2x2 grid, 16:9, choose 4 drastically different subjects: INPUT: TOPIC = [ANY TOPIC] EMOTIONAL_MODE = [OPTIONAL] VIEWER_CONTEXT = [OPTIONAL] <scale_physics> Make size readable through: - tiny observer or vehicle - foreground-to-background distance - atmospheric haze - shadow falloff - visible horizon compression - surface detail changing from macro to massive - known-size objects near the subject - architecture, trees, lights, roads, waves, clouds, or terrain Never rely only on “make it huge.” </scale_physics> <symbolic_physicalization> If TOPIC is physical, enlarge or recontextualize it. If TOPIC is abstract, convert it into a symbolic monument: - memory → buried archive, sleeping giant, endless room - time → eroded ring, colossal clock strata, fossilized wave - AI → black-box cathedral, data planet, thinking machine horizon - grief → sunken statue, hollow mountain, empty city - knowledge → library cliff, book continent, luminous archive - disease → cellular megastructure, shadow organism, invasive landscape - music → wave temple, frozen vibration, instrument horizon Choose only the metaphor that fits TOPIC. </symbolic_physicalization> <cinematic_atmosphere> Use: - minimal figures - wide negative space - dramatic rim light - dusty or misty depth - muted or inferred palette - silence and scale - foreground texture - distant horizon - monumental material detail The image should feel like discovery, not a poster. </cinematic_atmosphere> 2x2 网格,16:9,选择 4 个截然不同的主题:输入: TOPIC = [任何主题] EMOTIONAL_MODE = [可选] VIEWER_CONTEXT = [可选] <scale_physics> 通过以下方式让大小可见: - 微小的观察者或车辆 - 前景到背景的距离 - 大气雾霾 - 阴影衰减 - 可见的地平线压缩 - 表面细节从微观变为巨观 - 主体附近已知大小的物体 - 建筑、树木、灯光、道路、波浪、云朵或地形 永远不要只依赖“把它做大”。 </scale_physics> <symbolic_physicalization> 如果主题是实体的,将其放大或重新语境化。 如果主题是抽象的,将其转化为象征性的纪念碑: - 记忆 → 埋藏的档案馆、沉睡的巨人、无尽的房间 - 时间 → 侵蚀的圆环、巨大的时钟地层、化石波浪 - AI → 黑盒大教堂、数据星球、思维机器地平线 - 悲伤 → 沉没的雕像、空心的山、空城 - 知识 → 图书馆悬崖、书籍大陆、发光的档案馆 - 疾病 → 细胞巨型结构、阴影生物、入侵景观 - 音乐 → 波浪神庙、冰冷的振动、乐器地平线 仅选择适合主题的隐喻。 </symbolic_physicalization> <cinematic_atmosphere> 使用: - 极简的人物 - 宽阔的留白 - 戏剧性的轮廓光 - 尘土飞扬或雾气缭绕的深度 - 柔和或隐含的色调 - 寂静与尺度 - 前景纹理 - 遥远的地平线 - 宏伟的材质细节 图像应该感觉像是一种发现,而不是一张海报。 </cinematic_atmosphere>

Use the provided face reference as the exact identity. Preserve the facial structure, eyes, eyebrows, nose, lips, jawline, skin tone, hairstyle, facial proportions, and overall likeness with maximum consistency. Do not alter or reinterpret

{ "prompt": "A fit young man taking a mirror selfie in a locker room after a workout. He is shirtless with a muscular, well-defined physique, slightly sweaty skin, and wearing dark athletic shorts. He is sitting on a wooden bench, holding a

好莱坞级高端美妆大片,8K超清,阿莱摄影机微距镜头,极致的蒙太奇快剪手法。 核心创意:通过极高密度的碎片化特写镜头组合,将化妆过程压缩成一场视觉与节奏的狂欢,展现从素颜到极致华丽的惊人蜕变。 时长:15秒 场景:高级纯黑背景无影棚,顶级专业的伦勃朗光影布置。 分镜脚本 00:00-00:03 镜头1:真实底色与视觉苏醒 机位:面部特写,平缓推镜。 画面:纯黑背景下,女孩完全素颜,展现最真实的皮肤纹理和疲惫感。她抬起眼眸,眼神慵懒。突然,她对着镜头打了一个清脆的响指。 剪