OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

雀斑红发成年女性的双光影电影肖像

Cinematic Double-Lighting Portrait of a Freckled Redhead Adult Woman

A two-panel cinematic close-up portrait study of the same clearly adult freckled red-haired woman, contrasting warm dewy green-eyed light with a short copper-haired low-key brown-eyed look.

雀斑红发成年女性的双光影电影肖像 — AI image creation
雀斑红发成年女性的双光影电影肖像

提示词 / Prompt

中文提示词

以用户上传的参考图作为两个分区中同一位明确成年女性的面部身份依据。两个画面都保留她可辨认的身份，仅改变下文指定的发型、眼睛颜色呈现与光线。 创作一张精致的双分区电影感近景肖像，分界清楚，两个画面中的脸部大小一致。分区一：女性留较长的鲜红色头发，脸上有自然雀斑，绿色眼睛，嘴唇柔和微启，水润皮肤上可见少量水珠。她自信地直视镜头。采用温暖的黄金时段光线、戏剧但柔和的阴影和微妙体积光，背景为暗色柔焦。分区二：同一位成年女性留短而凌乱的铜红色头发，鼻梁与脸颊上有明显雀斑，深棕色眼睛，表情中性；皮肤只有自然微光，不表现夸张汗湿。采用电影感低调照明，以柔和轮廓光勾勒头发，背景保持暗沉柔焦。 两个画面都使用自然克制的妆容、真实皮肤纹理、眼睛清晰合焦和浅景深。两种光色氛围各自鲜明，同时确保脸部身份一致、裁切方式统一。追求照片写实的电影肖像效果，但不指定镜头硬件、光圈、RAW拍摄或承诺特定输出分辨率。不出现其他人物、文字、标志或水印。

English prompt

Use the uploaded reference image as the facial identity reference for the same clearly adult woman shown in both panels. Preserve her recognizable identity in each depiction while varying only the hair styling, eye-color presentation and lighting described below. Create a refined two-panel close-up portrait study, with a clear clean division and consistent face scale. Panel one: the woman has longer vivid red hair, natural freckles across her face, green eyes, softly parted lips and a few visible water droplets on dewy skin. She makes direct, confident eye contact. Illuminate this panel with warm golden-hour light, dramatic but gentle shadows and subtle volumetric glow against a dark softly blurred background. Panel two: the same adult woman appears with short, messy copper-red hair, prominent freckles across nose and cheeks, deep brown eyes and a neutral expression; her skin has a slight natural sheen, not exaggerated sweat. Use cinematic low-key lighting with a soft rim light along the hair and a dark moody blurred background. Both panels use natural understated makeup, realistic skin texture, sharp attention on the eyes and shallow background separation. Keep the two lighting/color moods distinct while maintaining the same facial identity and close crop. Aim for photorealistic cinematic portraiture without specifying lens hardware, aperture, RAW capture or a guaranteed output resolution. No extra people, text, logos or watermark.

查看 API 模型

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