提示词分享 🚨
分享链接 · https://openimages.ajiang.me/p/2058799706977746948/

camera_lens · shallow-depth-of-fieldcamera_lens · wide-anglecomposition · close-uplighting · hard-lightmood · tensesubject_type · character
提示词(原文,逐字保留)
Prompt Share 🚨
Too many legends. Not enough elevator capacity.
Prompt 👇
Preserve every uploaded character’s exact likeness, proportions, eye shape, spacing, stylization, silhouette, outfit, accessories, materials, and original color palette.
Do not humanize.
Do not alter anatomy or facial structure.
Do not add extra facial features if the original character does not have them.
CHARACTER COUNT LOCK (CRITICAL)
Use each uploaded character exactly once and only once.
Do not create duplicate instances, clones, alternate poses of the same character, repeated identities, or hidden extra occupants.
Every visible character inside the elevator must correspond to one unique uploaded character only.
Reflections must only reflect already existing characters physically present in the scene and must never resemble additional bodies.
Do not generate extra characters behind the doors, inside shadows, or in reflective surfaces.
The total number of characters visible in the image must exactly equal the number of uploaded references.
Empty space is allowed if necessary to preserve character-count accuracy.
SCENE
The elevator interior feels claustrophobic, overstressed, and physically uncomfortable. Some characters are pressed against scratched reflective steel panels, others squeezed into corners, crouched low, hanging from support rails, flattened awkwardly against walls or ceiling from pressure, or partially trapped between the narrowing elevator doors. Every pose must feel unique and reactive to the cramped environment.
PERSPECTIVE
Extremely close 24mm cinematic lens inside the elevator creating severe perspective distortion and emotional pressure. Camera positioned chest-height among the crowd as if physically trapped inside the elevator.
COMPOSITION
Dense overlapping silhouettes with minimal negative space, aggressive foreground obstruction, cropped limbs and partial body framing at the image edges. Intentionally unstable and uncomfortable composition. No centered symmetry.
ENVIRONMENT
Worn elevator interior with scratched stainless steel panels, flickering fluorescent ceiling lights, emergency panel glow, subtle grime accumulation, fingerprints, worn rubber flooring, slight atmospheric haze.
LIGHTING
Harsh overhead fluorescent lighting mixed with intermittent red emergency light, creating sculptural shadows and uneven exposure pockets across the compressed bodies. Strong highlight reflections on metal surfaces without clean mirror reflections.
DEPTH OF FIELD
Focus locked on the nearest central characters while foreground limbs and deeper background layers progressively collapse into cinematic blur.
TEXTURE
Fabric compression folds, subtle surface wear, scratches, tactile material realism, micro-detail in the focal plane, atmospheric dust particles.
MOOD
Absurd tension, collective exhaustion, accidental comedy, urban survival energy.
EFFECTS
Subtle handheld motion blur, restrained chromatic aberration near frame edges, cinematic film grain, slight lens breathing distortion.
Avoid flat lighting, avoid spacious composition, avoid duplicated characters, avoid clean reflections, avoid organized posing, avoid full-frame sharpness.
ar 4:5
中文译文
提示词分享 🚨
角色太多,电梯容量不够。
提示词 👇
保留每个上传角色的精确外观、比例、眼型、间距、风格化处理、轮廓、服装、配饰、材质以及原始色彩搭配。
不要拟人化。
不要改变解剖结构或面部结构。
如果原始角色没有额外的面部特征,请勿添加。
角色数量锁定(关键)
每个上传角色仅使用一次,且只能使用一次。
不要创建重复实例、克隆、同一角色的替代姿势、重复身份或隐藏的额外人物。
电梯内每个可见角色必须仅对应一个独特的上传角色。
反射中只能映出现场实际存在的角色,不得出现类似额外身体的内容。
不要在门后、阴影中或反射表面上生成额外角色。
图像中可见的角色总数必须严格等于上传参考的数量。
如有必要,允许留白以保持角色数量准确。
场景
电梯内部感觉幽闭、过载且身体不适。一些角色紧贴划伤的反射钢板,另一些被挤入角落、低蹲、悬挂在扶手上、被压力别扭地压平在墙壁或天花板上,或部分卡在收窄的电梯门之间。每一个姿势都必须独特,并对拥挤的环境作出反应。
视角
电梯内部使用极近距离 24mm 电影镜头,产生严重的透视变形和情绪压力。相机置于人群胸口高度,仿佛被物理困在电梯内部。
构图
密集重叠的轮廓,极少的负空间,激进的前景遮挡,画面边缘的肢体裁切和局部身体取景。刻意制造不稳定和不适的构图。无中心对称。
环境
磨损的电梯内饰,划伤的不锈钢面板,闪烁的荧光顶灯,应急面板辉光,微妙污垢堆积,指纹,磨损的橡胶地板,轻微的大气雾霾。
照明
刺眼的顶部荧光照明,混合间歇性红色应急灯,在被压缩的身体上形成雕塑般的阴影和不均匀的曝光区域。金属表面有强烈的高光反射,但没有干净的镜面反射。
景深
焦点锁定在最近的中心角色,前景肢体和更深背景层逐渐坍缩为电影感的模糊。
质感
布料压缩褶皱,微妙的表面磨损,划痕,触觉材质真实感,对焦平面内的微观细节,大气尘埃颗粒。
氛围
荒诞的紧张感、集体的疲惫、意外的喜剧感、都市生存能量。
效果
微妙的手持运动模糊,画面边缘附近克制的色差,电影颗粒感,轻微的镜头呼吸畸变。
避免平面照明,避免宽敞构图,避免重复角色,避免干净反射,避免有组织的姿势,避免全画面锐利。
ar 4:5
来源与署名
原文由 @edizkan_ 发布在 X:查看原推文。 本页逐字保留原文并提供机器翻译的中文解读;版权归原作者所有。
同工具的更多提示词
- 感谢分享提示词 ☺️ 这是我的 😁 一只魅力非凡的耳廓狐DJ在巨大的户外沙漠锐舞派对上转动着发光的黑胶唱片,头顶高耸的霓虹全息影像,夸张的大耳朵在脉动的赛博朋…@JasonHeatony7
- 派对之后 { "prompt": "Ultra-realistic cinematic fashion runway photography of a gla…@XSydneyFan
- 提示分享: 电影级超现实幻想场景,描绘一个破碎的人形正在溶解成深色灰烬状颗粒,同时因情绪崩溃双手抱头,与三个发光的空灵存在形成对比——他们手牵手站在弥漫着暖金…@freesoulz75
- “给我做一个会病毒式传播的东西” 结果它返回了一个主角收据生成器,以收据格式吐槽你的整个人格。 这只是用一句话做出来的。 试试看 👇 https://t.co…@marrowleaf
- 神话中由熔融转化中诞生的水晶女神,发光的金色裂痕如火山脉络般蔓延于她的身躯之上,优雅的女性形象静坐于倒映的黑暗水面,身后巨大的水晶构造宛如神圣之翼,头顶环绕天…@freesoulz75
- 一个 3D 立体 pop-up 立体书插画,主角为 [subject](主题/对象),分层纸张元素展开成一幅微型场景。柔和光线,纸质表面带有纹理,呈现俏皮的手…@azed_ai