Floating Catalogue of Creative Skills
A 24-second motion-design catalogue that cycles pastel browser cards through five fields before a grid reveal.
By @Y0urDumble
get_blueprint("REF-6263") paste it into Claude to read this blueprint. Connect Claude
All 1 shots, with timecodes
Why it works
The first card is readable at frame 0 while neighbouring cards promise more. One continuous 24-second composition cycles through categories, with a measured reframe at frame 385 aligned to a beat hit; the grid payoff appears as brightness peaks at 20.2 seconds.
Look and sound
Read from the frames and the soundtrack.
Palette
Sampled from the opening frames.
Sound
- Voice
- None
- Tempo
- 75 BPM
- Loudness
- -26.5 LUFS
Quiet tonal electronic bed at 75 BPM in G major, with soft whooshes and selected hits on the beat.
Motion and type
- Pace
- 0.0 cuts a minute
- First 10 s
- 0 cuts
- Longest shot
- 24.00 sec
Likely made with
- Motion
- Remotion
- 3d
- Blender
- Music
- A licensed or generated music track
Estimated from the frames and sound, not confirmed by the maker.
Claude can make this
Hook sheet
The first 3 seconds, every quarter second.
Cut from the film’s hook sheet. Claude reads the same sheet over MCP.
Blueprint
15 plates that take the film apart, from its format to a recipe you can run. Claude reads the same plates over MCP.
Plate 01
The format in one breath
A large floating catalogue card presents one item at a time while its neighbours peek into view. A sequence of category changes makes a growing collection feel navigable; the final pull-back reveals the whole set.
Why it works
- (inferred) The first item is readable at frame 0, so the hook starts with evidence rather than a setup.
- (inferred) The beat grid gives the carousel predictable momentum; the marked reframe at frame 385 renews it without a cut.
- (inferred) The brightness peak at 20.2 s supports the grid payoff.
Plate 02
Format card
| Platform · aspect · length | X · 4:5 · 720×900 source · 1080×1350 deliverable · 24 s · 60 fps |
| Pace | One continuous shot, zero cuts; item changes inside it |
| Script | Music-led; wordless; card names and category headings carry meaning |
| Voice | None |
| On screen | Floating browser cards, small dimensional vignettes, terminal-like inset, final grid |
| Type | Two measured levels: 32–58 and 66–74 px cap at 1080p |
| Sound | Quiet tonal bed, steady 75 BPM; measured −26.5 LUFS, −4.8 dBTP |
| Structure | Seven beats inside shot 1; grid payoff near 20.2 s; CTA belongs on a new end-card line |
Tools you'll need: Remotion for motion graphics · Blender for 3D · a licensed or generated music track · Claude alone: yes
Frame sheets
Frame by frame
Stills taken at measured moments, each labelled with its time. Claude gets the same sheets over MCP.
Plate 03
The hook, frame by frame (first 3 seconds)
| Time | Picture | On-screen text | Words and sound | What it does |
|---|---|---|---|---|
| 0.00–0.75 | First browser card, pastel objects, neighbouring preview | Category and first item | No speech; bed begins | Establishes the system instantly |
| 0.75–1.50 | Second card comes forward | New item name | Arrival whoosh near 1.03 s | Promises a sequence |
| 1.75–2.75 | Third card then a new vignette | Item names replace each other | Hit at frame 98; whoosh at 158 | Makes the repetition legible |
Hook formula (inferred): visible collection + first proof: {{CATEGORY}} / {{ITEM_NAME}} / {{VERIFIED_PROOF}}.
Why it holds (inferred): The card hierarchy stays stable, but the central object and title change before the viewer has exhausted the list.
Plate 04
Structure (beat sheet)
| # | Time (s) | Shots | Beat | What happens | Technique |
|---|---|---|---|---|---|
| 1 | 0–1 | 1 | Hook | First large item card is already present | Result first |
| 2 | 1–4 | 1 | Pattern | More items advance in the first category | Sliding stack |
| 3 | 4–8 | 1 | New field | Video examples enter; frame 385 reframes the stack | Accent and object swap |
| 4 | 8–12.5 | 1 | New field | Green dimensional examples advance | Familiar card shell |
| 5 | 12.5–16.5 | 1 | New field | Motion examples advance | Palette and object change |
| 6 | 16.5–20.2 | 1 | Final field | Brand examples complete the tour | Last category pivot |
| 7 | 20.2–24 | 1 | Payoff | The stack becomes a many-card grid under a large claim | Pull-back and collection view |
Retention map (inferred): category pivots around 4, 8, 12.5 and 16.5 s; payoff begins near 20.2 s; the end is a grid hold and hard stop. These are picture changes within shot 1.
For agents: Shots
The measured data and build specs agents work from. Over MCP they get it all, with the full timeline as JSON.
Shots (measured)
One continuous shot, 24.00 s = 1440 frames at 60 fps. Frames are at 60 fps, first and last inclusive: a shot "33–70 · 38 f" is a Remotion <Sequence from={33} durationInFrames={38}>; at another frame rate, use the seconds.
| # | in–out (s) | frames @60 (first–last · count) | dur (s) | transition in | changes inside | scales (inferred) |
|---|---|---|---|---|---|---|
| 1 | 0.00–24.00 | 0–1439 · 1440 f | 24.00 | start | f385 reframe | scales |
scales: fixed = keep this shot's length when the film gets longer or shorter · scales = stretch or shrink it with the length · optional = drop it for the 15 s cut.
Measurement notes: S1: 357 text items read; the 40 largest are kept · cue sheet keeps 24 of 47 sound events (all near cuts, then the strongest)
Frame sheets (measured): shots, one labelled frame per shot (labels S<n> m:ss.s; + = just after a change inside the shot) · the full-length timeline, the whole film at fixed steps (labels S<n> m:ss.s f<frame>), 2 pages · close-up strips, 14 frames around each cut and change: f385 reframe (sheet "transitions", 1 page; labels S<n> m:ss.sss f<frame>): call get_frames('2120197632500887856'), or see them on the film page (Pro, or free films).
Plate 05
Shot grammar
Mix: One continuous motion-graphic composite. Rhythm: The film changes cards frequently while keeping the same information hierarchy; the end replaces a large card with a compact grid.
Rules this editor follows (inferred):
- Keep neighbouring cards partially visible so each arrival reads as travel through a collection.
- Change category label, accent, item title and vignette as one unit; keep the outer card component consistent.
- Treat the camera values as screen-space evidence. The big cumulative drift includes many card moves, not a single uninterrupted pan.
At a glance
- Biggest moves: shot 1: complex over 24.00 s, irregular.
For agents: Camera and motion · Transitions
The measured data and build specs agents work from. Over MCP they get it all, with the full timeline as JSON.
Camera and motion (measured)
Per shot, measured on the frames. "Camera move" names the camera's move; the numbers are the image's motion: pan in % of frame width (x) or height (y) per second, +x = content moves right (so a pan left), +y = content moves down; zoom per second > 1 = content grows (push in). "Over the shot" is the total, with px at 1080×1350. Speed is the move through the shot, about one character per 0.5 s (▁ still → █ its peak), then the peak in % of width per second. Cadence: "ones" = a new picture every frame, "twos" = each picture held two frames. Blur is relative to this film (0 = its sharpest shots, 1 = its blurriest); motion = share of the frame changing, 0–1. "?" = confidence under 0.5: a hint, check the sheets. The labels describe how the image moves, not where a 3D camera went: an object flying toward a still camera can read as a push in.
| # | camera move | pan x · y (%/s) | zoom (/s) | over the shot | ease | speed · peak | cadence | blur | motion | conf |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | complex (shake 37.4%/s) | −9.1 · −4.1 | ×0.985 | x −218.9% (−2364 px) · y −98.8% (−1334 px) · ×0.70 scale · roll −13.7° · jump at f385: ×0.76, x −23.9%, y +2.8% | irregular | ▂▁▁▂▁▂▁▁▁▁▁▁▁▁▁▁ · 200%/s | ~51 fps | 0.40 | 0.19 | 0.74 |
| ↳ 0.00–0.33 s: static | +0.0 · +0.0 | ×1.000 | none | hold | ||||||
| ↳ 0.33–0.67 s: static | +0.0 · +0.0 | ×1.000 | none | hold | ||||||
| ↳ 0.67–1.17 s: complex | −82.4 · −34.0 | ×0.646 | x −41.2% (−445 px) · y −17.0% (−229 px) · ×0.80 scale · roll −5.8° | ease-out | ||||||
| ↳ 1.17–1.42 s: static | +0.0 · +0.0 | ×1.000 | none | hold | ||||||
| ↳ 1.42–2.00 s: pan right | −20.4 · −0.9 | ×0.889 | x −11.9% (−129 px) · y −0.6% (−7 px) · ×0.93 scale · roll +1.6° | ease-out | ||||||
| ↳ 2.00–2.42 s: static | +0.0 · +0.0 | ×1.000 | none | hold |
Focus (measured): shot 1: sharp inside 0 · 0 · 100 · 100, no defocus (deep focus). Box = x · y · w · h % of the frame. In 2D, blur the layers outside the box by that many px; in 3D, set the focus distance on the box and open the aperture until the background blur matches.
Biggest moves: shot 1: complex over 24.00 s, irregular. Locked-off shots: 0 of 1. Animation cadence: about 50.4 new pictures a second: 50.4 unique frames per second of 60 while things move (held-frame share 0.16, 605 moving frames). Match it, or the motion will feel smoother or choppier than the reference.
Transitions (measured)
One continuous shot: no cuts. Changes happen inside the shot (camera, light, layout); the measured ones:
| shot | at (s) | frame @60 | change |
|---|---|---|---|
| 1 | 6.42 | 385 | reframe: jump x0.76 |
What shot 1 shows (inferred from the sheets)
| # | Type | Framing | What's on screen | Text / overlay |
|---|---|---|---|---|
| 1 | Continuous catalogue composite | Large tilted browser card with off-edge neighbours, then a grid | Original item vignettes by category | Category, item, description, proof, counter and end claim |
Plate 06
Visual style guide
- Look (inferred): A bright paper-coloured field, white rounded cards, softly lit toy-scale objects and shallow contact shadows. The reference uses pink, green, violet and gold accents; map their category roles to {{BRAND_ACCENTS}} on {{BRAND_BG}}.
- Graphics (inferred): A browser strip, dark code-like inset, small badge, counter and bottom progress dashes. Preserve the hierarchy; replace all literal marks and claims.
- Transitions (inferred): Slide and rotate overlapping cards inside one composite, then reduce them into the grid. The transition table in plate 5 has the only detected discrete reframe.
- Branding and end card (inferred): Add {{LOGO}} and {{CTA_URL}} as text beneath the grid in the new version; preserve a readable final hold rather than using a QR code.
At a glance
- Type scale (cap px): 66–74 · 32–58 (2 levels) among the text above.
- Phone minimum: cap height ≥ 2% of frame width = 22 px at 1080×1350
- Whole film (time-weighted):
#F1EAE367% ·#CFC9C920% ·#D8B5A57% ·#211B194% ·#827F7D2%. - Curve facts: average 0.84 · darkest 0.82 at 18.0 s · brightest 0.91 at 20.2 s.
For agents: Type · Type spec · Palette by shot · Brightness curve · Layout boxes · Material card
The measured data and build specs agents work from. Over MCP they get it all, with the full timeline as JSON.
Type (measured)
Sizes are cap height (the height of a capital letter) in px at the 1080p deliverable, 1080×1350. Font size ≈ cap ÷ 0.71 for a typical sans (Inter 0.73, Geist 0.71, SF Pro 0.70, Instrument Sans 0.72): check your own font's cap-height ratio. Cap heights are estimated from text-recognition boxes (median error about 11% on real text; text under about 13 px at 1080p and non-Latin scripts are not read). Box = x · y · w · h as % of the frame, from the top-left.
| text | shot | on screen (s) | cap px | ≈ font px | cap % of height | box x·y·w·h % | position |
|---|---|---|---|---|---|---|---|
| "Design" | 1 | 0.0–19.2 | 52 | 74 | 3.9 | 12 · 9 · 17 · 6 | left, top |
| "anthropic skills" | 1 | 0.0–0.8 | 42 | 60 | 3.1 | 12 · 57 · 42 · 5 | left, lower third |
| "Video" | 1 | 4.0–8.0 | 55 | 77 | 4.1 | 13 · 9 · 14 · 5 | left, top |
| "video-shotcraft" | 1 | 5.0–6.2 | 35 | 49 | 2.6 | 41 · 68 · 32 · 4 | centre, lower third |
| "remotion skills" | 1 | 5.8–7.0 | 36 | 51 | 2.7 | 41 · 68 · 31 · 4 | centre, lower third |
| "animated shorts" | 1 | 6.5–8.0 | 36 | 51 | 2.7 | 20 · 61 · 37 · 4 | centre, lower third |
| "blender mcp" | 1 | 7.2–8.8 | 34 | 48 | 2.5 | 30 · 65 · 30 · 4 | centre, lower third |
| "unity mcp" | 1 | 9.0–10.2 | 32 | 46 | 2.4 | 42 · 70 · 21 · 4 | centre, lower third |
| "universal-modder" | 1 | 9.8–11.2 | 36 | 51 | 2.7 | 40 · 68 · 38 · 4 | centre, lower third |
| "MOTION." | 1 | 11.5–16.0 | 74 | 104 | 5.5 | 35 · 36 · 36 · 6 | centre, upper |
| "anything2explainer" | 1 | 13.8–15.0 | 46 | 65 | 3.4 | 12 · 57 · 54 · 5 | centre, lower third |
| "oil-motion" | 1 | 15.2–16.0 | 46 | 65 | 3.4 | 12 · 57 · 27 · 4 | left, lower third |
| "stitch skills" | 1 | 16.0–17.5 | 35 | 49 | 2.6 | 43 · 69 · 23 · 4 | centre, lower third |
| "Brand" | 1 | 16.2–20.0 | 58 | 81 | 4.3 | 12 · 9 · 17 · 5 | left, top |
| "architecture diagrams" | 1 | 17.0–18.2 | 39 | 55 | 2.9 | 11 · 57 · 60 · 4 | centre, lower third |
| "nothing design" | 1 | 18.5–20.0 | 34 | 48 | 2.5 | 42 · 69 · 31 · 4 | centre, lower third |
| "25 skills." | 1 | 20.2–23.8 | 70 | 99 | 5.2 | 4 · 9 · 25 · 6 | left, top (inside 5% of the edge) |
| "Zero excuses." | 1 | 20.2–23.8 | 66 | 93 | 4.9 | 4 · 16 · 39 · 6 | left, top (inside 5% of the edge) |
Plus 3 smaller or briefer text items, cap 26–44 px, mostly labels inside interfaces: all in the timeline JSON.
Type scale (cap px): 66–74 · 32–58 (2 levels) among the text above.
Left out: 6 reads that are not clean words or were seen in a single frame ("399", "hyperframes", "OpenMontage"…): blurred, moving or partly hidden text the recognizer misread. They stay in the timeline JSON, not in the type scale.
Phone minimum: cap height ≥ 2% of frame width = 22 px at 1080×1350. Why: in a phone feed the film fills the screen width, about 390 pt, and 2% of that is 7.8 pt, the cap height of 11 pt text, Apple's smallest recommended text size. Words the viewer must read go at or above it; smaller text can only be decoration. Every measured text item in the reference clears it.
Type spec (inferred)
Assume cap height is 0.71 of CSS font size. The closest free fallback for {{BRAND_FONT}} is Inter for the sans text; use the brand's own italic display face for category labels.
| Level | Cap px at 1080p | Font size | Weight · case · tracking | Colour pair | Motion |
|---|---|---|---|---|---|
| Display / payoff | 66–74 | 93–104 px | 500–700 · sentence case · −0.035em | foreground on {{BRAND_BG}} | 12 f opacity and y −12 px, cubic-bezier(.2,.8,.2,1) |
| Category / item | 32–58 | 45–82 px | category 500 italic, title 800 · original case · −0.025em | {{BRAND_ACCENTS}} on {{BRAND_BG}}; dark foreground on white card | 8 f opacity and y +10 px, cubic-bezier(.2,.8,.2,1); keep all readable words ≥22 px cap |
Palette by shot (measured)
Colours measured on the frames, with their share of the frame; luma 0 = black, 1 = white. These are the reference's brand colours: map them onto {{BRAND_BG}} and {{BRAND_ACCENTS}} by role (background, surface, accent), don't reuse them as the format. Neighbouring shots with the same palette share a row.
| # | luma mean · p95 | colours (share of frame) | highlights (bright pixels) |
|---|---|---|---|
| 1 | 0.84 · 1.00 | #F1EAE3 66% · #CFC9C9 20% · #D8B5A5 7% · #211B19 4% · #827F7D 2% | #FFFFFF ≈6500 K 99% |
Whole film (time-weighted): #F1EAE3 67% · #CFC9C9 20% · #D8B5A5 7% · #211B19 4% · #827F7D 2%.
Brightness curve (measured)
Mean luma per 1 s, one character each, grouped in tens; p95 = the bright highlights. Scale: ▁ = 0 (black) → █ = 1.0 (white); luma runs 0–1.
0s 10s 20s
mean ▇▇▇▇▇▇▇▇▇▇ ▇▇▇▇▇▇▇▇▇▇ █▇▇▇
p95 ██████████ ██████████ ████
Curve facts: average 0.84 · darkest 0.82 at 18.0 s · brightest 0.91 at 20.2 s.
Layout boxes (inferred)
| Element | Shots | Box x · y · w · h (%) | Source · confidence | Notes |
|---|---|---|---|---|
| Category heading | 1 | 12 · 9 · 17 · 6 | measured text box · high | First field |
| First item title | 1 | 12 · 57 · 42 · 5 | measured text box · high | Lower information band |
| Motion display | 1 | 35 · 36 · 36 · 6 | measured text box · high | Vignette text |
| Final count | 1 | 4 · 9 · 25 · 6 | measured text box · high | End card, upper left |
| Final payoff | 1 | 4 · 16 · 39 · 6 | measured text box · high | End card, second line |
| Card grid | 1 | 4 · 22 · 91 · 65 | timeline tile at 22.0 s · medium | Re-layout to leave CTA space |
End card (inferred): Use {{BRAND_BG}}, place {{LOGO}} and text {{CTA_URL}} beneath the grid in a 4 · 89 · 91 · 7% band. The reference has no such CTA; the measured large end text remains on screen to 23.8 s.
Material card (inferred)
:root { --brand-bg: {{BRAND_BG}}; --brand-accent: {{BRAND_ACCENTS}}; --card: #fff; --ink: #201f20; }
.stage { background: var(--brand-bg); perspective: 1400px; }
.card { background: var(--card); border: 1px solid color-mix(in srgb,var(--ink) 12%,transparent); border-radius: 20px; box-shadow: 0 18px 38px rgb(0 0 0 / .14); transform-style: preserve-3d; }
.window { background: radial-gradient(circle at 50% 18%,color-mix(in srgb,var(--brand-accent) 17%,white),var(--brand-bg) 70%); }
.object { filter: drop-shadow(0 15px 18px rgb(0 0 0 / .16)); }
Blender · optional vignette objects: Principled BSDF roughness .42, transmission 0, IOR 1.45, coat .12, emission 0; world strength .65; EEVEE 64 samples; 50 mm lens, f/8. Render RGBA objects at 1080×1350, 60 fps.
2D equivalent · deep focus across 0 · 0 · 100 · 100% (0 px defocus); soft area-light glow 50 px at .12 opacity; vignette .08 strength with 85% radius. Rounded rectangles and layered extrusion replace physical depth.
Plate 07
Voice and delivery
Measured: No speech detected.
Voice direction (inferred): No voice; the viewer reads each card. Use concise item names and let the post carry long explanations.
Plate 08
Sound design
- Music: A soft tonal electronic bed with a steady pulse supports the item procession.
- Effects: The cue sheet places hits on some arrivals and the frame-385 reframe; reserve whooshes for the cards named in the cue sheet.
- Space: Leave the final grid uncluttered by extra impacts; the measured energy lifts in the final section before falling away.
At a glance
- Tempo: 75 BPM, confidence 0.97 · first downbeat 2.440 s (bar start confidence 0.75) · beat 0.8000 s · bar of 4 = 3.200 s.
- Key: G major (confidence 1.00; runner-up G minor).
- Loudness of the reference (measured): −26.5 LUFS integrated, true peak −4.8 dBTP, range 2.0 LU. That is how this film was mastered, not a target.
For agents: Sound grid · Original music bed
The measured data and build specs agents work from. Over MCP they get it all, with the full timeline as JSON.
Sound grid (measured)
Tempo: 75 BPM, confidence 0.97 · first downbeat 2.440 s (bar start confidence 0.75) · beat 0.8000 s · bar of 4 = 3.200 s.
Beat grid in frames. A constant grid fitted to the music: beat k is at round((first downbeat + k × beat) × fps); round each beat on its own, since adding a rounded beat length drifts.
| per beat | per bar | first downbeat (frame) | downbeats (frame), first 7 | |
|---|---|---|---|---|
| 60 fps (this film) | 48.00 f | 192.00 f | 146 | 146 · 338 · 530 · 722 · 914 · 1106 · 1298 |
| 30 fps | 24.00 f | 96.00 f | 73 | 73 · 169 · 265 · 361 · 457 · 553 · 649 |
Beats in seconds: 0.040 · 0.840 · 1.640 · 2.440 · 3.240 · 4.040 · 4.840 · 5.640 · 6.440 · 7.240 · 8.040 · 8.840 … (30 in all, audio.tempo.beats in the timeline JSON).
Key: G major (confidence 1.00; runner-up G minor).
Sections (by energy of the whole mix):
| section | time (s) | shots | level (RMS) |
|---|---|---|---|
| steady | 0.0–20.0 | 1 | −30 dB |
| steady | 20.0–24.0 | 1 | −28 dB |
Energy map (RMS per 1 s, ▁ = −47 dB → █ = −24 dB):
▇▆▅▅▇▆▅▅▇▆ ▅▅▇▆▅▅▇▆▅▅ █▆▄▁
Cue sheet (24 cues: 0 at a cut, 12 on a beat). "On screen" names the picture event the cue hits within ±3 frames (a cut, a jump, a move starting, a line of text appearing, a brightness step), or says there is none:
| t (s) | frame @60 | shot | kind | strength | on screen (±3 f) | on beat |
|---|---|---|---|---|---|---|
| 0.24 | 14 | 1 | whoosh | 1.00 | no visible event | |
| 1.03 | 62 | 1 | whoosh | 0.83 | "open-design" appears | |
| 1.62 | 98 | 1 | hit | 0.94 | no visible event | ✓ |
| 2.63 | 158 | 1 | whoosh | 0.87 | "hallmark" appears | |
| 3.23 | 194 | 1 | hit | 1.00 | camera complex starts (shot 1) | ✓ |
| 3.44 | 206 | 1 | whoosh | 0.83 | no visible event | |
| 4.83 | 290 | 1 | hit | 1.00 | camera complex starts (shot 1) | ✓ |
| 6.43 | 386 | 1 | hit | 1.00 | jump ×0.76 in shot 1 | ✓ |
| 6.63 | 398 | 1 | whoosh | 0.85 | "animated shorts" appears | |
| 7.22 | 433 | 1 | hit | 0.84 | "blender mcp" appears | ✓ |
| 7.44 | 446 | 1 | whoosh | 0.85 | no visible event | |
| 8.82 | 529 | 1 | hit | 0.95 | no visible event | ✓ |
| 9.83 | 590 | 1 | whoosh | 0.83 | "universal-modder" appears | |
| 10.43 | 626 | 1 | hit | 1.00 | camera pan left starts (shot 1) | ✓ |
| 10.63 | 638 | 1 | whoosh | 0.83 | "gamedev skills" appears | |
| 11.43 | 686 | 1 | whoosh | 0.87 | "Motion" appears | |
| 13.03 | 782 | 1 | whoosh | 0.83 | no visible event | |
| 13.63 | 818 | 1 | hit | 1.00 | "anything2explainer" appears | ✓ |
| 15.23 | 914 | 1 | hit | 1.00 | "oil-motion" appears | ✓ |
| 16.83 | 1010 | 1 | hit | 1.00 | camera pan left starts (shot 1) | ✓ |
| 17.03 | 1022 | 1 | whoosh | 0.87 | "architecture diagrams" appears | |
| 17.62 | 1057 | 1 | hit | 0.91 | "logo design" appears | ✓ |
| 19.23 | 1154 | 1 | hit | 0.99 | camera complex starts (shot 1) | ✓ |
| 19.43 | 1166 | 1 | whoosh | 0.85 | no visible event |
Loudness of the reference (measured): −26.5 LUFS integrated, true peak −4.8 dBTP, range 2.0 LU. That is how this film was mastered, not a target.
Loudness targets for your version (common practice per platform; the brief overrides them):
| where it plays | integrated | true peak max | note |
|---|---|---|---|
| YouTube | −14 LUFS | −1 dBTP | louder masters are turned down; quieter ones are not turned up |
| X, LinkedIn (this reference) | −14 LUFS | −1 dBTP | no published normalisation: the file plays as mastered |
| TikTok, Reels, Shorts | −14 to −10 LUFS | −1 dBTP | played on phone speakers; short-form is often mastered hotter |
| Website or app (autoplay) | −16 LUFS | −1 dBTP | often starts muted; keep the unmuted level comfortable |
| Broadcast (EBU R128) | −23 LUFS | −1 dBTP | only when the brief asks for it |
Original music bed (inferred)
Write an original G-major, 75 BPM bed: muted pluck on beats, soft rounded bass, brushed shaker and airy two-note synth. Beats 1–2 keep a small motif; beats 3–6 change chord voicing at each category; beat 7 opens the voicing for the grid and lets the tail decay. Put transient accents on cue frames 98, 194, 290, 386, 626, 818, 914, 1010 and 1154; keep the remaining listed whooshes light. Music-model prompt: Original soft electronic catalogue pulse, 75 BPM, G major, muted pluck, rounded bass, airy synth, restrained transients, no vocals, spacious grid reveal.
Plate 09
Packaging
- Title (inferred):
{{COLLECTION_COUNT}} useful {{CATEGORY}} picks for {{PRODUCT}}with a count matching the actual grid. - Cover (inferred): One large first card, a clearly visible next-card edge and a short count promise.
- Post copy (inferred): Open with the verified benefit, then list items and links; put {{CTA_URL}} in plain text.
- Audience signals: The supplied context lists 56,649 views. No replay graph is supplied.
Plate 10
Script template
(inferred) This is screen copy, not narration. Keep the title and one-line benefit readable during each item's appearance; only verified claims enter the proof field.
[beat 1 · shot 1] {{CATEGORY}} / {{ITEM_NAME}} / {{ITEM_BENEFIT}}
[beat 2 · shot 1] Repeat the same three fields for subsequent items.
[beats 3–6 · shot 1] Change category and accent; keep each item benefit under 9 words.
[beat 7 · shot 1] {{COLLECTION_COUNT}} items. {{PAYOFF}}. {{LOGO}} {{CTA_URL}}
Plate 11
Production plan
Two routes (inferred)
Minimum: Claude alone. Remotion with CSS 3D builds the card shell, perspective stack, simple dimensional objects, grid and a synthesised 75 BPM bed end to end. It simplifies vignette geometry and physical lighting; it gives up nuanced reflections and soft 3D contact shadows. The user supplies real {{PRODUCT}} data, {{LOGO}} and any actual capture.
Full: with the toolkit. Model and light original vignette objects in Blender, then pass 1080×1350 RGBA PNG sequences at 60 fps to Remotion. Pass camera and card keyframes as JSON; Remotion assembles typography, browser chrome, masks and grid. Supply an original or licensed 48 kHz WAV plus 75 BPM click-track WAV for cue alignment; export H.264 MP4 at 1080×1350, 60 fps.
Asset honesty: Use real {{PRODUCT}} captures or UI rebuilt in code. Never generate product UI, fake screenshots, invented proof, reviews or results. The terminal-like inset must display genuine or clearly illustrative text, never a fabricated working command.
Plate 12
Prompts
Script prompt (inferred)
For {{PRODUCT}}, write a wordless catalogue of {{COLLECTION_COUNT}} real items across five {{CATEGORY}} groups. For each, return {{ITEM_NAME}}, a benefit under nine words, and {{VERIFIED_PROOF}} or none. End with a short {{PAYOFF}}. Verify every link and numeric claim; do not invent usage metrics.
Music prompt (inferred): Use the original-bed prompt in plate 8.
Plate 13
Edit recipe
How it's built (inferred)
Signature move, shot 1: Build a 3D card stack in screen space. Keep the active card on top, outgoing card left/back and incoming card right/back. Slide each group's cards along x while counter-rotating the browser face; clamp the primary title to a readable interval. At the measured frame-385 reframe, switch the composition anchor to the next card: scale the group to ×0.76 and move its image x −23.9%, y +2.8%. The transition strip shows adjacent cards overlapping across frames 383–390; implement a 12-frame masked hand-off centred on frame 385, with cubic-bezier(.22,.75,.18,1). The measurement records the jump at frame 385, so the composite must reach the new anchor there. There is no hard cut.
For the opening, follow plate 5's still segments, then the 0.67–1.17 s ×0.80 ease-out with x −41.2% and y −17.0%, followed by the 1.42–2.00 s ×0.93 pan. Use those as image motion, not physical camera travel. Card vignettes float independently with ±8 px y and ±3° rotation over 48–96 frames; category swaps replace their objects under the white information band. Keep about 50 new pictures per second during movement: keyframe at 60 fps, occasionally hold one frame, and use a 180° shutter equivalent (three subframe samples) for fast slides. Avoid making tiny terminal text essential.
For beat 7, parent the card stack to a grid transform and pull back, revealing repeated card instances beneath the measured top text boxes. Keep the grid bright; plate 6 records the film's brightness maximum at 20.2 s. End without a fade to black.
Build order
- Verify the data and copy for beats 1–7 and make one reusable card component.
- Build beat 1, then animate the stack through beats 2–6 inside shot 1, including the frame-385 hand-off.
- Render original vignette assets and composite them below the fixed browser chrome.
- Build beat 7's pull-back and grid; insert the text CTA in the adapted version.
- Align WAV accents to plate 8, render, and inspect every frame around 385 plus the grid payoff.
Plate 14
QA checklist
- One continuous shot, with no hard cuts; the reframe reaches the measured anchor at frame 385.
- Opening image motion agrees with plate 5's segments within ±1 frame and ±3% screen travel.
- Every required word has ≥22 px cap height at 1080×1350; decorative microtype is not needed to understand the card.
- Cue hits land within ±3 frames of the selected plate 8 events and the bed runs at 75 BPM.
- X/LinkedIn master measures near −14 LUFS integrated and ≤−1 dBTP true peak.
- Grid title and payoff sit inside their measured plate 6 boxes; no fade to black is added.
- Count, links and proof labels match the actual items.
Plate 15
Make it yours
Brand slots
| Slot | Replaces in the reference | Notes |
|---|---|---|
{{PRODUCT}} | The named skills collection | Own product or subject |
{{BRAND_FONT}} | Reference display and card fonts | Use your licensed face |
{{BRAND_BG}} | Warm cream field | Supply a light or otherwise legible background |
{{BRAND_ACCENTS}} | Category accent colours | Distinct roles, original palette |
{{LOGO}} | Reference mark and mascot | Own mark, text acceptable |
{{CTA_URL}} | External post links | Plain text destination |
{{CATEGORY}} | Five field labels | Own grouping |
{{ITEM_NAME}} | Skill names | Verified real items |
{{ITEM_BENEFIT}} | One-line descriptions | Under nine words |
{{VERIFIED_PROOF}} | Badges and counters | Omit if unverified |
{{COLLECTION_COUNT}} | Final item count | Must equal grid entries |
{{PAYOFF}} | Final slogan | Own short result claim |
What this film avoids (inferred): no voiceover, people, stock footage, long feature list, screen-recording walkthrough, QR code or store badges. The small browser chrome is decoration, not a real app demo.
Brand-distance warning (inferred): A cream field, pink italic heading, star-faced mascot and identical skill list together would read as the reference. Replace the mascot with your own object language first.
Adapt the aspect (inferred)
- 16:9: Re-lay out. A centre crop of 4:5 loses the upper headings and lower card text; spread the active card and neighbour horizontally, move category to a top safe band and CTA above the bottom platform controls. Required text ≥38 px cap.
- 9:16: Re-lay out. Move the active card into the middle vertical band and stack previews above/below; reserve top and bottom UI safe zones, and keep all required text ≥22 px cap.
- 1:1: Centre crop, then move category and final count/payoff back inside the top safe band; place the CTA above the bottom UI overlay. Required text ≥22 px cap.
Adapt the length (inferred)
- 15 s: Keep beats 1 and 7; compress item dwell and use fewer entries within beats 2–6, preserving the signature card hand-off.
- 30 s: Keep the seven-beat order and add verified items within beats 2–6; retain the grid reveal.
- 60 s: Add category chapters and more real items, with new vignettes and rhythmic changes; do not merely stretch the closing grid.
The creator's signature (don't copy): The exact skill list, mascot, post claim, palette and music.
The transferable format (do copy): A readable floating card system, beat-paced category tour, overlapping reframe and final collection pull-back.
Source: https://x.com/Y0urDumble/status/2120197632500887856 · measured 2026-11-11 · (measured) = film evidence · (inferred) = construction reading
For agents: Aspect conversions · Length cuts
The measured data and build specs agents work from. Over MCP they get it all, with the full timeline as JSON.
Aspect conversions (computed from the measured boxes)
Aspect conversions (computed from the measured boxes): a centre crop of the reference frame, checked against its text boxes and moving elements.
| to | deliver at | a centre crop keeps | measured boxes kept | boxes a crop loses | phone minimum | so |
|---|---|---|---|---|---|---|
| 16:9 | 1920×1080 | middle 45.0% of the height (cut 27.5% top and bottom) | 13 of 21 | "Design", "huashu design", "Video", "unity mcp" … | 38 px cap | re-lay out: a crop loses too much |
| 9:16 | 1080×1920 | middle 70.3% of the width (cut 14.8% each side) | 11 of 21 | "Design", "anthropic skills", "hallmark", "Video" … | 22 px cap | re-lay out: a crop loses too much |
| 1:1 | 1080×1080 | middle 80.0% of the height (cut 10.0% top and bottom) | 17 of 21 | "Design", "Video", "Brand", "25 skills." | 22 px cap | centre crop, then move the lost boxes inside |
Length cuts (computed from the scales flags)
Length cuts (computed from the scales flags): frames per shot at 60 fps; fixed shots keep their length, optional shots go first when the cut is shorter, the scaling shots share what is left. Snap each cut to the nearest beat when plate 8 has a steady pulse.
| # | flag | reference (f @60) | 15 s cut | 30 s cut | 60 s cut |
|---|---|---|---|---|---|
| 1 | scales | 1440 | 900 | 1800 | 2304 |
| total | 1440 f · 24.0 s | 900 f · 15.0 s | 1800 f · 30.0 s | 2304 f · 38.4 s + 21.6 s new beats |
60 s: stretching tops out at ×1.6 (38.4 s); add 21.6 s of new beats (more of the scaling beats' pattern with new content), never a longer hold on the fixed shots.
Recipe
One prompt that makes a film like this with Claude. Use the method; make the footage, words and brand your own.
Make a 24 s 4:5 film for {{PRODUCT}} in the format of REF-2120197632500887856 (floating catalogue). Brand: font {{BRAND_FONT}}, background {{BRAND_BG}}, accents {{BRAND_ACCENTS}}, logo {{LOGO}}, call to action {{CTA_URL}}. Supply real {{CATEGORY}}, {{ITEM_NAME}}, {{ITEM_BENEFIT}}, {{VERIFIED_PROOF}}, {{COLLECTION_COUNT}} and {{PAYOFF}} data. First load get_blueprint("2120197632500887856", format: "json"); build to its shot frames, camera numbers, type boxes and beat grid. This is music-led and wordless: establish the first large browser card immediately, cycle overlapping item cards through five fields, and pull back into a grid for the final payoff. Use Blender Python for original dimensional vignette objects, export RGBA PNG sequences at 1080×1350 and 60 fps, and assemble the cards, type, masks and grid in Remotion. As a simpler fallback, build the whole film in Remotion with CSS 3D and synthesised music. Match the frame-385 ×0.76 reframe, x −23.9%, y +2.8%, with a 12-frame masked overlap; keep it one continuous composition. Keep required words at least 22 px cap. Create an original G-major 75 BPM bed, align accents to the JSON cue frames, and master for X near −14 LUFS, true peak ≤−1 dBTP. Use real captures of {{PRODUCT}} or UI rebuilt in code; never generated product UI, fake screenshots or fake results. Render 1080×1350, 60 fps, H.264 MP4, about 24 seconds. End with text {{LOGO}} plus {{CTA_URL}}; no QR or badges. Never reuse the reference's hex colours, fonts, logo, mascot, slogans or music.Films like this
Tagged motion designcataloguebrowser mockuppastel 3dcard carouselgrid revealsilent4:5