# Type a numbered three-word staircase in overlapping beats and wipe it off line by line

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 "How to Schedule a Weekly Metrics Report With ChatGPT Work" (p_slDAvPjv0), published 2026-08-06.
Source interval [941, 1007) on the 24000/1001 fps master: 66 frames, 2.752750 s. Local frame 0 = source frame 941.
Create a 1920x1080 progressive timeline at 24000/1001 fps and a Fusion composition exactly 66 frames long. Keep integer frames; do not round the frame rate.
Background field: #6f6752. Export silent H.264 yuv420p to match the intended playback container.

## Observation
On a warm blurred interview background '2.' pops on at local ~1. The verb types on top-left from ~4 to ~15; the second word pops on whole to the right at ~9 while the verb is still typing; the third word types on lower-right from ~12 to ~21. The three words hold until ~55, then exit as staggered left-to-right wipes: number off at ~57, verb wiped ~57-62, third word wiped ~60-65, second word wiped ~62-66.

## Mechanism
Overlapping type-on staircase: a typed line, a popped line and a typed line whose reveals overlap in time, exited by per-line left-to-right wipes.

## Phases (local frames)
- 1: Step number pops on
- 4: Verb types on
- 9: Second word pops on
- 12: Third word types on
- 21: Complete, hold
- 57: Staggered wipe-off

## Layer inventory (absolute 1920x1080 top-left pixels, back to front)
- pillar [rect] at (1380, 0) size 260x1080 fill #8d846c
- figure [ellipse] at (620, 140) size 620x1200 fill #1e2230
- num [text] at (395, 210) text "2." 64px weight 600 #ffffff visible [1, 57)
- w1 [text] at (90, 340) text "Connect" 84px weight 600 #ffffff reveal chars frames 4-15 visible [4, 63)
- w1-wipe [rect] at (80, 320) size 0x130 fill #6f6752 visible [57, 66)
    keyframes w: f57=0, f62=620 (out)
- w2 [text] at (1240, 530) text "Your" 72px weight 600 #ffffff visible [9, 66)
- w2-wipe [rect] at (1230, 510) size 0x120 fill #6f6752 visible [62, 66)
    keyframes w: f62=0, f66=400 (out)
- w3 [text] at (1540, 720) text "Sources" 72px weight 600 #ffffff reveal chars frames 12-21 visible [12, 66)
- w3-wipe [rect] at (1530, 700) size 0x120 fill #6f6752 visible [60, 66)
    keyframes w: f60=0, f65=380 (out)

## Construction steps
1. Timeline 1920x1080 at 23.976 fps; Fusion comp 66 frames. Local 0 = master frame 941.
2. Background: warm plate (#6f6752) with a lighter pillar band at right and a dark figure ellipse centre-left, standing in for the blurred interview.
3. Text+ white weight 600: number 64 px at (395,210) from frame 1; verb 84 px at (90,340) Write On 4-15; second word 72 px at (1240,530) hard-on at 9; third word 72 px at (1540,720) Write On 12-21. No carets.
4. Exit: number off at 57; rectangle masks sweep left to right over the verb 57-62, the third word 60-65 and the second word 62-66.
5. Easy mistake: sequencing the reveals. The second word lands while the verb is still typing, and the third starts before the verb finishes; the overlaps are the rhythm.

## Content variants
Default copy:
- num: 2.
- w1: Connect
- w2: Your
- w3: Sources
Adaptation changes:
- num: 3.
- w1: Schedule
- w2: The
- w3: Report

## 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.
