电影感隧道人像
一张照片级写实风格的时尚编辑人像，主体蜷缩在黑暗的工业管道中心，周围环绕着圆形长曝光光轨。
gpt-image-2 · 产品营销 · 人像 / 自拍 · 摄影 · 时尚单品 · 电影 / 电影剧照 · 社交媒体帖子 · 网红 / 模特
中文提示词
创作一张超写实电影感时尚编辑摄影作品，主体为 {argument name="subject" default="一位留着黑色长发、戴圆框眼镜的年轻女性"}，她蜷缩在大型黑暗圆柱形工业隧道或管道内。采用 4:5 竖构图，从管道正前方拍摄，主体居中，圆形隧道像漩涡一样填满整个画面。主体低姿态坐在中心，双膝蜷起，双臂交叉放在膝盖上，双脚并拢，平静地注视着镜头，带着自信的浅笑。她穿着 {argument name="outfit" default="一件修身黑色 T 恤、宽松黑色长裤，以及带有白色鞋带和鞋底的黑色帆布鞋"}。隧道内部为黑色、有光泽、磨损且富有质感的金属或混凝土材质，带有圆形擦痕和粗糙的工业表面细节。在隧道内缘添加戏剧性的同心光轨，混合暖橙色高光和细长的白色光线，仿佛是由沿管道曲率的长曝光反射所产生。光影应具有情绪感和高对比度：面部和手臂上有柔和的正面光，头发上有温暖的轮廓高光，外缘周围有深邃的阴影，主体周围形成强烈的圆形框架。采用广角编辑镜头感，带有轻微的鱼眼畸变，主体对焦清晰，皮肤和织物纹理细节丰富，眼镜反射真实，比例自然，画面中不包含任何文字、标志、水印或额外人物。
原始提示词
Create an ultra-photorealistic cinematic fashion editorial photograph of {argument name="subject" default="a young woman with long dark hair and round glasses"} sitting curled up inside a large dark cylindrical industrial tunnel or pipe. Use a vertical 4:5 composition shot from directly in front of the pipe, centered on the subject, with the circular tunnel filling the entire frame like a vortex. The subject sits low in the center with knees drawn up, arms folded and resting across the knees, feet close together, looking calmly into the camera with a subtle confident smile. She wears {argument name="outfit" default="a fitted black T-shirt, loose black pants, and black canvas sneakers with white laces and soles"}. The tunnel interior is black, glossy, worn, textured metal or concrete, with circular scuffs and gritty industrial surface detail. Add dramatic concentric light trails around the inside rim of the tunnel, mixing warm orange highlights and thin white streaks, as if created by long-exposure reflections following the pipe’s curvature. Lighting should be moody and high contrast: soft frontal light on the face and arms, warm rim highlights in the hair, deep shadows around the outer edges, and a strong circular frame around the subject. Use a wide-angle editorial lens feel with slight fisheye distortion, sharp focus on the subject, detailed skin and fabric texture, realistic glasses reflections, natural proportions, and no text, logos, watermark, or extra people.