapes; do not blend identities, beautify, retouch skin, or shift age. Pose her in front of a deep-crimson red wall with vertical panels and small metal rivets. Show a three-quarter back pose with shoulders angled, one arm relaxed at her side, and her head turned back toward the camera over her shoulder with a warm, confident smile. She wears a deep-red backless satin gown or slip dress with a halter neck or thin straps, realistic satin sheen and folds, and a tasteful PG-13 open back; long dark glossy waves drape over one side of her back. Add glowy evening makeup, defined brows, soft highlight, red or wine satin-finish lipstick and subtle blush; small elegant earrings only, with no visible brands. Use a clean high-end fashion look with a cinematic nightlife mood, crisp but natural detail, very subtle film grain, natural dynamic range (not HDR), and realistic unretouched skin texture. No nudity or explicit sexual content, logos, watermark, or readable text. If an optional style reference is supplied, use it only as a loose guide for palette, mood, lighting and composition; without one, follow this written direction.
探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作
相关作品
- 打工人的七日循环：周一绝望到周五狂欢
Seven-sector clockwise chibi infographic of one repeating workweek, with the ref-10a-v1 character in all panels, day-specific fictional scenes, and an extra MONDAY on Sunday’s phone.
- 柔和渐变背景下的宾迪与耳饰肖像
A softly lit 3:4 close-up portrait with a pixie cut, centered black bindi, traditional gold-tone jhumka earrings, cream blouse, blush-pink cushion and lavender-pink gradient background. Identity uses ref-10a-v1.
- 现代职业多臂写实影棚肖像
A reusable 4:5 hyper-real studio portrait template showing a profession-specific multi-arm illusion, relevant tools, and a topic-matched seamless backdrop. Identity is supplied by ref-10a-v1; profession defaults to a creative professional and is editable.
- 电光蓝背景下爆发飞溅的剪影侧颜
A dramatic 9:16 side-profile portrait with a shadowed near-silhouette, intense directional light, and outward black and sapphire-blue paint splashes against a vivid electric-blue background. Use the provided identity reference.
- 夜间赛道旁 iPhone 闪光抓拍
使用 ref-10a-v1 保留人物身份与可见发型，创作新的夜间赛车场闪光抓拍；赛车夹克内搭译作修身内搭上衣，品牌贴章保持通用，不编造赞助商文字。
- 自然清晰度与色彩增强人像修复
以固定完整人像照片为底图，轻微提升真实可见的清晰度、细节和色彩；保持面部身份、衣着、姿势与背景，不人为做旧或制造损坏。