原始文档
Nano Banana · en
{
"image_analysis": {
"subject": {
"type": "Georgina Rodríguez",
"expression": "Neutral, slightly sultry, contemplative, looking off-camera to the right",
"pose": "Portrait shot, head and shoulders, slightly angled body"
},
"facial_features": {
"skin": "Fair with prominent natural freckles dusted across the nose and cheeks",
"eyes": "Dark, almond-shaped, gazing sideways, adorned with winged eyeliner and mascara",
"eyebrows": "Dark, well-defined, slightly arched",
"lips": "Full, plump, glossy, slightly parted",
"structure": "Soft jawline, youthful features"
},
"hair": {
"color": "Dark brown or black",
"style": "Shoulder-length, layered, tousled waves, messy texture",
"details": "Curtain bangs framing the face, loose flyaways catching the light"
},
"attire": {
"garment": "White tank top / camisole",
"details": "V-neck, thin spaghetti straps, simple texture"
},
"lighting_and_atmosphere": {
"type": "Flash photography",
"contrast": "High contrast black and white",
"shadows": "Hard shadow behind the subject indicating direct front-facing flash",
"vibe": "Vintage, 90s grunge aesthetic, lo-fi, candid"
},
"technical_enhancements": {
"resolution": "4k HD, 8k resolution",
"quality": "Ultra-detailed, photorealistic, sharp focus",
"texture": "Fine film grain, realistic skin texture"
}
},
"generated_prompt": "A hyper-realistic 4k HD black and white portrait of a Georgina Rodríguez with shoulder-length dark messy hair and curtain bangs. She has prominent freckles on her nose and cheeks, full glossy lips slightly parted, and dark eyes looking to the side with winged eyeliner. She is wearing a simple white V-neck camisole. The lighting is direct flash photography style creating a vintage 90s aesthetic with high contrast and a hard shadow against a plain background. The image should feature ultra-detailed skin texture, fine film grain, and sharp focus, capturing a candid, sultry mood."
}使用方法
- 复制完整原始提示词，并保留其中的参数。
- 按需求替换主体、文字和构图要求。
- 在兼容的外部工具中使用，核对当前使用条件并检查结果。
开始前需了解
- 先确认是否需要参考图。标注兼容模型，不代表该示例由这个模型生成。
- 核对名称、数字与画面文字；不同模型版本的结果可能不同。
