Three-panel character reference sheet: GPT Image 2 prompt

Based on the attached reference image, generate a three-panel character reference sheet. The composition is a horizontal image divided equally into three vertical panels with clean, thin divider lines. Strictly keep the character's face, hairstyle, hair color, body proportions, clothing, styling, and accessories exactly consistent with the reference image; all three panels show the same person in the same outfit. Left panel - front full-body view, headless. The body faces the camera from below the shoulders down to the feet. Both arms hang naturally at the sides, hands relaxed and open, weight evenly placed on both feet. No head, no hair, nothing above the shoulder line, and no hair falling over the chest or shoulders. The neck ends at the base of the throat with a clean, flat, sharply defined horizontal cut, like a headless mannequin with a crisp sculptural edge. It is not blurred, faded, smoky, transparent, ghosted, bloody, or anatomical; above the cut is only blank background. Preserve normal full-body spacing with enough empty space above the shoulders so the figure's scale and placement match a normal full-body portrait. Middle panel - back full-body view with head. The same person is photographed from directly behind, standing straight. Hair falls naturally down the back. The back structure of the clothing, hemline or trouser shape, and shoes are clearly visible. Both arms hang naturally at the sides, hands relaxed, weight evenly placed on both feet, with the full figure visible from the top of the head to the shoes. Right panel - tight chest-up close-up for identity locking. The same person is framed tightly from slightly above the head to the collarbone, showing only the upper edge of the clothing. The face occupies most of the frame in a true close-up. The body faces the camera, head level, eyes looking directly into the lens, lips closed and relaxed, with a neutral restrained expression. The eyes, eyebrows, lashes, lip texture, and key identity markers are clear at close range. Unified background across all three panels: flat, even 18% middle-gray seamless background with one solid color value, no gradients or falloff. Shadowless frontal illumination with even fill from all sides, no key-light side, no shadow side, no rim light or edge light. The background has no projected shadow and there is zero cast shadow under the feet. Lighting is completely consistent across all three panels. Photorealistic quality, with fine even pores and subsurface scattering visible in the skin, individually defined hair strands, clear fabric weave, and subtle film grain. Real human, real camera photography, never plastic-looking, never rendered, never CGI, no AI skin smoothing, and no excessive smoothing.

基于所附参考图，生成一张三格人物设定参考图。构图为一个横向画面，等分为三个并排的竖直分格 (three-panel character reference sheet)，格与格之间是干净纤细的分隔线。严格保持参考图中人物的面部、发型发色、身材比例、服装造型、配饰完全一致，三格中为同一人物、同一套造型。 左格 —— 正面全身，无头。人物身体正对镜头，从肩部以下直到脚部，双臂自然垂放身侧，双手放松张开，重心均匀落在双脚。没有头部、没有头发——没有任何东西高出肩线，胸前与肩上没有头发垂落。颈部在喉咙根部以一条干净、平整、边缘清晰的水平切口收束，如同无头人台模特，利落的雕塑感切边，不模糊、不虚化、不消散、无烟雾、无残影、无透明、无血迹、无解剖细节；切口之上只有空白背景。此格保留完整头顶空间，肩线上方留有充足空白，使人物在画面中的比例与位置与正常全身像一致。 中格 —— 背面全身，带头。同一人物从正后方拍摄，站姿笔直，头发从背后自然垂落，服装背面构造、裙摆/裤型、鞋履清晰可见，双臂自然垂放身侧，双手放松，重心均匀落在双脚，从头顶到鞋子完整入镜。 右格 —— 齐胸大特写，身份锁定。同一人物从头顶略上方到锁骨、仅露出服装最上缘的紧凑构图，脸部占据画面大部分，真正的近景特写。身体正对镜头，头部平视，双眼直视镜头，双唇闭合放松，中性克制的神情。眉眼、睫毛、唇部质感与关键身份标记在近距离下清晰可辨。 三格统一背景：平坦均匀的 18% 中灰无缝背景，单一色值，无渐变无衰减 (flat 18% gray seamless)。无方向无阴影的正面照明，四周补光均匀一致，无主光侧、无阴影侧、无轮廓光、无边缘光 (shadowless frontal illumination)，背景无任何投影，脚下无接触阴影 (zero cast shadow)，三格光照完全一致。 真实照片级质感 (photorealistic)，皮肤可见细腻均匀毛孔与次表面散射，发丝根根分明，织物纹理清晰 (fabric weave)，轻微胶片颗粒。真实人类，真实相机拍摄，绝非塑料感、绝非渲染、绝非 CGI、无 AI 磨皮、无过度平滑。