Teal-and-Amber Bedroom Portrait with a Bottle-Drinking Pose — AI Image Prompt Guide

Preserve reference identity and specify the drinking action step by step, using cool cyan and warm amber light in a rustic bedroom.
English translation of the public case notes.
Use a character reference sheet and a detailed scene prompt to create a vertical cinematic bedroom portrait of a seated person drinking from a bottle. The notes combine identity consistency, a cross-legged pose, clothing, rustic room materials and teal-orange light. They describe prompt structure, not an independently verified test result. Lock identity and facial proportions first. Then define clothing and body position, followed by room textures, directional cool and warm light and cinematic depth of field. Describe the action as a connected sequence: lift the left arm, bend the elbow, align the bottle opening with the lips, tilt the head back slightly and look upward. This makes the intended relationships more explicit than simply asking the person to drink. Avoid malformed hands, identity drift, a floating or intersecting bottle, a stiff drinking pose, over-smoothed skin, flat light, text and watermarks. The source cuts off its advice about allocating cyan and amber light across the face, hair, jeans and silhouette. [Source text is truncated.]
Original case notes (Chinese)
🔬 效果验证|青橙卧室里的举瓶肖像 复刻实验:文生图 / 图像编辑。 真正被测出来的是:这是一个偏场景应用的电影感人像提示词:把角色参考、姿势动作、服装、室内布景和青橙光效逐层锁定。值得学习的是对“喝瓶装饮料”这种易出错动作的对齐描述,适合改写成写真、海报或角色一致性测试。 边界在哪里: 类型:图生图 输入:上传的角色参考表,以及包含服装、坐姿、饮用动作、卧室环境和青橙光效的完整提示词 目标:生成一张保持参考人物身份一致的竖幅电影感室内举瓶饮用肖像 关键词:角色一致性 / 举瓶饮用 / 盘腿坐姿 / 青橙调色 / 乡村卧室 / 电影景深 / 真实皮肤 / 负面约束 重点:先锁定参考身份与五官比例,再描述服装和身体姿势,最后补充房间材质、双向冷暖光和电影调色;瓶口与嘴唇对齐、头部后仰、视线向上是动作成立的核心。 约束:避免身份漂移、手部畸形、瓶子角度悬空或穿模、喝水姿势僵硬、过度磨皮、平光、画面出现文字或水印。 💡 减少翻车的三个办法 • 把“左臂抬起—手肘弯曲—瓶口对齐嘴唇—头微后仰”写成连续动作链,降低姿势误解。 • 用冷青主光与暖琥珀反向光分工,明确面部、头发、牛仔裤和轮廓光的受光位置… #GPTImage2 #OpenAI #AI绘画 #电影人像 #青橙调色 #动作链提示