# Style — r1 (reference reel, 9:16 · 34.2 s · 22 shots)
Primary: yes · Kind: video 9:16

## Measurements
| Property | Measured | Evidence |
|---|---|---|
| Cut rhythm | 3.4 cuts/10 s, avg shot 2.9 s, faster in the hook (0.8 s shots) | shots.json, motion-01 |
| Captions | 2–3 words, appear per word, lower third ~72 % height, bold sans, white + shadow, 1 accent word | frames/hook.jpg t=0.42–1.3 |
| Framing | punch 1.0 ↔ ~1.15 on every jump cut; face kept upper third | motion-01..03 |
| Transitions | hard cuts; one whip-pan at 12.1 s (3 frames) | frames/whip.jpg |
| B-roll | 30 % of runtime, full-frame, 1.5–2 s each, screen recordings | shots 4, 7, 9 |
| Graphics | kinetic keyword over presenter at 2 points; counter at 21 s | motion-03 |
| Grade | warm, lifted blacks, light grain | contact sheet |
| Pace | 175 wpm, pauses ≤ 0.25 s | ref.transcript.json |
| Energy | hits on cut points every ~2 beats; zoom on punchlines | inference from motion pages |

## Technique to copy
- Alternate punch 1.0 / 1.12 on every jump cut, focus on the face.
- Captions: `word` style, max 3 words, lower position, one accent word per chunk.
- One whip at the topic change (~12 s); everything else a hard cut.

## Do not copy
- their footage, people, logos, illustrations, wording, story
- their fonts/colours — my design system replaces them

## Mapping to my toolkit
| Reference device | Equivalent |
|---|---|
| 2–3 word reveal captions | `captions.style: "word"`, `max_words: 3`, `position: "lower"` |
| punch on jump cuts | `clips[].punch` 1.0 / 1.12 alternating, `focus` on the face |
| top-half screen recording | `inserts[].layer: "split"`, `enter: "slide"` |
| kinetic keyword | motion-kit `kinetic` scene |
| warm grade | `grade: {"preset": "warm-daylight", "intensity": 0.5}` |

## Confidence
Cut rhythm and captions measured on every shot; energy inferred from motion pages; sound not heard.