Reverse-Engineer Any Short Video, Shot by Shot

---
videomd_version: "1.0"
report_id: "rep_example_06"
title: "Hybrid AI Scroll-Stopper Hook + Real Demo Ad"
duration_sec: 12.7
resolution: "540x960"
aspect_ratio: "9:16"
fps: 30.0
shot_count: 6
bpm: 77.1
cut_on_beat_ratio: 0.4
generator: "ShotBreakdown (Unofficial Cloud Workspace · Built on JuneYaooo/rea-video Apache-2.0)"
---

# Hybrid AI Scroll-Stopper Hook + Real Demo Ad

> **Open-Source Attribution**: Generated via ShotBreakdown built on open-source [JuneYaooo/rea-video](https://github.com/JuneYaooo/rea-video) (Apache-2.0) · Official project site: [https://reavideo.ai](https://reavideo.ai) · Audio SFX model: BandIt v2 (CC BY-SA 4.0).

## Layer 1 · Executive Summary & Hook Architecture

- **Duration**: `12.7s` (`6` shots, avg `2.12s/shot`)
- **Pacing Profile**: `fast-cut` · **Remake Difficulty**: `advanced`
- **3-Second Hook Strategy**: The opening shot uses a fantastical AI-generated elf character to create a visual pattern interrupt, immediately grabbing attention with its surreal aesthetic. This is paired with a 'POV' text overlay, directly engaging the viewer and setting up a relatable scenario.
- **Audio & Rhythm**: `77.1 BPM` · Cut-on-Beat `40%` · Voiceover `human_ugc` · Upbeat, modern electronic pop with a driving beat

### Visual Production Mix

| Visual Category | Shots | Duration (s) | Share |
| :--- | :---: | :---: | :---: |
| `ai_video` | 1 | 2.5s | 19.7% |
| `avatar_lipsync` | 2 | 2.6s | 20.5% |
| `live_action` | 1 | 2.5s | 19.7% |
| `screen_recording` | 1 | 2.6s | 20.5% |
| `code_ui_motion` | 1 | 2.5s | 19.7% |

## Layer 2 · Master Shot Timeline

| # | Timecode | Dur | Beat | Visual Type | Framing & Motion | Confidence |
| :---: | :--- | :---: | :---: | :--- | :--- | :---: |
| 01 | `0.00s - 2.50s` | `2.50s` | Yes | `ai_video` | CU · Eye-level · Slow push-in | `98%` |
| 02 | `2.50s - 3.90s` | `1.40s` | Yes | `avatar_lipsync` | MCU · Eye-level · Static locked | `95%` |
| 03 | `3.90s - 5.10s` | `1.20s` | — | `avatar_lipsync` | MCU · Eye-level · Static locked | `95%` |
| 04 | `5.10s - 7.60s` | `2.50s` | — | `live_action` | CU · Eye-level · Handheld micro-shake | `97%` |
| 05 | `7.60s - 10.20s` | `2.60s` | — | `screen_recording` | WS · Top-down · Static locked | `99%` |
| 06 | `10.20s - 12.70s` | `2.50s` | Yes | `code_ui_motion` | WS · Isometric · Code spring slide | `98%` |

## Layer 3 · Shot-by-Shot Forensics & Remake Prompts

### Shot 01 (`0.00s -> 2.50s`) — AI Generated Video

- **Forensic Verdict**: `ai_video` (`98%` confidence · Offline baseline: `live_action` · VLM: `flash`)
- **Forensic Evidence**: Unnaturally smooth skin texture and perfect symmetry typical of AI generation, Subtle inconsistencies in hair strands and ear details across frames, Lack of natural micro-reflections in eyes and skin pores, The 'POV' text overlay is perfectly static and sharp, suggesting digital overlay on an AI base — This shot is classified as AI video due to its hyper-real, almost plastic-like skin texture, perfect symmetry, and lack of natural imperfections, despite the offline detector's low confidence.
- **Camera & Lighting**: `CU` | `Eye-level` | `Slow push-in` | Lighting: Soft, ethereal glow with subtle rim lighting, fantasy-style | Gear/Engine: Kling 2.0 / Sora
- **On-Screen / Spoken Copy**: "POV: Your competitor uses AI scroll-stopper hooks"
- **Recommended Remake Tool**: `Kling 2.0` (`9:16`, motion: `low`)
- **Reverse-Engineered English AI Prompt**:
  ```text
  Close-up shot of a beautiful female elf with long, flowing auburn hair and pointed ears, wearing intricate green leather armor. She gazes directly at the camera with a slight, enigmatic smile. The camera slowly pushes in on her face. Soft, magical glow emanates from behind, creating a dreamy, ethereal atmosphere. Fantasy art style, highly detailed, cinematic.
  ```

### Shot 02 (`2.50s -> 3.90s`) — AI Talking Avatar / Lip-Sync

- **Forensic Verdict**: `avatar_lipsync` (`95%` confidence · Offline baseline: `ai_video` · VLM: `flash`)
- **Forensic Evidence**: Facial expressions and mouth movements appear slightly unnatural and repetitive for a human speaker., Skin texture is too smooth and lacks fine detail, even under compression., Background is perfectly static and uniform, suggesting a generated environment., Eyes lack the subtle micro-movements and reflections of live action. — This shot is an avatar lipsync, indicated by the slightly artificial facial movements and overly smooth skin, despite the offline detector labeling it as generic AI video.
- **Camera & Lighting**: `MCU` | `Eye-level` | `Static locked` | Lighting: Flat, soft studio lighting from the front | Gear/Engine: AvatarSay Talking Character
- **On-Screen / Spoken Copy**: "Direct-to-camera presenter explaining the workflow"
- **Recommended Remake Tool**: `AvatarSay Talking Character` (`9:16`, motion: `low`)
- **Reverse-Engineered English AI Prompt**:
  ```text
  Medium close-up of a smiling male presenter, mid-30s, with short dark hair and a white collared shirt, speaking directly to the camera. His mouth movements are synchronized to a voiceover, conveying enthusiasm and clarity. The background is a clean, neutral light gray studio wall. The lighting is even and bright, highlighting his face. Professional, engaging, and friendly tone.
  ```

### Shot 03 (`3.90s -> 5.10s`) — AI Talking Avatar / Lip-Sync

- **Forensic Verdict**: `avatar_lipsync` (`95%` confidence · Offline baseline: `ai_video` · VLM: `flash`)
- **Forensic Evidence**: Consistent artificiality in facial expressions and mouth movements., Lack of natural skin texture and pore visibility, even with compression., Static, uniform background without any environmental nuances., Eyes appear somewhat glassy and lack genuine human micro-expressions. — Similar to shot 2, this is an avatar lipsync due to the consistent artificiality in facial movements and skin texture, which the offline detector misclassified as generic AI video.
- **Camera & Lighting**: `MCU` | `Eye-level` | `Static locked` | Lighting: Flat, soft studio lighting from the front | Gear/Engine: AvatarSay Talking Character
- **On-Screen / Spoken Copy**: "Direct-to-camera presenter explaining the workflow"
- **Recommended Remake Tool**: `AvatarSay Talking Character` (`9:16`, motion: `low`)
- **Reverse-Engineered English AI Prompt**:
  ```text
  Medium close-up of a male presenter, mid-30s, with short dark hair and a white collared shirt, speaking directly to the camera with a confident and engaging expression. His mouth movements are perfectly synchronized to a voiceover. The background is a clean, neutral light gray studio wall. The lighting is even and bright, emphasizing his clear communication. Professional, informative, and friendly.
  ```

### Shot 04 (`5.10s -> 7.60s`) — Live Action Footage

- **Forensic Verdict**: `live_action` (`97%` confidence · Offline baseline: `live_action` · VLM: `flash`)
- **Forensic Evidence**: Visible skin pores, subtle wrinkles, and natural reflections in the glasses and eyes., Authentic motion blur and slight micro-shakes consistent with handheld smartphone footage., Natural variations in lighting and shadows on the face and background., The 'Shot on iPhone' text overlay is a clear indicator of live-action capture. — This is clearly live-action footage, confirmed by the natural skin details, realistic motion, and the explicit 'Shot on iPhone' text, aligning with the offline detector's classification.
- **Camera & Lighting**: `CU` | `Eye-level` | `Handheld micro-shake` | Lighting: Natural desk light, soft and diffused from the front-side | Gear/Engine: iPhone 16 Pro 24mm
- **On-Screen / Spoken Copy**: "Shot on smartphone under natural studio desk light"
- **Recommended Remake Tool**: `Live Smartphone Shoot` (`9:16`, motion: `medium`)
- **Reverse-Engineered English AI Prompt**:
  ```text
  Close-up shot of a smiling East Asian man in his 60s, wearing glasses and a doctor's coat over a blue tie. He looks directly at the camera with a warm, friendly expression. The camera has a slight, natural handheld micro-shake. Soft, natural desk light illuminates his face, creating gentle shadows. The background is a subtly blurred office environment. Authentic, professional, and approachable.
  ```

### Shot 05 (`7.60s -> 10.20s`) — Screen / App Capture

- **Forensic Verdict**: `screen_recording` (`99%` confidence · Offline baseline: `code_ui_motion` · VLM: `flash`)
- **Forensic Evidence**: Perfectly sharp and static UI elements with no camera lens distortion., Pixel-perfect text rendering and consistent font smoothing., Mouse cursor movement and clicks are precise and programmatic., Absence of any real-world reflections or environmental light interaction. — This is a genuine screen recording of a software dashboard, evidenced by the pixel-perfect UI, programmatic cursor, and lack of real-world camera artifacts, confirming the offline detector's classification.
- **Camera & Lighting**: `WS` | `Top-down` | `Static locked` | Lighting: Flat UI canvas lighting | Gear/Engine: Screen recording software
- **On-Screen / Spoken Copy**: "Real product dashboard walkthrough"
- **Recommended Remake Tool**: `Screen Recording` (`9:16`, motion: `low`)
- **Reverse-Engineered English AI Prompt**:
  ```text
  A clean, modern software dashboard UI walkthrough. The screen displays various input fields, buttons, and text areas for 'Create', 'Make it talk', 'Compare', and 'Pricing'. A mouse cursor navigates smoothly, highlighting different sections and demonstrating text input for '2. Write the script'. The interface is minimalist with white and light gray elements, and subtle blue accents. Crisp, functional, and user-friendly.
  ```

### Shot 06 (`10.20s -> 12.70s`) — Code / Remotion UI Motion

- **Forensic Verdict**: `code_ui_motion` (`98%` confidence · Offline baseline: `code_ui_motion` · VLM: `flash`)
- **Forensic Evidence**: Vector-based graphics with perfectly smooth edges and consistent color fills., Programmatic animation of cards sliding and text appearing, typical of motion graphics., Absence of any real-world camera noise or lens aberrations., The 'ROAS 3.8X BRIEF' text is rendered with perfect clarity and anti-aliasing. — This shot is programmatic UI motion graphics, characterized by its vector-based elements, smooth programmatic animations, and perfect digital rendering, aligning with the offline detector's classification.
- **Camera & Lighting**: `WS` | `Isometric` | `Code spring slide` | Lighting: Flat UI canvas lighting | Gear/Engine: Remotion 4 / After Effects
- **On-Screen / Spoken Copy**: "Clone This Winning Formula — AI Hook + Real Proof"
- **Recommended Remake Tool**: `Remotion React` (`9:16`, motion: `medium`)
- **Reverse-Engineered English AI Prompt**:
  ```text
  A dynamic motion graphics sequence featuring three identical, minimalist rectangular cards with rounded corners, sliding into view from the left. Each card displays 'ROAS 3.8X BRIEF' at the top, followed by 'Clone This Winning Formula' in a larger, bold font, and then 'AI Hook (0-3s) + Real Proof (3-10s)' below. The cards are a dark gray with a subtle yellow accent line at the bottom. The background is a clean, dark gray. Smooth, professional, and impactful.
  ```
- **Remotion Component Spec (`JSON`)**:
  ```json
  {"componentLayout": "Three identical rectangular cards, horizontally aligned, sliding in from left to right.", "springPhysics": {"fps": 60, "damping": 20, "stiffness": 150}, "paletteHexes": ["#282828", "#3A3A3A", "#FFD700", "#FFFFFF"], "textAnimationKeyframes": [{"element": "ROAS 3.8X BRIEF", "animation": "fade_in", "delay": 0.1}, {"element": "Clone This Winning Formula", "animation": "slide_up_and_fade", "delay": 0.2}, {"element": "AI Hook (0-3s) + Real Proof (3-10s)", "animation": "slide_up_and_fade", "delay": 0.3}]}
  ```

## Layer 4 · Audio Beats, SFX Hits & Transcript

- **SFX / Stem Triggers (BandIt v2)**: `5.40s` (whoosh), `12.12s` (whoosh), `12.38s` (impact)
- **Beat Grid (`77.1 BPM`)**: `0.22s, 1.00s, 1.78s, 2.51s, 3.33s, 4.10s, 4.88s, 5.65s, 6.43s, 7.21s, 7.99s, 8.75s, 9.53s, 10.32s, 11.09s, 11.87s`

### Aligned Captions (`SRT`)

```srt
1
00:00:00,000 --> 00:00:02,500
POV: Your competitor uses AI scroll-stopper hooks

2
00:00:02,500 --> 00:00:03,900
Direct-to-camera presenter explaining the workflow

3
00:00:03,900 --> 00:00:05,100
Direct-to-camera presenter explaining the workflow

4
00:00:05,100 --> 00:00:07,600
Shot on smartphone under natural studio desk light

5
00:00:07,600 --> 00:00:10,200
Real product dashboard walkthrough

6
00:00:10,200 --> 00:00:12,700
Clone This Winning Formula — AI Hook + Real Proof
```

Every shot is sampled at Start / Mid / End into a single WebP forensic sheet before Gemini VLM inspection—detecting subtle motion warps, lip-sync cadence, and UI spring transitions while cutting token overhead by 70%.