Vidrush
Toggle Sidebar
AI 视频生成器
Language
领取积分
AI 视频生成器
加载中...
导航
我的创作
Vidrush 工具
Agent
AI 视频
Faceless Channel
AI 音乐
爆款预设
提示词库
图片生成器
AI 配音
AI 广告生成器
Shorts Studio
模板
价格
视频编辑
视频加音频
视频超分
视频增强
人声背景音分离
视频转 Prompt
AI 智能切片
长视频转短视频
AI 游戏高光剪辑
AI B-roll 生成器
YouTube 缩略图生成器
视频 OCR
视频摘要生成器
剧本还原
AI 漫剧转绘
长视频理解
产品视频生成器
Sora 2 视频生成器
编辑图片
图片去背景
图片放大
图片扩图
背景虚化
AI 智能扩图
黑底背景
白底背景
图片去物体
图片增强
图片去模糊
图片锐化
图片转 Prompt
AI 发型切换器
AI 发色切换器
产品图片生成器
生成器
AI 女生生成器
AI 人脸生成器
AI 表情生成器
AI 角色生成器
AI 涂鸦生成器
AI 暗黑美学生成器
音频工具
音频翻译
AI 旁白配音
文本转语音
女声生成器
AI ASMR 声音生成器
AI 有声书
播客生成器
AI 音乐生成器
AI 音效生成器
口音生成器
英式口音生成器
澳洲口音生成器
字幕工具
视频转文本
MP4 转 SRT
视频转 SRT
字幕翻译器
SRT 翻译器
MP3 转 SRT
字幕转语音
AI 写作
AI 语气改写器
中文
3D Render
模型
Gemini Omni
10s
50 credits
提示词
模板
Screen recording of a 3D software viewport, vertical 9:16, 7 seconds, one continuous capture with no cuts. The only sounds are faint scroll-wheel ticks and mouse clicks synced to the zoom steps; no music. THE VIEWPORT: the whole frame is the viewport of a professional 3D modelling app. A thin, very dark grey toolbar strip runs along the top edge with tiny illegible menu text and icons, and a matching thin strip along the bottom edge with tiny icons and a small timeline bar. A tiny greyish text label sits in the bottom-right corner just above the bottom strip. The viewport background is a smooth dark charcoal-grey gradient, slightly lighter in the center where the light falls and darker toward the corners, with a soft vignette. When the camera is far enough away, a fine dark grid of thin lines is visible on the floor plane, plus a soft circular pool of light on the floor around the model, like a single soft spotlight from above. One small white dot, a pivot marker, floats near the model's chin, and a small white cursor arrow drifts very slightly. Nothing else is in the scene. THE MODEL: a photogrammetry-quality 3D scan of a real young fashion model, rendered with real photographic skin texture, real fabric weave and real hair strands, never smooth, waxy or cartoon-like. She is in her early twenties, very pale skin with light freckles across the nose and cheeks, full lips with a natural rose tint, light eyes, a calm neutral pout, looking straight into the camera. Long, thick, wavy copper-red hair parted in the middle, falling past her shoulders and down her chest. She wears a plain fitted light-blue crew-neck cotton t-shirt with short sleeves (no collar, no buttons), a flared white mini skort of stiff cotton, sheer salmon-pink tights, and pale-yellow pointed-toe slingback kitten heels. The scan is completely static: she never moves, blinks or breathes; her hair never moves. THE POSE: she sits on the floor facing the camera with her knees drawn up and her legs bent and crossed at the shins in front of her, the right leg on top so the right knee in pink tights points up toward the camera and the pale-yellow heel of that foot is aimed at the viewer near the bottom of the frame; the left foot is tucked under. Her torso is upright and leaning slightly forward, both arms wrapped loosely around her shins with the hands resting on the ankles. Compact, closed, elegant editorial pose, weight on the floor, hair falling forward over the shoulders. THE CAMERA: nothing in the scene moves; only the viewport camera navigates, the way an artist inspects a model with the mouse: quick scroll-wheel zoom steps that each take about a third of a second and settle with a tiny ease, and slow, subtle orbit drifts in between while the mouse is dragged. The framing is always centered on her. TIMELINE: 0 to 1.0 s, wide shot: the entire seated figure sits small in the middle of the frame with lots of dark empty viewport around her, the floor grid faint, the light pool soft under her; the view drifts very slightly. 1.0 to 1.4 s, a fast zoom straight into her face. 1.4 to 3.5 s, a tight close-up: her face fills the upper half of the frame, hair around it, and the pink-tights knee and the blue shirt fill the lower half; the camera orbits slowly a few degrees to the side and back so her face is seen from slightly changing angles, freckles, pores and single hair strands clearly visible. 3.5 to 3.9 s, one zoom step back out to a medium shot showing her head, torso, arms and the raised knee, white skort visible at the bottom. 3.9 to 4.6 s, a second zoom step further out to the wide shot: the whole figure with the heels, the floor grid lines now clearly visible, and the bright soft spotlight pool on the floor under her fading out into the dark. 4.6 to 6.2 s, hold the wide shot while the camera orbits slowly around her by about fifteen degrees, the grid perspective shifting under her, the light pool staying under her. 6.2 to 6.6 s, a fast zoom straight into her midsection. 6.6 to 7.0 s, hold a close-up of her crossed legs and arms: pink tights, white skort, the hands resting on the ankles, the hem of the blue shirt and her chin at the very top of the frame. NEVER: she never moves, changes pose, looks away, blinks, or animates in any way; the render never looks like a plastic doll, a cartoon or a video game character; no collar or buttons on the shirt; no bright white background; the top and bottom UI strips never disappear or move; no cuts, no transitions, no fades; no text other than the tiny UI labels. Photoreal scan render, soft studio spotlight from above, neutral grey tones, crisp detail, calm and editorial.
4666 / 50000
参考图片(可选)
参考图片
可选:最多上传 2 张参考图来引导生成(JPEG、PNG、WebP,最大 10MB)
画面比例
16:9
9:16
时长
10s
分辨率
720P
1080P
生成音频
2 倍积分
Basic+
所需积分
50
加载中...
视频预览
中文