PROMPT RADAR / VIDEO视频叙事
One storyboard. One prompt. Endless possibilities. This entire animation
One storyboard. One prompt. Endless possibilities. This entire animation was created with AI. Create with GPT IMAGE 2 + Seedance 2.0 Prompt: Use the uploaded storyboard image as a visual reference only. Create a single 1
@HaniaAi12 564 热度 7.2k 浏览
编辑先看
它为什么值得进入灵感库
这条内容值得留下的,不是表面的风格词,而是它提供了一条可以继续实验的画面路径。 把一段描述拆成起始状态、动作变化、镜头响应和结束画面,连续性比形容词更重要。
01 / 原始方法
动作与镜头 Prompt
One storyboard. One prompt. Endless possibilities. This entire animation was created with AI. Create with GPT IMAGE 2 + Seedance 2.0 Prompt: Use the uploaded storyboard image as a visual reference only. Create a single 10-second cinematic 3D animation by recreating every storyboard panel in sequence. Do NOT animate the storyboard page itself. Rebuild every scene as a real animated environment while keeping the exact same girl, outfit, hairstyle, facial features, kitchen, colors, lighting, props, and overall style from the storyboard. 0.00–1.25s: The girl happily waves at the camera and points toward the cupcake ingredients on the table. 1.25–2.50s: She pours flour into a clear glass bowl. Soft flour particles float naturally. 2.50–3.75s: She gently whisks the cupcake batter while smiling. 3.75–5.00s: Extreme close-up of creamy batter being poured into colorful cupcake liners. 5.00–6.25s: She sprinkles rainbow sprinkles over the cupcake batter. 6.25–7.50s: Top-down shot of the cupcake tray as she carefully places it into the oven. 7.50–8.75s: Inside the oven, the cupcakes rise into fluffy golden cupcakes with warm glowing light. 8.75–10.00s: She removes the cupcakes, proudly presents them to the camera, smiles brightly, and gives a cute thumbs-up. Style: Pixar-quality 3D animation, ultra-cute, cozy bakery kitchen, warm golden lighting, soft pastel colors, cinematic depth of field, smooth camera movements, realistic hand movements, natural facial expressions, seamless transitions, high detail, 4K look. Camera: Follow the storyboard camera angles exactly (wide shot, medium shot, close-up, extreme close-up, top-down, close-up, wide shot). Important: Treat the uploaded storyboard only as a reference. Never animate the storyboard page itself. Recreate every scene from scratch in full 3D animation. Do not show storyboard borders or text.
02 / 复刻路径
不要一次改完所有变量
用 GPT Image 2 复刻时,建议先单独生成稳定首帧,再把动作拆成“开始、变化、收束”三段;每段只保留一个主要运动。
- 主体:One storyboard. One prompt. Endless possibilities. This entire animation was created with AI. Create with GPT IMAGE 2 + Seedance 2.0 Prompt: Use the uploaded storyboard image as a visual reference only. Create a single 1
- 视觉方向:实时 AI 视频案例;查看原帖媒体与作者说明
- 镜头与构图:按原帖镜头流程执行;查看原帖媒体与工作流
- 交付约束:auto,保留原作者方法与归属;复刻时避免水印、低清、畸变、主体漂移和无关文字。
负向约束
保留原作者方法与归属;复刻时避免水印、低清、畸变、主体漂移和无关文字。
模型建议
GPT Image 2 / Seedance / Veo / Kling。先用原始 Prompt 建立基准结果,再根据模型对自然语言、镜头运动或文字排版的能力做局部改写。
来源与编辑原则