OpenZhenZhen AI 提示词画廊 · MiniMax H3 · video
男友视角·雨靴溅湿鞋尖
Boyfriend's View: Rain Boots Splashing Wet Toes
男友视角·雨靴溅湿鞋尖，使用同名原始提示词配对并提供中英文版本
提示词 / Prompt
中文提示词
[参考生成] 15秒、16:9、真人写实、男友第一人称眼睛视角，一镜到底。<Subject 1>来自<Picture 1>，始终保留同一成年女性的脸、肤色、深色眼睛及深棕色长发；参考图片只用于身份与明确可见的衣着，不是本片开场帧。自然表演与真实物理，普通话对白按下述原句发声，不出现字幕或文字。 雨后庭院。女方保持参考脸与长发，戴帽放下的黄色雨衣、长裤和雨靴；男友站在一级浅而干的门槛上，双脚都在上层，右侧浅色帆布鞋尖刚露过台阶边。台阶下紧贴一处浅水洼，女方站在对面一大步处，旁边有干路。镜头从男友眼睛轻微俯视，优先看清她脸，不是从头俯拍双脚。她抬头问「走哪边？」男友右手短暂指向干路，答「那边。」她先半步假装走干路，发现仍干燥且暴露的右鞋尖后故意变向，只用一只雨靴踩进靠近台阶的一侧水洼一次。水从接触点飞起，跨短距离击中原来的鞋尖。男友短暂低头，看同一鞋尖出现清楚深色湿斑，再把右脚完全收回干台阶；随即抬眼。她仍有雨靴在浅水中，看过鞋子原处后对上视线说「晚了。」结尾看她的脸，不勉强同框远处脸与收回的鞋。雨后自然声、三句对白、一次清晰溅水及鞋撤回声；无第二次跺水、字幕或音乐。 仅指定人物和物件出现，动作因果与位置连续；男友的脸和身体不入画，只在需要时出现指定的手或鞋。环境声及对白与实际动作同步，无旁白、笑声罐头或额外音乐。
English prompt
subject_definitions: <Subject 1> is the adult woman in <Picture 1>. Preserve her visible facial identity, fair skin, dark eyes and long center-parted dark-brown hair. Her hood-down yellow rain jacket, trousers, rain boots and courtyard are original to this episode. <Picture 1> supplies identity, not the opening frame. summary: [reference generation] A 15-second, 16:9 live-action boyfriend eye-POV story. He stands one step above <Subject 1> and points her toward a dry path around a shallow puddle. She pretends to take it but chooses to step into the water toward his exposed light sneaker toe. One splash actually wets that same toe before he withdraws it. She sees the wet mark and tells him he is too late. retention_analysis: <Subject 1> (appears in [Shot 1]): fully_preserved - her referenced facial identity and long hair remain visible; rainwear, courtyard and water are original episode elements. detailed_description: mpov, first-person perspective pov video, high angle. One continuous 15-second 16:9 live-action view from my eyes while I stand on a single shallow dry doorstep, slightly higher than <Subject 1> below. My natural attention begins on her face and body at a gentle downward angle, not an overhead view of my feet. Both my feet are on the upper step; the clean toe of my right light-canvas sneaker projects just past the step edge near the bottom of my sight. Directly against the base of the step is the near rim of one shallow puddle. <Subject 1> stands about one long pace away at its farther rim, with a dry path alongside it. Her hood is down, so her complete face is visible above her yellow jacket and rain boots while she speaks and makes her choice. After the splash I have a brief reason to look farther down at the shoe, then return promptly to her. [Shot 1] She looks up and asks, <Subject 1> (S1), <d>[Chinese] 走哪边？</d> My right hand briefly points toward the dry strip, and I answer from my own eye position, me (S2), <d>[Chinese] 那边。</d> My pointing hand lowers, but my right shoe toe remains exposed and still dry. She notices it. First she shifts a half step toward the indicated dry path, as if accepting my choice. Then she deliberately changes direction and plants one rain boot into the puddle close to the step-side rim. Water visibly bursts outward from that boot contact, crosses the short gap and strikes the original exposed light sneaker toe. At the sound and sight of that one splash, my eyes dip naturally for a brief look at the shoe. Before I move it, a broad dark wet patch is plainly visible on the same canvas toe, not a sudden change to another shoe. I then pull the wet right foot completely back onto the dry upper tread, out of the forward view; the brief downward glance can follow this retreat to its actual resting place. My eyes promptly rise to her again. She is lifting her gaze from the place where the wet shoe just was, showing that she noticed it. With her rain boot still in the shallow water, she meets my returning gaze and says, <Subject 1> (S1), <d>[Chinese] 晚了。</d> End on her face receiving my too-late retreat, not on an impossible shoe-and-far-face combination. There is no second stomp or unexplained water at the camera. Only my pointing hand, a little forearm and the right shoe tip enter when needed; my face and the rest of my body are never seen. overall_soundscape: Quiet rain-after courtyard ambience, her Mandarin question and final tease, my close off-screen reply, one clear rain-boot splash against the shallow puddle and the quick scuff of my retreating shoe. No repeated splashes, narration, captions or added music. non_diegetic_music: N/A
主题与风格
探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作
相关作品
- 风暴崖上的巨龙吐息
制作一段20秒、16:9的电影感奇幻视频。一条巨龙立于嶙峋的岩石峭壁上方，背景是昏暗的暴风雨天空，黑曜石般的鳞片映着微弱光线。镜头从低角度缓慢推近龙首；它鼻翼翕动，双眼燃起凶猛光芒，深深吸气。随后巨龙从口中喷出强劲的火焰洪流。镜头以流畅的环绕弧线跟随宽阔的火焰，让橙红色光芒照亮岩石与远处群山，投下锐利阴影，并使灼热空气泛起扭曲波动。短暂停留，展现巨龙的威势，再略微拉远，直至火焰渐渐熄弱。鳞片与火焰呈现丰富细节与写实质感，暴风雨戏剧性光影，镜头运动流畅，电影感高反差构图。不要文字、字幕或标志。
- 雨落叶面
创作一段20秒、16:9的写实微距短片，呈现雨滴落在一片叶子上。以超慢动作展示雨滴撞击叶面、弹起，并形成细致涟漪向外扩散。采用柔和漫射光、低饱和电影色调、清晰自然的细节，以及宁静空灵的氛围。构图聚焦叶片与雨滴，使用浅景深和稳定镜头。
- 雨中极速赛道
生成一段 20 秒、16:9 的电影感视频。以高速摄影风格呈现一辆采用法拉利风格设计的红色一级方程式赛车，在雨天绕赛道疾驰。采用高角度视点和对称构图。展现轮胎扬起的水雾与飞溅水花，并以明显的动态模糊营造强烈速度感。画面细节丰富，采用电影感光影，并呈现真实的湿滑赛道倒影。确保赛车和赛道在整个画面中清晰可辨。
- 雨夜巷中寻线索
生成一段20秒、16:9的电影感写实视频。夜晚，纽约一条典型的小巷里，雨水打湿的地面在几盏昏暗路灯下映出鲜艳色彩。巷道两侧是布满各色涂鸦的老砖墙，旁边放着垃圾桶。透过部分遮挡视线的墙壁，可以看到几栋高楼的一角，以及远处有车辆和行人的繁忙街道，呈现城市环境。一名侦探注意到地上的一张纸，弯腰将它拾起，并以警觉、专注的神情仔细查看。镜头平稳缓慢地推近侦探，最后近距离呈现侦探手中的纸张。全程保持真实的电影光线、鲜艳色彩和连贯自然的动作。
- 大阪雨巷霓虹末世
制作一段20秒、16:9画幅的电影感数字艺术视频。场景是黄昏时分大阪一条被雨水浸透的反乌托邦赛博朋克后街。开头用宽幅建立镜头展现狭窄的湿街、密集霓虹招牌、头顶电缆、自动售货机和有雾的纵深。镜头以街面高度缓慢向前推进，几名穿深色衣物、身份模糊的人物背向镜头行走并经过画面；保持其身份、面孔和性别不明确。洋红、青色与琥珀色霓虹在水洼和湿沥青上形成倒影，雨滴落下时激起一圈圈扩散的水纹。通过浓重阴影、远处剪影、闪烁招牌、飘散蒸汽和偶发电气故障营造克制的恐怖张力，同时保持阴郁的学院派氛围与精致的超写实质感。使用真实雨景、体积雾、细致湿润表面、轻微手持运动和高反差电影灯光。不要对白、字幕、标志或可辨认的受版
- 大阪黄昏霓虹雨巷
20秒电影感数字艺术视频，画幅比例16:9。黄昏时分，大阪一条被雨水浸透的后街在霓虹招牌下闪烁，积水倒映着斑斓灯光。昏暗而无法辨认身份的人影穿过雾气与雨幕，沿狭窄巷道自然行走。采用层次丰富的数字绘画质感、超写实的湿润表面、具有氛围感的景深、戏剧性的霓虹色彩对比，以及细致的实时渲染光照。开场以缓慢的空旷雨巷全景建立环境，随后在街道高度跟随经过的人影，最后停留在积水中的霓虹倒影上。无对白、字幕、标志或可读文字。