即梦AI肢体与动态光影氛围
#肢体与动态#光影氛围
中文翻译
使用 seedance 2.5 制作
提示词:
主体:来自图1的人物。面部和发型必须与图1完全一致;不要变成另一个人。忽略图1的服装、背景、姿势和底部文字。休闲黑色健身房着装:合身背心 + 紧身裤/短裤,头发扎起。即使图像质量低,也要保持面部特征清晰。16:9 横向。
场景:健身房更衣室,有全身镜、储物柜,长凳上有健身包。没有可读的文字、标牌、储物柜编号或品牌标识。
时间线:
- 0–4秒:走向镜子,将手机靠在储物柜/包上,检查倒影,调整背心肩带。
- 4–8秒:坐下,系鞋带并打双结,站起来轻轻弹跳测试鞋子。
- 8–11秒:从包里拿出耳机,插入并调整。
- 11–14秒:面对镜子,转动肩膀,对着倒影点头,轻声说:“好吧……开始吧。”
- 14–17秒:摆臂/拉伸,短暂看向镜头。
- 17–19秒:拿起手机,转身自信地走向门口;在迈步中途定格。
镜头/风格:旧款 iPhone 日常视频外观,标准 1x 镜头,静态/支撑角度,松散的 1–1.5 秒剪辑。刺眼的荧光更衣室灯光,略带冷/绿色调,微弱的镜子眩光。全程有严重的噪点和压缩伪影。自然的皮肤纹理、毛孔、碎发和油光。无调色、电影感/胶片感、美颜滤镜、CG 纹理、鱼眼、0.5x、桶形畸变、暗角或镜头边缘。无场景跳跃或合并剪辑。
声音:无背景音乐。更衣室回声,远处储物柜的叮当声,系鞋带声,耳机盒咔嗒声,外面微弱的健身房噪音,以及近距离/闷闷的自然人声。
[回复]
原帖
Made with seedance 2.5
prompt:
Subject: Person from Image1. Face and hairstyle must match Image1 completely; do not transform into a different person. Ignore Image1’s outfit, background, pose, and bottom text. Casual black gym outfit: fitted tank + leggings/shorts, hair tied back. Keep facial features sharp even with low image quality. 16:9 horizontal.
Setting: Gym locker room with full-length mirror, lockers, and gym bag on bench. No legible text, signage, locker numbers, or branding.
Timeline:
- 0–4s: Walks to mirror, props phone against locker/bag, checks reflection, adjusts tank straps.
- 4–8s: Sits, ties and double-knots shoelaces, stands and lightly bounces to test shoes.
- 8–11s: Takes earbuds from bag, inserts and adjusts them.
- 11–14s: Faces mirror, rolls shoulders, nods at reflection and quietly says, “okay... let’s get it.”
- 14–17s: Arm swings/stretching, briefly glances at camera.
- 17–19s: Picks up phone, turns toward door and walks off confidently; freeze mid-stride.
Camera/Style: Old iPhone everyday-video look, standard 1x lens, static/propped angle, loose 1–1.5s cuts. Harsh fluorescent locker-room lighting, slightly cool/green tint, faint mirror glare. Heavy noise and compression artifacts throughout. Natural skin texture, pores, flyaways, and oil sheen. No grading, cinematic/film look, beauty filter, CG texture, fisheye, 0.5x, barrel distortion, vignette, or lens edges. No scene skipping or consolidated cuts.
Sound: No BGM. Locker-room echo, distant locker clanks, shoelace tightening, earbud-case click, faint gym noise outside, and close/muffled natural voice.
