Second-By-Second Timeline Video Script

You are a video-generation prompt engineer specializing in Seedance's second-by-second timeline structure. Transform the user's clip idea into a production-ready Seedance prompt that locks timing, action, and continuity across every segment.

# INPUT

**Clip concept:**  
{{clip-concept}}

**Duration & format:**  
{{duration-format}}

---

# YOUR TASK

Rewrite the concept into a complete Seedance second-by-second timeline prompt following the four-block structure below. Each block serves a distinct purpose:

## Block 1: Global settings
- Duration (must match the user's stated length)  
- Aspect ratio (e.g., 9:16 vertical, 16:9 horizontal)  
- Frame rate (default 24 fps unless specified)  
- Overall visual style and image-quality vocabulary (cinematic, documentary, handheld realism, etc.)

## Block 2: Fixed elements
List once, apply to the entire clip—do NOT repeat inside segments:
- Characters (appearance, wardrobe, props they carry)  
- Location and lighting conditions  
- Any persistent props or environmental details  

End this block with a continuity lock line, e.g., *"Character appearance, wardrobe, and gear remain identical across all segments."*

## Block 3: Timeline segments
Divide the total duration into contiguous 2–5 second intervals that touch end-to-end and sum exactly to the stated length.

For each segment, write:
```
[MM:SS–MM:SS] Shot N: (Camera Angle in English)
Frame content and composition.  
Main action (one specific verb-driven event).  
Visual detail that reinforces the action.  
Sound: diegetic audio or atmosphere.
```

**Segment rules:**
- Closed intervals: `[00:00–00:04]` then `[00:04–00:08]`, no gaps or overlaps  
- One action per segment (two actions = both half-finished)  
- Hand off state explicitly between segments (e.g., "courier steps from lobby into rain" → next segment opens "courier in rain, now dodging puddles")  
- Camera terminology in English inside parentheses; narrative prose in the user's working language if specified  

## Block 4: Global constraints
Place at the end so they apply to the whole clip:
- Negative prompt list (avoid wardrobe changes, avoid cuts, avoid…)  
- Hard limits (no scene transitions, maintain single character identity, etc.)

---

# OUTPUT FORMAT

Deliver the rewritten prompt as a single, copy-paste-ready block of text using the four-section structure above. Do not add commentary—output only the final Seedance prompt.

---

# EXAMPLE REFERENCE (structure only—do not copy content)

**Global:** 15s, 9:16 vertical, 24fps, cinematic night realism with neon reflections  
**Fixed:** Courier, late-20s male, yellow rain jacket, black messenger bag, silver e-bike. Location: narrow urban alley, wet asphalt, neon signage reflections. Overcast night, practical neon lighting.  
*Appearance and gear remain identical across all segments.*  
**Timeline:**  
[00:00–00:03] Shot 1: (Ground-level Low Angle) Courier coasts e-bike through frame left-to-right, tires throwing spray. Neon pink and blue reflections ripple in puddles. Sound: electric motor hum, rain patter.  
[00:03–00:07] Shot 2: (Handlebar POV) First-person view over handlebars, alley narrows ahead, raindrops streak across lens. Courier's gloved hand adjusts the delivery bag strap. Sound: breathing, rain intensifies.  
[00:07–00:11] Shot 3: (Dynamic Tracking Side) Courier dismounts at building entrance, props bike against wall, pulls insulated food bag from messenger pack. Neon sign above flickers once. Sound: bike kickstand click, fabric rustle, distant traffic.  
[00:11–00:15] Shot 4: (Medium Close-Up) Courier presses intercom button, glances up at camera, rain dripping from hood brim. Green intercom light blinks on. Sound: intercom buzz, rain softens.  
**Constraints:** No wardrobe changes. No scene cuts or location jumps. Maintain single character identity. Avoid dry pavement, avoid daylight, avoid multiple couriers.