# Cut from the title card into the interview and build the lower third: name first, role three frames later

Manual DaVinci Resolve / Fusion construction guide generated from the recipe. It is not a tested .drp.
The accompanying Remotion project (recipe.json) is the executable definition; the numbers below are copied from it.
Neither construction establishes source fidelity: this is an original vector approximation of an inspected editing moment.

## Source and clock
Reference: OpenAI's "Previewing Ultrafast mode: GPT‑5.6 Sol at up to 14X the speed" (EDWbX_-RYws), published 2026-08-14.
Source interval [198, 294) on the 24000/1001 fps master: 96 frames, 4.004000 s. Local frame 0 = source frame 198.
Create a 1920x1080 progressive timeline at 24000/1001 fps and a Fusion composition exactly 96 frames long. Keep integer frames; do not round the frame rate.
Background field: #2a2420. Export silent H.264 yuv420p to match the intended playback container.

## Observation
Local 0-1 (master 198-199) are the last two frames of the opening title card: two lines of very large white bold type, top-left and bottom-left, over a warm interview shot. At local 2 (200) a hard cut lands on a different speaker in a book-lined room with no graphics. At local 5 (203) the speaker's name pops on at bottom-left in small bold white type; at local 8 (206) the role line pops on to its right, separated by a wide gap. Both hold unchanged to the end of the window; neither slides, fades or has a plate.

## Mechanism
Two-stage lower-third build on hard steps after a cut: name lands first, the role follows three frames later, with no plate or motion.

## Phases (local frames)
- 0: Title card, last two frames
- 2: Hard cut to interview
- 5: Name pops on
- 8: Role pops on
- 95: Lower third holds

## Layer inventory (absolute 1920x1080 top-left pixels, back to front)
- title-shot [group] at (0, 0) size 1920x1080 visible [0, 2)
  - t-wall [rect] at (0, 0) size 1920x1080 fill #a89478
  - t-edge-l [rect] at (0, 0) size 90x1080 fill #2a2420
  - t-edge-r [rect] at (1830, 0) size 90x1080 fill #2a2420
  - t-figure [rect] at (560, 300) size 800x900 radius 200 fill #e9e4dc
  - t-head [ellipse] at (800, 120) size 320x380 fill #c9a37c
  - t-hair [rect] at (790, 90) size 340x140 radius 70 fill #1b1512
  - t-top [text] at (90, 40) text "Turbo Mode" 108px weight 700 #ffffff
  - t-bottom [text] at (90, 900) text "on Model Six" 108px weight 700 #ffffff
- interview-shot [group] at (0, 0) size 1920x1080 visible [2, 96)
  - i-wall [rect] at (0, 0) size 1920x1080 fill #3a2e26
  - i-shelf [rect] at (560, 0) size 800x1080 fill #5a3d2b
  - i-shelf-1 [rect] at (580, 120) size 760x22 fill #2e1d14
  - i-shelf-2 [rect] at (580, 380) size 760x22 fill #2e1d14
  - i-shelf-3 [rect] at (580, 640) size 760x22 fill #2e1d14
  - i-books-1 [rect] at (600, 160) size 720x210 fill #8a5a3a
  - i-books-2 [rect] at (600, 420) size 720x210 fill #6f7f7a
  - i-door-l [rect] at (200, 0) size 360x1080 fill #d9d6cf
  - i-curtain [rect] at (1500, 0) size 420x1080 fill #1d1a1c
  - i-window [rect] at (1360, 60) size 160x620 fill #e8e2d8
  - i-body [rect] at (560, 640) size 820x600 radius 220 fill #8d2f38
  - i-head [ellipse] at (820, 300) size 300x360 fill #d6ac86
  - i-beard [ellipse] at (830, 470) size 280x220 fill #3a2a22
  - i-hair [rect] at (815, 270) size 310x120 radius 60 fill #2b1f1a
  - interview-shot-floor [rect] at (0, 880) size 1920x200 fill #0e0b09
- lt-name [text] at (140, 954) text "Rhea Castellanos" 28px weight 700 #ffffff visible [5, 96)
- lt-role [text] at (540, 954) text "Member of the Platform Team, Northwind Labs" 28px weight 500 #ffffff visible [8, 96)

## Construction steps
1. Timeline 1920x1080 at 23.976 fps; Fusion composition 96 frames. Local 0 = master frame 198.
2. Picture track: title-card stand-in for frames 0-1 (warm wall, pale figure, two 108 px white bold Text+ lines at (90,40) and (90,900)), then a straight cut at frame 2 to the interview stand-in (dark wall, bookshelf block, door and curtain panels, seated figure).
3. Lower third: two separate Text+ nodes, 28 px white, baseline row at y=954. Name at x=140 in weight 700; role at x=540 in weight 500. No background plate, no drop shadow beyond a light one for legibility.
4. Timing: name node starts at frame 5, role node at frame 8. Both are hard cuts on (no Write On, no opacity ramp) and hold to the end.
5. Easy mistake: revealing name and role together or sliding them in. The source builds them as two separate hard steps three frames apart, and nothing else on screen changes when they land.

## Content variants
Default copy:
- titleTop: Turbo Mode
- titleBottom: on Model Six
- name: Rhea Castellanos
- role: Member of the Platform Team, Northwind Labs
Adaptation changes:
- titleTop: Instant Replies
- titleBottom: on Model Seven
- name: Omar Delacroix
- role: Staff Engineer, Northwind Labs

## Limits
Easing names map to Fusion spline shapes: linear = linear, in/out/inOut = ease curves on the same frames, hold = step, spring/overshoot = slight overshoot past the target before settling.
Text reveal frames give the first and last unit; distribute characters/words evenly between them. Cursor click rings last 14 frames.
Vector illustration is an approximation of layout and motion, not the product UI or recovered production assets.
