Compare commits
1
Commits
| Author | SHA1 | Date | |
|---|---|---|---|
 Claude
|
d4b3cf1b85 |
No files matched your search
Whitespace-only changes.
@@ -0,0 +1,84 @@
|
||||
---
|
||||
name: after-effects
|
||||
description: This skill should be used when the user asks anything about Adobe After Effects — to "write an After Effects expression", "make a wiggle / loopOut / inertial bounce", "rig a control with a slider / null", "build a MOGRT / Essential Graphics template", "export AE to Lottie or Media Encoder"; OR to "export from After Effects", "export with transparency / an alpha channel", "fix a black background on export", "choose ProRes vs H.264", "which codec / what bitrate for 1080p / 4K", "stop H.264 disappearing from the Render Queue"; OR when "After Effects is slow / laggy / crashing / froze", "RAM preview won't play in real time", "AE is using all my memory / render takes forever", "how do I speed up After Effects"; OR to "organize my After Effects project", "relink missing footage", "fix broken links after moving files", "set a folder structure / naming conventions", "batch rename layers", or "write an ExtendScript / .jsx script". Covers expressions & rigging, export recipes, performance triage, and project hygiene.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# After Effects
|
||||
|
||||
After Effects is GUI software, so this skill delivers knowledge and ready-to-paste assets — expressions, rig setups, export checklists, diagnostic flows, naming schemes, and ExtendScript — rather than a rendered file. It is a hub covering four domains; route to the matching reference for depth and complete recipes.
|
||||
|
||||
## Routing
|
||||
|
||||
| If the user wants to… | Go to |
|
||||
|---|---|
|
||||
| Write/debug an expression, rig sliders & nulls, build a MOGRT, export to Lottie | `references/expressions.md` (deep cookbook in `references/expression-library.md`) |
|
||||
| Choose a codec/container, export with alpha, fix a black background, set bitrate, recover H.264 | `references/export-recipes.md` |
|
||||
| Fix slow/laggy previews, slow renders, crashes, freezes, out-of-memory | `references/performance-triage.md` |
|
||||
| Organize the project, relink missing footage, set naming/folders, batch-automate with .jsx | `references/project-hygiene.md` |
|
||||
|
||||
The `.jsx` automation scripts live in `scripts/`.
|
||||
|
||||
## Expressions & rigging
|
||||
|
||||
Expressions are JavaScript evaluated on the property they are pasted on; `value` is that property's value, `time` is comp time in seconds, and `index`, `thisLayer`, `thisComp` are in scope. An expression overrides keyframes unless it incorporates `value`.
|
||||
|
||||
```js
|
||||
wiggle(2, 30); // 2 wiggles/sec, amplitude 30 (property units)
|
||||
loopOut("cycle"); // repeat keyframed range forever; "pingpong" / "offset" / "continue"
|
||||
seedRandom(index, true); // stable-but-unique per layer (timeless) before random()/wiggle()
|
||||
thisComp.layer("Lead").transform.position.valueAtTime(time - 0.2 * index); // staggered echo
|
||||
```
|
||||
|
||||
Inertial bounce (overshoot/settle after the last keyframe) and the full loop/wiggle/rigging set are in `references/expressions.md`.
|
||||
|
||||
Rig controllable motion: add a null with **Slider / Checkbox / Color / Angle / Point** control effects, then reference them and parent transforms to drive many layers from one.
|
||||
|
||||
```js
|
||||
ctrl = thisComp.layer("Controls");
|
||||
amt = ctrl.effect("Intensity")("Slider"); // Checkbox → 1/0, Color → [r,g,b,a]
|
||||
value * amt;
|
||||
```
|
||||
|
||||
Expose those controls via **Window → Essential Graphics** and **Export Motion Graphics Template** (.mogrt) for Premiere/AE editors.
|
||||
|
||||
## Export recipes
|
||||
|
||||
Two engines, and choosing wrong is the root of most export confusion:
|
||||
- **Render Queue (RQ)** — masters/intermediates: ProRes, image sequences, lossless, alpha.
|
||||
- **Adobe Media Encoder (AME)** — delivery: H.264 MP4, H.265, social presets.
|
||||
|
||||
Since AE CC 2014/2015 the native **H.264 output module was removed from the RQ** — make MP4s in AME, not the RQ. **H.264/MP4 cannot store alpha**, so a transparent comp exported to H.264 always fills with black; that is the classic "black background" complaint and not a bug. For transparency use **ProRes 4444**, a **PNG/TIFF sequence**, or **HEVC-alpha (Safari) + VP9/WebM-alpha (Chrome)**, with Output Module **Channels: RGB + Alpha**.
|
||||
|
||||
Quick targets: editor handoff → **ProRes 422 HQ**; master/grade or alpha → **ProRes 4444**; web/social/client → **H.264 via AME**, VBR 2-pass (~10–16 Mbps 1080p, ~35–45 Mbps 4K). Full matrix, social specs, and step-by-step alpha exports in `references/export-recipes.md`.
|
||||
|
||||
## Performance triage
|
||||
|
||||
Work top-down by symptom, cheapest fix first, and stop when it clears. Most settings are in `Edit > Preferences` (Win) / `After Effects > Settings` (Mac).
|
||||
|
||||
- **Slow previews:** Half/Third resolution → enable **Multi-Frame Rendering** → raise RAM allocation (leave ~6 GB for OS) → Adaptive fast previews → purge cache → Disk Cache on SSD/NVMe → solo heavy layers/effects → HW-accelerated decode → proxy heavy comps.
|
||||
- **Slow renders:** confirm MFR on → RQ for masters / AME for delivery → HW decode H.264 → pre-render slow precomps → cheaper effects → transcode long-GOP to ProRes.
|
||||
- **Crashes/freezes:** purge + empty disk cache → update **NVIDIA Studio** (not Game-Ready) drivers → reset prefs (`Ctrl/Cmd+Alt+Shift` on launch) → pull third-party plugins → free disk space → disable MFR if MFR-specific → roll back AE after a bad update.
|
||||
|
||||
MFR needs several GB RAM per active core. Ordered flows, exact preference paths, and the proxy workflow are in `references/performance-triage.md`.
|
||||
|
||||
## Project hygiene
|
||||
|
||||
Keep projects organized, portable, and automatable.
|
||||
|
||||
- **Folders:** numeric-prefixed bins (`01_Comps`, `02_Precomps`, `03_Footage`, …, `99_Old`) for stable sort order.
|
||||
- **Naming:** `PROJECT_section_descriptor_v01`; precomps `PRE_`; rename solids/nulls to their role (`CTRL_master`); no spaces/special chars for farm/cross-OS safety.
|
||||
- **Broken links:** AE references footage by path, so moving files breaks links. Double-click a missing item → **Replace Footage > File** and AE relinks the whole folder at once; enumerate with `File > Dependencies > Find Missing Footage`. **Always hand off projects via `File > Dependencies > Collect Files`**, never a bare `.aep`.
|
||||
- **ExtendScript (.jsx):** wrap batches in `app.beginUndoGroup(...) / app.endUndoGroup()`; apply effects by **matchName** (e.g. `"ADBE Gaussian Blur 2"`), not the display name. Ready scripts: `scripts/batch-rename.jsx`, `scripts/sequential-offset.jsx`, `scripts/apply-effect-to-layers.jsx`.
|
||||
|
||||
Full naming scheme, label-color system, relink/Collect-Files/Reduce workflows, and scripting setup in `references/project-hygiene.md`.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/expressions.md` — expressions, rigging, MOGRT, Lottie/Media Encoder basics.
|
||||
- `references/expression-library.md` — deeper expression cookbook: tuned inertial bounce, all loop variants, advanced wiggle, valueAtTime echoes, seedRandom grids, full rigging + coordinate spaces, MOGRT and Lottie/Bodymovin steps, Media Encoder presets.
|
||||
- `references/export-recipes.md` — full codec/alpha/bitrate matrix, per-platform social specs, step-by-step transparency exports, AME preset cautions.
|
||||
- `references/performance-triage.md` — ordered preview/render/crash flows, preference paths and values, MFR/GPU/disk-cache tuning, pre-render & proxy workflow.
|
||||
- `references/project-hygiene.md` — folder template, naming and label systems, relink/Collect Files/Reduce details, scripting setup.
|
||||
- `scripts/batch-rename.jsx`, `scripts/sequential-offset.jsx`, `scripts/apply-effect-to-layers.jsx` — ExtendScript automations.
|
||||
</content>
|
||||
@@ -0,0 +1,168 @@
|
||||
# Export Recipes — Codec, Container, Alpha, Bitrate
|
||||
|
||||
Pick the correct codec, container, alpha handling, and bitrate for a given delivery target, and avoid the classic failures: black backgrounds where transparency was expected, H.264 that vanished from the Render Queue, and bloated or muddy files.
|
||||
|
||||
## The two export engines
|
||||
|
||||
After Effects has two output paths. Choosing the wrong one is the root of most export confusion.
|
||||
|
||||
- **Render Queue (RQ)** — `Composition > Add to Render Queue`. Best for mastering/intermediate formats: ProRes, image sequences, lossless, alpha. Renders inside AE.
|
||||
- **Adobe Media Encoder (AME)** — `Composition > Add to Adobe Media Encoder Queue` (or the RQ "Queue in AME" button). Best for delivery/compressed formats: H.264 MP4, H.265, social presets. Renders in a separate app so AE stays usable.
|
||||
|
||||
**Why H.264 disappeared from the Render Queue:** Since AE CC 2014/2015, Adobe removed the native H.264, H.264 Blu-ray, and MPEG-2 output modules from the RQ. To make an MP4, use AME — not the RQ. There is no setting to restore it in the RQ; that workflow is intentional.
|
||||
|
||||
## Decision tree
|
||||
|
||||
```
|
||||
Need transparency (alpha) in the final file?
|
||||
├─ YES, for an editor/compositor (NLE) → ProRes 4444 (.mov), RGB+Alpha, RQ
|
||||
│ └─ smaller, still alpha? → PNG sequence or TIFF sequence (RGB+Alpha)
|
||||
├─ YES, for the web/browser → dual export:
|
||||
│ • HEVC with alpha (.mov, Safari/iOS)
|
||||
│ • VP9/WebM with alpha (Chrome/Firefox/Edge)
|
||||
└─ NO alpha needed
|
||||
├─ Final delivery / upload (YouTube, social, client review) → H.264 MP4 via AME
|
||||
├─ Editorial / intermediate / grading → ProRes 422 HQ or 422 (.mov), RQ
|
||||
├─ Mastering / archival, highest fidelity → ProRes 4444 or lossless / image seq
|
||||
└─ Tiny looping clip, limited colors → GIF (via AME / Photoshop), or WebM
|
||||
```
|
||||
|
||||
## Transparency — the black background problem
|
||||
|
||||
H.264/MP4 **cannot store an alpha channel**. Exporting a comp with transparent areas to H.264 always fills them with black (or white). This is the single most common export complaint and it is not a bug.
|
||||
|
||||
To actually keep transparency:
|
||||
|
||||
1. Use a format that supports alpha: **ProRes 4444**, **PNG/TIFF sequence**, **HEVC-with-alpha**, or **VP9/WebM-with-alpha**.
|
||||
2. In the RQ Output Module, set **Channels: RGB + Alpha** and **Depth** to Millions of Colors+ (the "+" is the alpha).
|
||||
3. Choose a matte type: **Premultiplied (Matted with color)** for most cases; **Straight (Unmatted)** if the compositor explicitly wants straight alpha. Straight alpha avoids dark/light fringing on edges when re-composited, but only some apps import it cleanly.
|
||||
|
||||
For web transparent video, no single file plays everywhere. Ship two and use the HTML `<video>` element with multiple `<source>` tags: HEVC-alpha for Safari/iOS, VP9/WebM-alpha for Chromium/Firefox.
|
||||
|
||||
## Bitrate guidance (H.264 via AME)
|
||||
|
||||
Set in AME Export Settings → Video → Bitrate Settings.
|
||||
|
||||
- **Encoding:** VBR, 2 pass (better quality per byte than CBR or 1-pass).
|
||||
- **1080p30:** Target ~10–16 Mbps, Max ~20 Mbps.
|
||||
- **1080p60:** Target ~16–24 Mbps.
|
||||
- **4K30:** Target ~35–45 Mbps, Max ~55 Mbps.
|
||||
- **4K60:** Target ~50–68 Mbps.
|
||||
- Higher target/max for high-motion or grainy footage; lower for flat graphics/motion design.
|
||||
- For uploading to YouTube/Vimeo, lean to the high end — the platform re-encodes, so feed it a clean, high-bitrate master rather than over-compressing twice.
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Target | Engine | Codec / Container | Channels | Bitrate / Notes |
|
||||
|---|---|---|---|---|
|
||||
| YouTube / Vimeo upload | AME | H.264 .mp4 | RGB | VBR 2-pass, 1080p ~16 Mbps / 4K ~45 Mbps |
|
||||
| Instagram / TikTok / Reels | AME | H.264 .mp4 | RGB | 1080×1920 or 1080×1080, ~10–14 Mbps |
|
||||
| Client review | AME | H.264 .mp4 | RGB | VBR 2-pass ~10–16 Mbps |
|
||||
| Editorial / hand to editor | RQ | ProRes 422 HQ .mov | RGB | Mezzanine; large but edit-friendly |
|
||||
| Mastering / grading | RQ | ProRes 4444 .mov | RGB(+A) | Highest fidelity, optional alpha |
|
||||
| Transparency → NLE | RQ | ProRes 4444 .mov | RGB+Alpha | Millions+; premultiplied usually |
|
||||
| Transparency → frames | RQ | PNG or TIFF sequence | RGB+Alpha | Lossless per-frame, robust |
|
||||
| Transparent web (Safari) | AME/RQ | HEVC w/ alpha .mov | RGB+Alpha | Pair with VP9 below |
|
||||
| Transparent web (Chrome) | AME/3rd-party | VP9 / WebM w/ alpha | RGB+Alpha | `yuva420p` alpha mode |
|
||||
| Looping clip, few colors | AME/PS | GIF or WebM | — | GIF = large/limited; prefer WebM |
|
||||
| Archival lossless | RQ | Lossless / PNG seq | RGB(+A) | Huge files |
|
||||
|
||||
## Gotchas
|
||||
|
||||
- **H.264 in the RQ is gone** — use AME. Don't search for a missing module.
|
||||
- **Black background = wrong codec.** Switching matte/straight settings won't help if the codec has no alpha. Change the codec first.
|
||||
- **"Millions of Colors" has no alpha; "Millions of Colors+" does.** The `+` is the alpha channel in RQ Depth.
|
||||
- **ProRes on Windows** is fully supported for export in modern AE; no extra codec install needed.
|
||||
- **GIF is rarely the right answer** — large files, dithered color, no real alpha softness. Prefer WebM/VP9 or short H.264 loops.
|
||||
- **Don't double-compress.** Master to ProRes/lossless, then make one H.264 from that — not H.264 from H.264.
|
||||
- **AME "Match Source"** presets can silently downscale or drop the bitrate; verify resolution and bitrate before queuing.
|
||||
- **Render multiple outputs efficiently:** in the RQ, one render item can have several Output Modules (e.g. ProRes master + PNG seq) from a single render pass.
|
||||
|
||||
---
|
||||
|
||||
# Codec & Export Matrix — Detailed Reference
|
||||
|
||||
## ProRes family — when to use which
|
||||
|
||||
| Variant | Use for | Alpha | Relative size |
|
||||
|---|---|---|---|
|
||||
| ProRes 422 Proxy | Offline edit, rough cuts | No | Smallest |
|
||||
| ProRes 422 LT | Lightweight editorial | No | Small |
|
||||
| ProRes 422 | General editorial | No | Medium |
|
||||
| ProRes 422 HQ | High-quality editorial / delivery to editor | No | Large |
|
||||
| ProRes 4444 | Mastering, grading, **transparency** | Yes | Larger |
|
||||
| ProRes 4444 XQ | Highest fidelity, HDR, max-quality alpha | Yes | Largest |
|
||||
|
||||
Rule of thumb: hand an editor **422 HQ**; master/keep alpha with **4444**.
|
||||
|
||||
## Full target → settings matrix
|
||||
|
||||
| Delivery target | Engine | Container | Codec | Resolution | Channels / Depth | Bitrate / Quality | Audio |
|
||||
|---|---|---|---|---|---|---|---|
|
||||
| YouTube 1080p | AME | .mp4 | H.264 | 1920×1080 | RGB, 8-bit | VBR 2-pass, Tgt 16 / Max 20 Mbps | AAC 320 kbps 48 kHz |
|
||||
| YouTube 4K | AME | .mp4 | H.264 | 3840×2160 | RGB, 8-bit | VBR 2-pass, Tgt 45 / Max 55 Mbps | AAC 384 kbps |
|
||||
| Vimeo 1080p | AME | .mp4 | H.264 | 1920×1080 | RGB | VBR 2-pass, Tgt 20 Mbps | AAC 320 kbps |
|
||||
| Instagram Feed | AME | .mp4 | H.264 | 1080×1080 or 1080×1350 | RGB | ~10–12 Mbps | AAC 128–256 kbps |
|
||||
| Instagram Reels / TikTok | AME | .mp4 | H.264 | 1080×1920 | RGB | ~12–14 Mbps | AAC 256 kbps |
|
||||
| X / Twitter | AME | .mp4 | H.264 | 1280×720 or 1080p | RGB | ~6–10 Mbps | AAC |
|
||||
| Broadcast delivery | RQ | .mov | ProRes 422 HQ | as spec (often 1080i/p) | RGB, 10-bit | — | PCM 48 kHz |
|
||||
| Editorial handoff | RQ | .mov | ProRes 422 HQ | comp size | RGB | — | PCM |
|
||||
| Color grading master | RQ | .mov | ProRes 4444 | comp size | RGB, 12-bit | — | PCM |
|
||||
| VFX plate / alpha to NLE | RQ | .mov | ProRes 4444 | comp size | RGB+Alpha, Millions+ | premultiplied | — |
|
||||
| Compositing frames | RQ | seq | PNG or 16-bit TIFF | comp size | RGB+Alpha | lossless | — |
|
||||
| EXR for VFX pipeline | RQ | seq | OpenEXR | comp size | RGB+Alpha, 32-bit float | linear/ACES | — |
|
||||
| Transparent web (Apple) | AME or RQ | .mov | HEVC w/ alpha | comp size | RGB+Alpha | quality ~ 60–80% | optional |
|
||||
| Transparent web (Chrome) | 3rd-party/ffmpeg | .webm | VP9 (yuva420p) | comp size | RGB+Alpha | CRF ~30 | optional |
|
||||
| Animated icon / loop | AME or PS | .gif / .webm | GIF / VP9 | small | indexed / RGBA | minimize frames | — |
|
||||
|
||||
## Transparency — step by step
|
||||
|
||||
### ProRes 4444 with alpha (Render Queue)
|
||||
1. `Composition > Add to Render Queue`.
|
||||
2. Click the blue **Output Module** text.
|
||||
3. Format: **QuickTime**. Click **Format Options** → Codec: **Apple ProRes 4444**.
|
||||
4. **Channels: RGB + Alpha**. **Depth: Millions of Colors+**.
|
||||
5. **Color: Premultiplied (Matted)** for most uses; choose **Straight (Unmatted)** only if the downstream app asks for straight alpha.
|
||||
6. Set Output To path, then **Render**.
|
||||
|
||||
### PNG / TIFF sequence with alpha (Render Queue)
|
||||
1. Add to RQ → Output Module.
|
||||
2. Format: **PNG Sequence** (or **TIFF Sequence**).
|
||||
3. **Channels: RGB + Alpha**, Depth **Millions of Colors+** (PNG) or 16-bit (TIFF).
|
||||
4. Output To a **dedicated empty folder** — sequences write one file per frame.
|
||||
|
||||
### HEVC with alpha (Apple/Safari)
|
||||
- In AME, choose **H.265 (HEVC)** and enable an alpha/transparency option where available, or use the QuickTime HEVC export with RGB+Alpha. Verify playback in Safari; Chrome will not show its alpha.
|
||||
|
||||
### VP9 / WebM with alpha (Chrome)
|
||||
- AME does not reliably produce alpha WebM. Render a ProRes 4444 or PNG sequence from AE, then encode with ffmpeg:
|
||||
- `ffmpeg -i master.mov -c:v libvpx-vp9 -pix_fmt yuva420p -b:v 0 -crf 30 out.webm`
|
||||
- `yuva420p` is the alpha-carrying pixel format; without it WebM has no transparency.
|
||||
|
||||
### HTML for cross-browser transparent video
|
||||
```html
|
||||
<video autoplay loop muted playsinline>
|
||||
<source src="clip.mov" type="video/quicktime"> <!-- HEVC alpha, Safari/iOS -->
|
||||
<source src="clip.webm" type="video/webm"> <!-- VP9 alpha, Chrome/Firefox -->
|
||||
</video>
|
||||
```
|
||||
|
||||
## Alpha matte types explained
|
||||
- **Straight (Unmatted):** color stored independent of alpha; cleanest edges when re-composited, but fewer apps import correctly.
|
||||
- **Premultiplied (Matted with color):** color blended with a matte color (usually black). Most compatible; can show fringing if the wrong matte color is assumed downstream.
|
||||
- When in doubt for NLE handoff, **premultiplied with black** is the safe default.
|
||||
|
||||
## AME preset cautions
|
||||
- **Match Source – High bitrate** can downscale to the source clip rather than the comp; confirm output resolution.
|
||||
- Always check **VBR 2-pass** is selected for final masters; presets sometimes default to 1-pass.
|
||||
- Use **Render at Maximum Depth** and **Use Maximum Render Quality** for scaling-heavy comps (slower but cleaner).
|
||||
- For social verticals, set the resolution explicitly — don't trust a 16:9 preset.
|
||||
|
||||
## Bitrate cheat sheet (H.264 VBR 2-pass)
|
||||
| Resolution / fps | Target Mbps | Max Mbps |
|
||||
|---|---|---|
|
||||
| 720p30 | 5–8 | 10 |
|
||||
| 1080p30 | 10–16 | 20 |
|
||||
| 1080p60 | 16–24 | 30 |
|
||||
| 4K30 | 35–45 | 55 |
|
||||
| 4K60 | 50–68 | 80 |
|
||||
@@ -0,0 +1,184 @@
|
||||
# After Effects Expression Library, Rigging & Export
|
||||
|
||||
A deeper cookbook. All expressions are JavaScript (the modern AE expression engine). They evaluate on the property they are pasted on, where `value`, `time`, `index`, `thisLayer`, `thisComp` are in scope.
|
||||
|
||||
## Inertial bounce (parameterized)
|
||||
|
||||
```js
|
||||
// Overshoot/settle after the most recent keyframe. Tune amp/freq/decay.
|
||||
amp = 0.10; // 0.05 = subtle, 0.2 = bouncy
|
||||
freq = 3.0; // oscillations per second
|
||||
decay = 5.0; // higher = settles faster
|
||||
|
||||
n = 0;
|
||||
if (numKeys > 0) {
|
||||
n = nearestKey(time).index;
|
||||
if (key(n).time > time) n--;
|
||||
}
|
||||
if (n === 0) {
|
||||
value;
|
||||
} else {
|
||||
t = time - key(n).time;
|
||||
v = velocityAtTime(key(n).time - thisComp.frameDuration / 10);
|
||||
value + v * amp * Math.sin(freq * t * 2 * Math.PI) / Math.exp(decay * t);
|
||||
}
|
||||
```
|
||||
|
||||
Inertial bounce works on multi-dimensional properties because `velocityAtTime` returns a vector; the arithmetic is per-component.
|
||||
|
||||
## Loop variants
|
||||
|
||||
```js
|
||||
loopOut("cycle"); // A B C A B C ...
|
||||
loopOut("pingpong"); // A B C B A B C ...
|
||||
loopOut("offset"); // continues accumulating the delta each cycle
|
||||
loopOut("continue"); // linear extension of the final velocity (no loop)
|
||||
loopOut("cycle", 3); // loop only the last 3 keyframes
|
||||
loopIn("cycle", 2); // loop the first 2 keyframes, backwards in time
|
||||
loopOutDuration("cycle", 2); // loop the last 2 SECONDS of keyframes
|
||||
```
|
||||
|
||||
Use `offset` for endless conveyor/scroll motion; `continue` for inertia past the last keyframe without oscillation.
|
||||
|
||||
## Wiggle — advanced
|
||||
|
||||
```js
|
||||
// Full signature: wiggle(freq, amp, octaves, amp_mult, time)
|
||||
wiggle(2, 30, 3, 0.5); // 3 octaves adds finer detail
|
||||
|
||||
// Per-axis wiggle (different amplitude X vs Y)
|
||||
x = wiggle(2, 50)[0];
|
||||
y = wiggle(2, 10)[1];
|
||||
[x, y];
|
||||
|
||||
// Wiggle only rotation while keeping keyframed position
|
||||
wiggle(1, 5); // on Rotation
|
||||
|
||||
// Smooth, looping ambient motion without true wiggle (loopable):
|
||||
amp = 20; period = 4; // seconds
|
||||
[Math.sin(time / period * 2 * Math.PI) * amp,
|
||||
Math.cos(time / period * 2 * Math.PI) * amp] + value;
|
||||
```
|
||||
|
||||
`wiggle` is non-looping noise; for seamless loops use periodic sin/cos as above.
|
||||
|
||||
## Stable randomness across a grid
|
||||
|
||||
```js
|
||||
seedRandom(index, true); // timeless=true → constant over time, varies per layer
|
||||
startDelay = random(0, 0.5); // each layer gets a stable random delay
|
||||
opacity = ease(time, startDelay, startDelay + 0.4, 0, 100);
|
||||
```
|
||||
|
||||
`seedRandom(index, false)` (timeless=false) lets `random()` vary over time — use for flicker; `true` for stable offsets.
|
||||
|
||||
## Time remap & valueAtTime echoes
|
||||
|
||||
```js
|
||||
// Trailing/echo: sample the lead layer in the past, staggered by index
|
||||
delay = 0.08 * (index - 1);
|
||||
thisComp.layer("Lead").transform.position.valueAtTime(time - delay);
|
||||
|
||||
// Speed-ramp a time-remapped layer using easing on the source time
|
||||
src = thisComp.layer("Source");
|
||||
linear(time, 0, 2, 0, src.source.duration); // play a 2s window of the source
|
||||
|
||||
// Sample own value in the past for motion smear / trail
|
||||
valueAtTime(time - 0.1);
|
||||
```
|
||||
|
||||
## Rigging walkthrough
|
||||
|
||||
### Control types and how to read them
|
||||
|
||||
```js
|
||||
ctrl = thisComp.layer("Controls");
|
||||
ctrl.effect("Intensity")("Slider"); // number
|
||||
ctrl.effect("Visible")("Checkbox"); // 1 or 0
|
||||
ctrl.effect("Brand")("Color"); // [r,g,b,a] 0..1
|
||||
ctrl.effect("Spin")("Angle"); // degrees
|
||||
ctrl.effect("Anchor")("Point"); // [x,y]
|
||||
```
|
||||
|
||||
### One slider, staggered reveal across N layers
|
||||
|
||||
```js
|
||||
// On each layer's Opacity:
|
||||
master = thisComp.layer("Controls").effect("Reveal")("Slider"); // 0..100 driver
|
||||
per = 6; // frames of stagger per layer
|
||||
shift = framesToTime(per * (index - 1));
|
||||
ease(master, 0, 100, 0, 100) * (time > shift ? 1 : 0);
|
||||
```
|
||||
|
||||
### Coordinate-space conversions
|
||||
|
||||
```js
|
||||
// Place this layer where another parented layer appears in comp space
|
||||
target = thisComp.layer("Marker");
|
||||
fromComp(target.toComp(target.anchorPoint));
|
||||
|
||||
// World position of a point on this layer
|
||||
toComp([0, 0]);
|
||||
```
|
||||
|
||||
`toComp` converts layer space → composition space; `fromComp` does the inverse. Use them whenever layers are parented or live in different (nested) comps.
|
||||
|
||||
### Linking many layers to a master null
|
||||
|
||||
1. Select all children → set Parent to the null.
|
||||
2. Animate only the null; children inherit transform.
|
||||
3. For non-transform links (effects, opacity), pick-whip to the control layer's effect.
|
||||
|
||||
## Essential Graphics / MOGRT export steps
|
||||
|
||||
1. **Window → Essential Graphics**, choose the master comp.
|
||||
2. Drag exposed properties (Slider, Checkbox, Color, Text Source, Media replacement) into the panel.
|
||||
3. Add **Group** headers, rename controls, fill comment fields for editors.
|
||||
4. Use **Drop-Zone / Media Replacement** for swappable footage/logos.
|
||||
5. Click **Export Motion Graphics Template**, choose destination (Local Templates Folder / Library / file). Toggle "Make fonts available" and check "Raster/vector" compatibility for Premiere.
|
||||
6. Test in Premiere Pro's Essential Graphics → Browse panel.
|
||||
|
||||
Keep all motion driven by expressions referencing the exposed controls so editors only touch the exposed params; hide internal layers with shy.
|
||||
|
||||
## Lottie / Bodymovin export
|
||||
|
||||
Supported reliably:
|
||||
- Shape layers, solids, transforms (position/scale/rotation/opacity/anchor).
|
||||
- Masks and track mattes (alpha), parenting, basic gradients.
|
||||
- Trim Paths, simple repeaters.
|
||||
- Expressions: enable in Bodymovin settings, but verify on the target renderer — `lottie-web`, lottie-ios, and lottie-android differ in expression and effect support.
|
||||
|
||||
Not supported / risky (bake or avoid):
|
||||
- Blend modes, most native Effects, layer blurs, 3D layers/cameras, time remapping, adjustment-layer effects.
|
||||
|
||||
Workflow:
|
||||
1. Install Bodymovin extension; **Window → Extensions → Bodymovin**.
|
||||
2. Select comp(s), set destination, choose Standard/Demo.
|
||||
3. Settings: enable "Glyphs"/"Hidden", "Expressions" only if needed; "Assets" → export images or use base64.
|
||||
4. Render `data.json`. Preview in a Lottie player; check file size and that masks/mattes survived.
|
||||
5. For web, load with `lottie-web`:
|
||||
|
||||
```js
|
||||
import lottie from 'lottie-web';
|
||||
lottie.loadAnimation({
|
||||
container: document.getElementById('lottie'),
|
||||
renderer: 'svg', // 'canvas' for heavy fills, better perf
|
||||
loop: true, autoplay: true,
|
||||
path: '/data.json',
|
||||
});
|
||||
```
|
||||
|
||||
## Media Encoder presets
|
||||
|
||||
| Use | Codec / format | Settings |
|
||||
|---|---|---|
|
||||
| Web/social delivery | H.264 (MP4) | VBR 2-pass, 1080p ~12–16 Mbps; 4K ~35–50 Mbps |
|
||||
| Master/intermediate | ProRes 422 HQ (MOV) | full range, matched fps |
|
||||
| With alpha | ProRes 4444 (MOV) | enable alpha channel |
|
||||
| Lossless interchange | PNG sequence or ProRes 4444 | per-frame, alpha |
|
||||
| GIF | via Media Encoder/Photoshop | reduce palette, dither, cap dimensions |
|
||||
|
||||
Tips:
|
||||
- Match the comp frame rate; avoid frame-blending unless intentionally retiming.
|
||||
- H.264 cannot store alpha — composite onto a background or use ProRes 4444 / PNG.
|
||||
- For crisp text, render at the delivery resolution (don't upscale); set quality to maximum and use 2-pass VBR.
|
||||
@@ -0,0 +1,134 @@
|
||||
# Expressions, Rigging & MOGRT
|
||||
|
||||
After Effects is GUI software, so this delivers knowledge: ready-to-paste JavaScript expressions, rigging setups that make motion controllable, and template/export pipelines. Expressions read on the property they are pasted on; `value` is that property's keyframed/static value, `time` is the composition time in seconds, and `index`, `thisLayer`, `thisComp` are in scope.
|
||||
|
||||
For a deeper cookbook (tuned inertial bounce, every loop variant, advanced per-axis wiggle, `valueAtTime` echoes, `seedRandom` grids, full rigging walkthrough, Essential Graphics → MOGRT steps, Lottie/Bodymovin + Media Encoder presets), see `expression-library.md`.
|
||||
|
||||
## Core expressions
|
||||
|
||||
### Inertial bounce (overshoot after the last keyframe)
|
||||
|
||||
Paste on a Position, Scale, or Rotation property. It continues the velocity at the final keyframe into a damped sine — the classic "settle" motion.
|
||||
|
||||
```js
|
||||
amp = 0.08; // overshoot amount
|
||||
freq = 2.5; // oscillations per second
|
||||
decay = 6.0; // how fast it settles
|
||||
n = 0;
|
||||
if (numKeys > 0) {
|
||||
n = nearestKey(time).index;
|
||||
if (key(n).time > time) n--;
|
||||
}
|
||||
if (n > 0) {
|
||||
t = time - key(n).time;
|
||||
v = velocityAtTime(key(n).time - 0.001);
|
||||
value + v * (amp * Math.sin(freq * t * 2 * Math.PI) / Math.exp(decay * t));
|
||||
} else {
|
||||
value;
|
||||
}
|
||||
```
|
||||
|
||||
### Looping
|
||||
|
||||
```js
|
||||
loopOut("cycle"); // repeat keyframed range forever
|
||||
loopOut("pingpong"); // bounce back and forth
|
||||
loopOut("offset"); // continue, accumulating (e.g. continuous travel)
|
||||
loopOut("continue"); // extend final velocity linearly
|
||||
loopOut("cycle", 2); // loop only the last 2 keyframes
|
||||
loopIn("cycle"); // mirror, before the first keyframe
|
||||
```
|
||||
|
||||
### Wiggle and stable randomness
|
||||
|
||||
```js
|
||||
wiggle(2, 30); // 2 wiggles/sec, amplitude 30 (units match the property)
|
||||
seedRandom(index, true); // unique-but-stable per layer; timeless=true
|
||||
random(0, 100); // stable after seedRandom
|
||||
wiggle(freq, amp, octaves=1, amp_mult=0.5, t=time); // full signature
|
||||
```
|
||||
|
||||
`seedRandom(index, true)` before `random()` or `wiggle()` makes each layer differ but stay constant across renders — essential for grids of layers that must not flicker.
|
||||
|
||||
### Time remap & value-at-time
|
||||
|
||||
```js
|
||||
// Sample another property/layer at a shifted time (echo/trailing motion)
|
||||
delay = 0.2 * index;
|
||||
thisComp.layer("Lead").transform.position.valueAtTime(time - delay);
|
||||
|
||||
// Speed up a precomp via time remap
|
||||
linear(time, 0, thisComp.duration, 0, 6); // 6s source mapped over comp duration
|
||||
```
|
||||
|
||||
## Rigging
|
||||
|
||||
Make motion controllable instead of hand-keyed:
|
||||
|
||||
1. Add a null/control layer (often a "Controls" layer). Add **Slider Control**, **Checkbox Control**, **Color Control**, **Angle Control**, or **Point Control** effects to it.
|
||||
2. Reference them with the pick-whip or by path:
|
||||
|
||||
```js
|
||||
ctrl = thisComp.layer("Controls");
|
||||
amt = ctrl.effect("Intensity")("Slider"); // Slider Control → ADBE Slider Control-0001
|
||||
on = ctrl.effect("Enable")("Checkbox"); // Checkbox: 1 or 0
|
||||
col = ctrl.effect("Tint")("Color");
|
||||
|
||||
if (on == 1) value * amt else value;
|
||||
```
|
||||
|
||||
3. Parent transforms to a null so one move drives many layers. Use `toComp`/`fromComp` to convert coordinate spaces when layers are parented or in different comps.
|
||||
|
||||
```js
|
||||
// Stagger N layers off one master slider
|
||||
master = thisComp.layer("Controls").effect("Reveal")("Slider");
|
||||
delay = 4 * index; // 4 frames per layer
|
||||
master.valueAtTime(time - framesToTime(delay));
|
||||
```
|
||||
|
||||
## Essential Graphics / MOGRT
|
||||
|
||||
1. Open **Window → Essential Graphics**, pick the comp as the Master.
|
||||
2. Drag the Slider/Checkbox/Color/Text source properties into the panel to expose them.
|
||||
3. Group, rename, and set comment fields for editors.
|
||||
4. **Export Motion Graphics Template** → a `.mogrt` usable in Premiere Pro / AE. Keep driving expressions referencing the exposed controls so editors change only the exposed params.
|
||||
|
||||
Set "Allow fonts to substitute" off and embed/package fonts where possible so the template renders consistently elsewhere.
|
||||
|
||||
## Export from expressions/template work
|
||||
|
||||
### To Lottie (Bodymovin / `@lottiefiles/bodymovin`)
|
||||
|
||||
- Stick to Lottie-supported features: shape layers, transforms, masks, track mattes, basic expressions converted to keyframes.
|
||||
- Avoid blur-heavy effects, blend modes, 3D layers, and most native effects — they don't translate.
|
||||
- Bake expressions to keyframes if the renderer is strict, or use the `expressions` flag in Bodymovin and test on the target player (web/iOS/Android renderers differ).
|
||||
- Run **Bodymovin → Render** to produce `data.json` (+ optional images). Verify in a Lottie preview before handoff.
|
||||
|
||||
### To video (Media Encoder)
|
||||
|
||||
- **File → Export → Add to Adobe Media Encoder Queue** (or render queue for AE-native).
|
||||
- Delivery: **H.264** (MP4), bitrate ~10–20 Mbps for 1080p, VBR 2-pass.
|
||||
- Master/intermediate: **ProRes 422 HQ** (or ProRes 4444 when alpha is needed).
|
||||
- Transparency: ProRes 4444 or PNG/QuickTime with an alpha channel; H.264 cannot carry alpha. (See `export-recipes.md` for the full codec/alpha/bitrate matrix.)
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Goal | Expression |
|
||||
|---|---|
|
||||
| Loop keyframes | `loopOut("cycle")` |
|
||||
| Continuous travel | `loopOut("offset")` |
|
||||
| Subtle life | `wiggle(2, 20)` |
|
||||
| Stable per-layer random | `seedRandom(index,true); random(...)` |
|
||||
| Settle/overshoot | inertial bounce (above) |
|
||||
| Echo another layer | `valueAtTime(time - delay)` |
|
||||
| Read a slider | `effect("Name")("Slider")` |
|
||||
| Frames ↔ seconds | `framesToTime(n)` / `timeToFrames(t)` |
|
||||
|
||||
## Gotchas
|
||||
|
||||
- An expression overrides keyframes on that property unless it incorporates `value`; start most expressions from `value`.
|
||||
- `index` is the layer's stacking number — it changes if layers are reordered; prefer named lookups for stable rigs.
|
||||
- Expression errors disable the expression and show a yellow banner; check the property's path names (effect names must match exactly, including duplicates' suffixes).
|
||||
- Pre-compose for reuse; name layers; use shy/lock and guide layers for safe areas before handoff.
|
||||
</content>
|
||||
</invoke>
|
||||
@@ -0,0 +1,138 @@
|
||||
# Performance Triage — Slow, Laggy, Crashing, Frozen
|
||||
|
||||
Diagnose slow previews, slow renders, freezes, and crashes in a fixed order, from cheapest fix to most invasive. Work top-down and stop when the symptom clears — most issues are solved in the first few steps.
|
||||
|
||||
Most fixes live in Preferences. Open `Edit > Preferences` (Windows) or `After Effects > Settings/Preferences` (macOS).
|
||||
|
||||
## First: identify the symptom
|
||||
|
||||
- **Slow / non-real-time previews (RAM preview)** → Flow A
|
||||
- **Slow renders (final output)** → Flow B
|
||||
- **Crashes / freezes / hangs** → Flow C
|
||||
|
||||
## Flow A — Slow previews (do in order)
|
||||
|
||||
1. **Set Resolution to Auto/Half/Third.** In the Composition viewer, change Resolution from Full to **Half** or **Third**. Adaptive/Draft resolution gives the biggest instant win.
|
||||
2. **Enable Multi-Frame Rendering (MFR).** `Preferences > Memory & Performance > Enable Multi-Frame Rendering`. Uses all CPU cores for preview and render (AE 2022+). Confirm "CPUs reserved for other applications" isn't set too high.
|
||||
3. **Check RAM allocation.** `Preferences > Memory & Performance`. Leave ~6 GB for the OS; give the rest to AE. Too little RAM = a tiny green preview bar.
|
||||
4. **Set Fast Previews to Adaptive Resolution.** Viewer's lightning-bolt icon → **Adaptive Resolution** (degrades resolution only while interacting).
|
||||
5. **Purge the cache.** `Edit > Purge > All Memory & Disk Cache`. Stale/full cache stalls previews.
|
||||
6. **Verify the Disk Cache is on a fast drive.** `Preferences > Media & Disk Cache` → Enable Disk Cache, point it at a fast **SSD/NVMe**, give it 50–100+ GB. Never on a slow external HDD.
|
||||
7. **Solo the offending layer/effect.** Toggle effects off (the fx switch) or solo layers to find the bottleneck. Heavy culprits: blurs, glows, Mocha/tracking, particles, expressions, large 3D, and many third-party plugins.
|
||||
8. **Enable GPU / hardware-accelerated decoding.** `Preferences > Video Preview` (Mercury GPU) and `Preferences > Import > Hardware Accelerate H.264/HEVC decoding`. Helps long-GOP source footage.
|
||||
9. **Pre-render or proxy heavy comps.** Pre-render a finished/slow precomp to ProRes and set it as a **proxy** (`File > Create Proxy`, or right-click footage → Set Proxy). Frees compute for the part being worked on.
|
||||
10. **Trim the work area** to just the section in progress so RAM preview fills faster.
|
||||
|
||||
## Flow B — Slow renders (do in order)
|
||||
|
||||
1. **Confirm Multi-Frame Rendering is on** (`Memory & Performance`) — the single largest render speedup on multi-core CPUs.
|
||||
2. **Render with the Render Queue for masters, AME for delivery** — and don't run heavy work in AE while AME renders.
|
||||
3. **Enable hardware-accelerated decoding** for H.264/HEVC sources (`Preferences > Import`).
|
||||
4. **Pre-render slow precomps** once and reuse them, instead of recomputing every frame. Heavy effect stacks, time-remapping, and 3rd-party renderers benefit most.
|
||||
5. **Replace effects with cheaper equivalents** where possible (e.g., Fast Box Blur over Gaussian; native over plugin).
|
||||
6. **Convert long-GOP source to an intermediate** (ProRes/DNxHR). Decoding H.264 every frame is slow; ProRes is cheap to read.
|
||||
7. **Check the renderer:** `Composition > Composition Settings > 3D Renderer`. Classic 3D is fastest; Cinema 4D/Ray-traced renderers are far slower.
|
||||
8. **Disk Cache on NVMe** so repeated frames aren't recomputed.
|
||||
9. **Close other apps**, especially other Adobe apps and browsers, to free RAM/CPU for MFR.
|
||||
|
||||
## Flow C — Crashes / freezes / hangs (do in order)
|
||||
|
||||
1. **Purge cache** (`Edit > Purge > All`). A full/corrupt cache causes hangs.
|
||||
2. **Clear the Disk Cache folder** in `Preferences > Media & Disk Cache > Empty Disk Cache`. Corrupt cache files cause crash-on-open.
|
||||
3. **Update GPU drivers** (NVIDIA Studio driver / AMD). Outdated/Game-Ready drivers cause Mercury GPU crashes. As a test, switch Project renderer GPU→CPU (`File > Project Settings > Video Rendering and Effects: Mercury Software Only`).
|
||||
4. **Reset Preferences:** hold `Ctrl+Alt+Shift` (Win) / `Cmd+Opt+Shift` (Mac) while launching AE. Corrupt prefs are a top crash cause.
|
||||
5. **Disable third-party plugins** by temporarily moving them out of the plug-ins folder, then reintroduce to find the offender. Old plugins are a leading crash/freeze source after AE updates.
|
||||
6. **Check Disk Cache and Output drives for free space.** A full scratch/cache disk causes freezes mid-render.
|
||||
7. **Lower or disable MFR** if crashes appear only during multi-frame render — some plugins are not MFR-safe; reserve more CPUs or disable MFR for those projects.
|
||||
8. **Roll back the AE version** if crashes started right after an update. Install a previous version via Creative Cloud (Other Versions) — many crash waves are version-specific regressions.
|
||||
9. **Look for a single corrupt asset** — import comps one at a time, or replace footage, to isolate a bad file.
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Symptom | First three moves |
|
||||
|---|---|
|
||||
| Choppy RAM preview | Half resolution → enable MFR → purge cache |
|
||||
| Tiny green preview bar | Raise RAM allocation → SSD disk cache → trim work area |
|
||||
| Slow final render | Enable MFR → HW decode H.264 → pre-render precomps |
|
||||
| Crash on launch | Reset prefs (modifier keys) → empty disk cache → update GPU driver |
|
||||
| Crash mid-render | Free disk space → disable/reserve CPUs for MFR → disable 3rd-party plugins |
|
||||
| Started after update | Roll back AE version → update plugins → reset prefs |
|
||||
| Laggy with footage | HW-accelerated decode → transcode to ProRes → proxy heavy comps |
|
||||
|
||||
## Gotchas
|
||||
|
||||
- **MFR needs RAM per core.** With many cores but little RAM, MFR can underperform or stall — aim for several GB per active core.
|
||||
- **Game-Ready drivers crash AE.** Prefer the **Studio** driver branch on NVIDIA.
|
||||
- **Disk Cache on an external HDD is worse than none** — it must be SSD/NVMe to help.
|
||||
- **Plugins lag behind AE updates.** After a major AE upgrade, a sudden crash wave usually means a plugin needs updating or is not MFR-safe.
|
||||
- **Purging clears the green bar.** Previews must rebuild afterward; that initial slowness is expected, not a new problem.
|
||||
- **CPU/GPU effects are not equal.** A few GPU-accelerated effects are fast; mixing many CPU-only effects forces round-trips that kill preview speed.
|
||||
- **Don't render in AE and AME simultaneously** on the same machine — they fight for cores and RAM.
|
||||
|
||||
---
|
||||
|
||||
# AE Performance — Detailed Diagnostics Reference
|
||||
|
||||
## Preference locations and recommended values
|
||||
|
||||
| Setting | Path | Recommended |
|
||||
|---|---|---|
|
||||
| RAM for AE | Preferences > Memory & Performance | Total RAM minus ~6 GB for OS |
|
||||
| Multi-Frame Rendering | Memory & Performance > Enable MFR | On; reserve 0–2 CPUs for other apps |
|
||||
| Disk Cache | Media & Disk Cache > Enable Disk Cache | On, SSD/NVMe, 50–150 GB |
|
||||
| Conformed media cache | Media & Disk Cache | On SSD, separate from project if possible |
|
||||
| HW decode (H.264/HEVC) | Import > Hardware Accelerate Decoding | On |
|
||||
| Mercury GPU | Project Settings > Video Rendering and Effects | GPU (CPU as crash test) |
|
||||
| Fast Previews | Viewer lightning icon | Adaptive Resolution |
|
||||
|
||||
## Multi-Frame Rendering (MFR) tuning
|
||||
- Introduced in AE 2022 (22.x); renders multiple frames in parallel across CPU cores for both preview and final render.
|
||||
- **RAM is the limiter.** Each rendering frame needs memory; with too little RAM, AE quietly reduces the cores it uses. Target several GB of free RAM per logical core you want active.
|
||||
- "CPUs reserved for other applications" trades render speed for system responsiveness. Set 0–2 on a dedicated machine.
|
||||
- Some legacy/3rd-party effects are **not MFR-safe** and can crash or render incorrectly under MFR. If a specific project crashes only with MFR on, disable MFR for that project or update the plugin.
|
||||
|
||||
## GPU and drivers
|
||||
- Use the **NVIDIA Studio Driver** (not Game Ready) for stability with Adobe apps; keep AMD drivers current too.
|
||||
- Mercury GPU acceleration speeds some effects and the renderer. If AE crashes on launch or while playing back, switch `Project Settings > Video Rendering and Effects` to **Mercury Software Only** to confirm GPU/driver fault, then update drivers.
|
||||
- Hardware-accelerated **decoding** (separate from rendering) speeds reading H.264/HEVC source footage; enable in `Preferences > Import`.
|
||||
|
||||
## Disk cache sizing and placement
|
||||
- Put the disk cache on the **fastest local drive** (internal NVMe ideal). Never on a slow USB HDD or a network drive.
|
||||
- 50–150 GB is typical; more helps long projects retain cached frames between sessions.
|
||||
- Symptoms of a bad cache: hangs on open, garbage frames, refusal to play. Fix via `Empty Disk Cache` and `Edit > Purge > All`.
|
||||
- Keep the cache drive from filling up — a full cache/scratch disk causes mid-render freezes.
|
||||
|
||||
## Pre-render and proxy workflow
|
||||
1. Identify a slow, finished precomp (heavy effects, particles, 3D, tracking).
|
||||
2. Add it to the Render Queue and render to **ProRes 422/422 HQ** (or PNG sequence).
|
||||
3. Right-click the footage/comp in the Project panel → **Set Proxy > File**, select the rendered file. A box icon shows the proxy is active.
|
||||
4. Toggle the proxy off for final render if you want to re-render at full quality, or keep it if the pre-render is final.
|
||||
5. Alternatively `File > Create Proxy > Movie/Still` to queue a proxy directly.
|
||||
- Proxies let AE substitute a cheap-to-read file for an expensive comp/footage during editing.
|
||||
|
||||
## Heavy effect culprits (preview/render cost)
|
||||
- Gaussian Blur, Camera Lens Blur, Glow → use Fast Box Blur where acceptable.
|
||||
- Particle systems (Particular, Form), Element 3D, Plexus.
|
||||
- Mocha tracking, Warp Stabilizer (analysis), motion tracking.
|
||||
- Expressions that read many layers/time, especially per-frame heavy ones.
|
||||
- Nested precomps with collapsed transformations and continuous rasterization.
|
||||
- Ray-traced / Cinema 4D 3D renderer vs Classic 3D.
|
||||
|
||||
## Version-specific culprits
|
||||
- After a **major AE update**, a sudden crash/slowdown wave usually traces to a third-party plugin not yet updated for that AE version — update plugins or remove them to test.
|
||||
- If problems began immediately after updating AE, install the **previous version** from Creative Cloud → app → Other Versions, and report the regression.
|
||||
- Reset corrupt preferences by holding `Ctrl+Alt+Shift` (Win) / `Cmd+Opt+Shift` (Mac) during launch; this resolves many post-update oddities.
|
||||
|
||||
## Isolation checklist for crashes
|
||||
1. Reset preferences (modifier-key launch).
|
||||
2. Empty disk cache + Purge All.
|
||||
3. Update GPU driver (Studio branch).
|
||||
4. Move 3rd-party plugins out, relaunch, reintroduce one at a time.
|
||||
5. Switch Mercury to Software Only to rule out GPU.
|
||||
6. Import comps/footage incrementally to find a corrupt asset.
|
||||
7. Disable MFR for the project if crashes are MFR-specific.
|
||||
8. Roll back AE version if the crash wave is version-correlated.
|
||||
|
||||
## When to pre-compose vs pre-render
|
||||
- **Pre-compose** to organize and to apply an effect to a group — does not by itself speed anything up (it can even add overhead).
|
||||
- **Pre-render** (render to a file + use as proxy/footage) when a precomp is finished and slow — this is the actual performance win.
|
||||
@@ -0,0 +1,188 @@
|
||||
# Project Hygiene — Organize, Relink, ExtendScript
|
||||
|
||||
Keep AE projects organized, portable, and automatable: consistent naming, a clear folder/bin structure, label discipline, pre-comp discipline, reliable relinking when files move, and ExtendScript (.jsx) automation for repetitive tasks.
|
||||
|
||||
## Folder (bin) structure
|
||||
|
||||
Create folders in the Project panel (the folder icon at the bottom). A reliable layout:
|
||||
|
||||
```
|
||||
01_Comps/ finished/active compositions
|
||||
02_Precomps/ nested precomps
|
||||
03_Footage/ imported video
|
||||
04_Images/ stills, logos, textures
|
||||
05_Audio/
|
||||
06_Solids/ (AE's auto "Solids" folder — keep it here)
|
||||
07_Assets_3D/
|
||||
99_Old/ deprecated comps/layers, not deleted yet
|
||||
```
|
||||
|
||||
Numeric prefixes force a stable sort order. After importing, immediately drag items into the right bin — `File > Reduce Project` and `File > Consolidate All Footage` help later, but discipline up front prevents chaos.
|
||||
|
||||
## Naming conventions
|
||||
|
||||
- **Comps:** `PROJECT_section_descriptor_v01` (e.g., `BRAND_intro_logoBuild_v03`). Version with zero-padded `v01`, `v02`.
|
||||
- **Precomps:** prefix `PRE_` or suffix `_pre` so they're obvious in lists.
|
||||
- **Layers:** rename solids/nulls to their role (`CTRL_master`, `BG_gradient`, `MATTE_text`). Never ship "Black Solid 12".
|
||||
- **No spaces / special chars** in filenames destined for render farms or cross-OS sharing — use `_` and `-`.
|
||||
- Keep names short but unique; long names truncate in the timeline.
|
||||
|
||||
## Label colors
|
||||
|
||||
Use the Label color (the swatch left of each Project/Timeline item) as a visual system. Pick a convention and apply it consistently, e.g.:
|
||||
- Red = control/null layers
|
||||
- Blue = footage
|
||||
- Green = precomps
|
||||
- Yellow = text
|
||||
- Purple = adjustment/effects layers
|
||||
|
||||
Right-click an item → **Label** to set. Select multiple, then `Edit > Label > Select Label Group` to act on a whole color at once. Customize label names/colors in `Preferences > Labels`.
|
||||
|
||||
## Pre-comp discipline
|
||||
|
||||
- Pre-compose (`Layer > Pre-compose`, `Ctrl/Cmd+Shift+C`) to **group** related layers and to apply one effect to a group — not reflexively for everything.
|
||||
- When pre-composing a single layer, choose **"Leave all attributes in"** vs **"Move all attributes into the new composition"** deliberately: "Move" carries transforms/effects inside; "Leave" keeps them on the outer layer.
|
||||
- Name the precomp immediately and file it in `02_Precomps`.
|
||||
- Avoid deep nesting beyond a few levels — it slows navigation and makes relinking/debugging painful.
|
||||
- Match precomp duration and frame rate to need; an over-long precomp wastes cache.
|
||||
|
||||
## The "moved one file and everything broke" rescue
|
||||
|
||||
AE references footage by file path. Moving, renaming, or sending a project without its assets breaks links (shown as **colored bars / "Missing Footage"** placeholders).
|
||||
|
||||
### Fix options, in order
|
||||
1. **Relink one, fix many:** double-click a missing item (or right-click → **Replace Footage > File**). Point it at the moved file; AE offers to relink *all* missing items from the same folder automatically. This single step usually fixes the whole project.
|
||||
2. **Find Missing Footage:** `File > Dependencies > Find Missing Footage` lists every broken link to resolve.
|
||||
3. **Replace Footage:** right-click footage → **Replace Footage > File** to swap an asset while keeping every use, effect, and keyframe intact.
|
||||
|
||||
### Prevent it: Collect Files
|
||||
Before moving or sharing a project, use `File > Dependencies > Collect Files`. This copies the project plus every used asset into one new folder with relative structure, so links survive on any machine. Always hand off projects as a Collect Files package, not a bare `.aep`.
|
||||
|
||||
Related: `File > Dependencies > Consolidate All Footage` (merge duplicate imports), `Remove Unused Footage`, and `Reduce Project` (strip everything not used by selected comps) to slim a project before collecting.
|
||||
|
||||
## Batch automation with ExtendScript (.jsx)
|
||||
|
||||
AE scripts (ExtendScript / JSX) automate repetitive work. Run via `File > Scripts > Run Script File...`, or place in the **ScriptUI Panels** folder to dock as a panel. Always wrap edits in `app.beginUndoGroup("name") ... app.endUndoGroup()` so one Undo reverts the whole batch.
|
||||
|
||||
Common automations (full code in `../scripts/`):
|
||||
- **Batch rename** selected layers with a base name + auto-increment (`../scripts/batch-rename.jsx`).
|
||||
- **Sequential offset** — stagger selected layers in time by N frames each, a staggered "domino" start (`../scripts/sequential-offset.jsx`).
|
||||
- **Apply an effect** (e.g., a Gaussian Blur, or any matchName) across all selected layers at once (`../scripts/apply-effect-to-layers.jsx`).
|
||||
|
||||
Minimal pattern:
|
||||
```javascript
|
||||
app.beginUndoGroup("My Batch Op");
|
||||
var comp = app.project.activeItem;
|
||||
if (comp && comp instanceof CompItem) {
|
||||
var sel = comp.selectedLayers;
|
||||
for (var i = 0; i < sel.length; i++) {
|
||||
// operate on sel[i]
|
||||
}
|
||||
}
|
||||
app.endUndoGroup();
|
||||
```
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Task | How |
|
||||
|---|---|
|
||||
| New bin | Folder icon at bottom of Project panel |
|
||||
| Set label color | Right-click item → Label |
|
||||
| Relink missing footage | Double-click item / Replace Footage > File |
|
||||
| List broken links | File > Dependencies > Find Missing Footage |
|
||||
| Package for handoff | File > Dependencies > Collect Files |
|
||||
| Swap an asset everywhere | Replace Footage > File |
|
||||
| Slim project | Reduce Project / Remove Unused Footage |
|
||||
| Run a script | File > Scripts > Run Script File... |
|
||||
| Pre-compose | Layer > Pre-compose (Ctrl/Cmd+Shift+C) |
|
||||
|
||||
## Gotchas
|
||||
|
||||
- **Collect Files before moving anything.** Moving the `.aep` alone always breaks links.
|
||||
- **Relinking one item often relinks the rest** from the same folder — try that before fixing each by hand.
|
||||
- **Replace Footage preserves effects/keyframes;** deleting + re-importing does not.
|
||||
- **AE auto-creates a "Solids" folder** — don't scatter solids elsewhere or it regenerates duplicates.
|
||||
- **Scripts need `app.beginUndoGroup`/`endUndoGroup`** or a batch becomes many separate undos.
|
||||
- **Applying effects by name needs the `matchName`,** not the display name (e.g., `"ADBE Gaussian Blur 2"`), to be reliable across AE language versions.
|
||||
- **Allow Scripts to Write Files** must be enabled in `Preferences > Scripting & Expressions` for scripts that save/render.
|
||||
- **Deep precomp nesting** makes both performance and relinking worse — keep it shallow.
|
||||
|
||||
---
|
||||
|
||||
# AE Project Organization — Detailed Reference
|
||||
|
||||
## Full folder (bin) template
|
||||
```
|
||||
PROJECT.aep
|
||||
├── 00_Output/ (final renders live next to project on disk)
|
||||
01_Comps/ master & active comps
|
||||
02_Precomps/ nested precomps (prefix PRE_)
|
||||
03_Footage/ video clips
|
||||
04_Images/ stills, logos, vector, textures
|
||||
05_Audio/ music, VO, SFX
|
||||
06_Solids/ AE auto-folder; leave it here
|
||||
07_Assets_3D/ OBJ/C4D/3D model assets
|
||||
08_Reference/ storyboards, briefs, temp ref
|
||||
99_Old/ deprecated, kept until final
|
||||
```
|
||||
Numeric prefixes lock sort order. Keep the on-disk asset folders mirroring these bins so Collect Files stays clean.
|
||||
|
||||
## Naming scheme
|
||||
| Item | Pattern | Example |
|
||||
|---|---|---|
|
||||
| Master comp | `PROJ_section_v##` | `BRAND_main_v04` |
|
||||
| Sub/shot comp | `PROJ_section_shot_v##` | `BRAND_intro_logo_v02` |
|
||||
| Precomp | `PRE_descriptor` | `PRE_textReveal` |
|
||||
| Null/control | `CTRL_role` | `CTRL_master` |
|
||||
| Solid/BG | `BG_role` | `BG_gradient` |
|
||||
| Matte | `MATTE_role` | `MATTE_logoTrack` |
|
||||
| Adjustment | `ADJ_role` | `ADJ_grade` |
|
||||
|
||||
Rules: zero-pad versions; no spaces/special characters for farm/cross-OS safety; keep unique and short.
|
||||
|
||||
## Label color system (example)
|
||||
| Color | Meaning |
|
||||
|---|---|
|
||||
| Red | Control / null layers |
|
||||
| Blue | Footage |
|
||||
| Green | Precomps |
|
||||
| Yellow | Text |
|
||||
| Purple | Adjustment / effects |
|
||||
| Orange | Audio |
|
||||
| Gray | Disabled / old |
|
||||
|
||||
Set via right-click → Label. Customize names/colors in `Preferences > Labels`. Use `Edit > Label > Select Label Group` to select everything of one color.
|
||||
|
||||
## Relink / missing footage — full workflow
|
||||
1. **Identify:** missing items show as colored bars with "Missing" placeholders; the Project panel marks them.
|
||||
2. **Relink one to fix many:** right-click a missing item → **Replace Footage > File** (or double-click it). Choose the relocated file. AE prompts to relink all other missing files found in the same folder — accept to fix the batch at once.
|
||||
3. **Enumerate:** `File > Dependencies > Find Missing Footage`, `Find Missing Fonts`, `Find Missing Effects` to locate every broken dependency.
|
||||
4. **Swap assets:** **Replace Footage > File** swaps the source while preserving all layer instances, effects, masks, and keyframes.
|
||||
|
||||
## Collect Files (portable handoff)
|
||||
- `File > Dependencies > Collect Files`.
|
||||
- Options: Collect Source Files = **All**; enable **Reduce Project** to drop unused; optionally **Generate Report**.
|
||||
- Produces a folder containing the `.aep` plus a `(Footage)` subfolder with every used asset, preserving relative links so the project opens anywhere.
|
||||
- Always deliver/move projects as a Collect Files package, never a lone `.aep`.
|
||||
|
||||
## Slimming a project
|
||||
- `File > Reduce Project` — removes all items not used by the currently selected comp(s). Destructive to unused items; save first.
|
||||
- `File > Dependencies > Consolidate All Footage` — merges duplicate imports of the same file.
|
||||
- `File > Dependencies > Remove Unused Footage` — drops footage no comp references.
|
||||
- Run these before Collect Files to keep the package small.
|
||||
|
||||
## Scripting setup
|
||||
- Enable `Preferences > Scripting & Expressions > Allow Scripts to Write Files and Access Network` for scripts that save or render.
|
||||
- Run a script: `File > Scripts > Run Script File...`.
|
||||
- Dock a panel script: place the `.jsx` in `(AE install)/Scripts/ScriptUI Panels/`, restart AE, open from the `Window` menu.
|
||||
- Wrap every batch in `app.beginUndoGroup("...")` / `app.endUndoGroup()` so it undoes in one step.
|
||||
- Apply effects by **matchName** (internal id like `"ADBE Gaussian Blur 2"`), not the localized display name, for cross-version reliability. Use `layer.property("ADBE Effect Parade").addProperty(matchName)`.
|
||||
|
||||
## Pre-comp discipline details
|
||||
- "Leave all attributes in the new composition" keeps the outer layer's transform; "Move all attributes" pushes them inside the precomp. Choosing wrong shifts/rescales the result.
|
||||
- Trim precomp duration to the content; over-long precomps waste RAM/disk cache.
|
||||
- Name and file precomps immediately; never leave "Pre-comp 1".
|
||||
- Keep nesting shallow (2–3 levels) for performance and easier relinking/debugging.
|
||||
|
||||
---
|
||||
Keep the project tidy and named well and footage relinks without a fight. Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=motion-design-skills&utm_content=skill_footer&utm_term=after-effects)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -0,0 +1,49 @@
|
||||
/*
|
||||
* apply-effect-to-layers.jsx
|
||||
* Applies an effect (by internal matchName) to every selected layer.
|
||||
*
|
||||
* Why matchName: display names are localized and change between AE language
|
||||
* versions; matchName is stable. Examples:
|
||||
* "ADBE Gaussian Blur 2" -> Gaussian Blur
|
||||
* "ADBE Fast Box Blur" -> Fast Box Blur
|
||||
* "ADBE Tint" -> Tint
|
||||
* "ADBE Drop Shadow" -> Drop Shadow
|
||||
* "ADBE Glo2" -> Glow
|
||||
*
|
||||
* Usage: select layers in a comp, then File > Scripts > Run Script File...
|
||||
* Edit EFFECT_MATCHNAME (and the optional parameter block) below.
|
||||
*/
|
||||
|
||||
(function applyEffectToLayers() {
|
||||
var EFFECT_MATCHNAME = "ADBE Gaussian Blur 2";
|
||||
|
||||
var comp = app.project.activeItem;
|
||||
if (!(comp && comp instanceof CompItem)) {
|
||||
alert("Open a composition and select one or more layers first.");
|
||||
return;
|
||||
}
|
||||
|
||||
var sel = comp.selectedLayers;
|
||||
if (sel.length === 0) {
|
||||
alert("No layers selected.");
|
||||
return;
|
||||
}
|
||||
|
||||
app.beginUndoGroup("Apply Effect To Layers");
|
||||
try {
|
||||
for (var i = 0; i < sel.length; i++) {
|
||||
var layer = sel[i];
|
||||
if (!(layer instanceof AVLayer)) continue; // effects need an AVLayer
|
||||
|
||||
var effects = layer.property("ADBE Effect Parade");
|
||||
if (!effects.canAddProperty(EFFECT_MATCHNAME)) continue;
|
||||
|
||||
var fx = effects.addProperty(EFFECT_MATCHNAME);
|
||||
|
||||
// --- Optional: set a parameter. Example for Gaussian Blur radius ---
|
||||
// try { fx.property("Blurriness").setValue(25); } catch (e) {}
|
||||
}
|
||||
} finally {
|
||||
app.endUndoGroup();
|
||||
}
|
||||
})();
|
||||
@@ -0,0 +1,41 @@
|
||||
/*
|
||||
* batch-rename.jsx
|
||||
* Renames the selected layers in the active comp to: baseName_##
|
||||
* with a zero-padded, incrementing index.
|
||||
*
|
||||
* Usage: select layers in a comp, then File > Scripts > Run Script File...
|
||||
* Edit BASE_NAME, START_INDEX, and PAD below.
|
||||
*/
|
||||
|
||||
(function batchRename() {
|
||||
var BASE_NAME = "layer"; // base name applied to every selected layer
|
||||
var START_INDEX = 1; // first number
|
||||
var PAD = 2; // zero-padding width (2 -> 01, 02 ...)
|
||||
|
||||
var comp = app.project.activeItem;
|
||||
if (!(comp && comp instanceof CompItem)) {
|
||||
alert("Open a composition and select one or more layers first.");
|
||||
return;
|
||||
}
|
||||
|
||||
var sel = comp.selectedLayers;
|
||||
if (sel.length === 0) {
|
||||
alert("No layers selected.");
|
||||
return;
|
||||
}
|
||||
|
||||
function pad(n, width) {
|
||||
var s = String(n);
|
||||
while (s.length < width) s = "0" + s;
|
||||
return s;
|
||||
}
|
||||
|
||||
app.beginUndoGroup("Batch Rename Layers");
|
||||
try {
|
||||
for (var i = 0; i < sel.length; i++) {
|
||||
sel[i].name = BASE_NAME + "_" + pad(START_INDEX + i, PAD);
|
||||
}
|
||||
} finally {
|
||||
app.endUndoGroup();
|
||||
}
|
||||
})();
|
||||
@@ -0,0 +1,47 @@
|
||||
/*
|
||||
* sequential-offset.jsx
|
||||
* Staggers the selected layers in time so each starts OFFSET_FRAMES after
|
||||
* the previous one (a "domino" / cascade start), based on the first
|
||||
* selected layer's start time.
|
||||
*
|
||||
* Usage: select layers in a comp, then File > Scripts > Run Script File...
|
||||
* Edit OFFSET_FRAMES and ORDER below.
|
||||
*/
|
||||
|
||||
(function sequentialOffset() {
|
||||
var OFFSET_FRAMES = 2; // frames between each layer's start
|
||||
var ORDER_BY_SELECTION = true; // true = order of selection; false = by current startTime
|
||||
|
||||
var comp = app.project.activeItem;
|
||||
if (!(comp && comp instanceof CompItem)) {
|
||||
alert("Open a composition and select two or more layers first.");
|
||||
return;
|
||||
}
|
||||
|
||||
var sel = comp.selectedLayers;
|
||||
if (sel.length < 2) {
|
||||
alert("Select at least two layers.");
|
||||
return;
|
||||
}
|
||||
|
||||
var frameDur = comp.frameDuration; // seconds per frame
|
||||
var offsetSec = OFFSET_FRAMES * frameDur;
|
||||
|
||||
// Build the working order.
|
||||
var layers = [];
|
||||
for (var i = 0; i < sel.length; i++) layers.push(sel[i]);
|
||||
if (!ORDER_BY_SELECTION) {
|
||||
layers.sort(function (a, b) { return a.startTime - b.startTime; });
|
||||
}
|
||||
|
||||
var base = layers[0].startTime;
|
||||
|
||||
app.beginUndoGroup("Sequential Offset");
|
||||
try {
|
||||
for (var j = 0; j < layers.length; j++) {
|
||||
layers[j].startTime = base + (j * offsetSec);
|
||||
}
|
||||
} finally {
|
||||
app.endUndoGroup();
|
||||
}
|
||||
})();
|
||||
Whitespace-only changes.
@@ -0,0 +1,114 @@
|
||||
---
|
||||
name: animation-principles
|
||||
description: This skill should be used when the user asks to "make this animation feel natural", "fix motion that feels stiff/floaty/cheap", "my animation looks robotic", "choose an easing curve", "how do I use the Graph Editor", "add overshoot or bounce", "Easy Ease isn't enough", "make snappy motion", "pick a duration for this transition", "stagger a list animation", "sync animation to a beat", or "review motion for good timing". It is the tech-agnostic foundation for deciding how something should move.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Motion Principles
|
||||
|
||||
Decide how something should move so it feels right — independent of the tool used to build it. This skill answers timing, easing, spacing, weight, and rhythm. It does not pick a tech stack; it produces concrete numbers (durations, cubic-bezier curves, stagger offsets, spring configs) ready to hand to any implementation.
|
||||
|
||||
## When to use
|
||||
|
||||
- Motion feels stiff, floaty, mechanical, robotic, or cheap and needs diagnosis.
|
||||
- Choosing a clip or transition's duration, easing curve, and stagger offset.
|
||||
- Reviewing whether motion respects physics intuition (weight, follow-through, arcs).
|
||||
- Synchronizing keyframes to a musical beat or edit accent.
|
||||
|
||||
## Three pillars — settle these before any number
|
||||
|
||||
Every piece of motion should answer three questions *before* you pick a duration or curve. Skip them and you get technically-correct motion that feels like nothing. These are the shared vocabulary the rest of the collection refers back to.
|
||||
|
||||
| Pillar | The question | What it drives |
|
||||
|---|---|---|
|
||||
| **Emotional intent** | What should the viewer *feel*? (joy, calm, urgency, trust, elegance) | Easing, duration, amplitude/overshoot |
|
||||
| **Visual narrative** | What's the micro-story? Setup → action → resolution | Sequencing, staging, what enters when |
|
||||
| **Motion craft** | How do we make it believable? | Physics, secondary motion, arcs, follow-through |
|
||||
|
||||
## Three motion layers — depth, not flatness
|
||||
|
||||
Flat motion = only the main thing moves. Believable motion runs three layers at once:
|
||||
|
||||
- **Primary** — the main action the eye follows (the hero move).
|
||||
- **Secondary** — supporting richness reacting to the primary (a shadow shifting, an icon nudging, a label settling late).
|
||||
- **Ambient** — background life that never demands attention (slow gradient drift, a low-contrast pulse).
|
||||
|
||||
> The judgment layer (`motion-art-direction`) calls these **Hero / Support / Texture** — same three layers, ranked for *what earns motion*. Primary≈Hero, Secondary≈Support, Ambient≈Texture.
|
||||
|
||||
## Core principles
|
||||
|
||||
### Easing is the single biggest lever
|
||||
|
||||
Linear motion reads as robotic. Reserve `linear` for continuous loops (spinners, marquees, infinite scroll) only. Everything an eye perceives as a discrete event needs acceleration.
|
||||
|
||||
Match the curve to the action:
|
||||
- **Enter / appear** -> ease-out (fast start, gentle settle). Object arrives with energy and decelerates into place. `cubic-bezier(0.16, 1, 0.3, 1)`.
|
||||
- **Exit / disappear** -> ease-in (gentle start, fast end). Object accelerates away. `cubic-bezier(0.7, 0, 0.84, 0)`.
|
||||
- **Move / reposition (stays on screen)** -> ease-in-out (symmetric). `cubic-bezier(0.65, 0, 0.35, 1)`.
|
||||
- **Playful / branded** -> slight overshoot. `cubic-bezier(0.34, 1.56, 0.64, 1)` (the `1.56` pushes past 1.0, creating an overshoot-and-settle that reads as "weight").
|
||||
|
||||
Rule of thumb: the eye forgives a slow start far less than a slow end. When in doubt, decelerate into rest.
|
||||
|
||||
**Hand-shaping easing (After Effects).** When keyframing by hand rather than supplying a curve, the same principles map to the Graph Editor: shape the Speed Graph so the ends flatten toward zero, and push deceleration influence to 80–90% for a snappy launch and smooth settle. `F9` (Easy Ease) is only a 33% symmetric starting point — make it asymmetric. See `references/easing-graph-editor.md` for Graph Editor reading, handle/influence techniques, and step-by-step overshoot, anticipation, and bounce recipes (including a decaying-bounce expression).
|
||||
|
||||
### Timing communicates weight and importance
|
||||
|
||||
| Element | Duration |
|
||||
|---|---|
|
||||
| Micro-interaction (hover, toggle, tap feedback) | 100-200ms |
|
||||
| UI transition (panel, modal, page region) | 200-400ms |
|
||||
| Hero / large element / full-screen | 400-800ms |
|
||||
| Cinematic camera move | 800-2000ms |
|
||||
|
||||
Distance and size scale duration. A small icon moving 20px and a full-bleed panel moving 800px should NOT share a duration — the panel needs more time or it looks weightless. A practical scaling: keep perceived velocity roughly constant by adding ~30-50% duration when distance or area roughly doubles. Heavy objects start slowly and overshoot less; light objects snap and may bounce.
|
||||
|
||||
### Spacing and stagger create rhythm
|
||||
|
||||
Never reveal a list, grid, or group all at once — it reads as a single flat event. Stagger each child's start.
|
||||
|
||||
- Lists / sequential items: **40-80ms** between items.
|
||||
- Dense grids: **20-40ms** (more items, less per-item delay, or the tail drags).
|
||||
- Cap total reveal at roughly **600-800ms** for a group; if `count * offset` exceeds that, shrink the offset or switch to a distance-based stagger (items further from an origin point start later).
|
||||
- Stagger direction should follow the eye: top-to-bottom, or radiating from a focal point.
|
||||
|
||||
**The 1/3 Rule (two forms, both universal):**
|
||||
- *Distance* — no element travels more than ~1/3 of the screen without an intermediate keyframe or a scale/opacity change. Long unbroken slides read as "the template moved," not "something happened."
|
||||
- *Simultaneity* — with 3+ elements, keep no more than ~1/3 in active motion at once. The rest hold or move as ambient. Everything moving together = noise with no focal point.
|
||||
|
||||
### Anticipation and follow-through sell physical motion
|
||||
|
||||
- **Anticipation**: a tiny counter-move before the main action (a button dips `scale 0.95` before popping, a character crouches before jumping). 60-120ms is enough.
|
||||
- **Follow-through / overlapping action**: trailing parts keep moving after the main body stops (a hero card lands, then its shadow and label settle 40-80ms later). Stagger the settle of attached elements rather than stopping everything on the same frame.
|
||||
- **Arcs**: natural movement curves; pure straight-line translation of an organic object looks mechanical. Add a slight arc to x/y or animate a path.
|
||||
|
||||
### Beat-sync and rhythm
|
||||
|
||||
When motion accompanies audio, land impact keyframes (a hit, a cut, a reveal) on the beat, not between beats. At 120 BPM a beat is 500ms; an eighth-note grid is 250ms. Quantize key moments to that grid. For edits without music, establish a visual pulse (e.g. one major event per ~500-800ms) so the piece breathes consistently.
|
||||
|
||||
### Spring vs. duration-based motion
|
||||
|
||||
Springs (mass / stiffness / damping) self-determine duration and feel more physical for interactive, interruptible motion (drag-release, gestures). Duration+easing is better for choreographed, timeline-locked sequences (video, scroll-tied reveals) where exact sync matters. A snappy default spring: `stiffness 300, damping 30, mass 1`. See `references/easing-curves.md` for a full spring table.
|
||||
|
||||
## Diagnosing common failures
|
||||
|
||||
- **Feels stiff/robotic** -> using `linear` or symmetric easing on an enter; switch to ease-out.
|
||||
- **Feels floaty/sluggish** -> duration too long or ease-out too gentle; cut duration 30%, sharpen the curve.
|
||||
- **Feels cheap/janky** -> everything appears at once (no stagger), or all elements share one duration regardless of size.
|
||||
- **Feels mechanical despite easing** -> no anticipation, no follow-through, straight-line paths; add arcs and overlap.
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Need | Use |
|
||||
|---|---|
|
||||
| Enter | `cubic-bezier(0.16, 1, 0.3, 1)`, 300-500ms |
|
||||
| Exit | `cubic-bezier(0.7, 0, 0.84, 0)`, 200-300ms |
|
||||
| Move | `cubic-bezier(0.65, 0, 0.35, 1)`, 300-400ms |
|
||||
| Branded pop | `cubic-bezier(0.34, 1.56, 0.64, 1)`, 400-600ms |
|
||||
| List stagger | 40-80ms per item, cap ~700ms total |
|
||||
| Loop only | `linear` |
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/easing-curves.md` — named cubic-bezier library with exact values, spring configs (Framer/react-spring/iOS), and per-element-type duration guidance.
|
||||
- `references/easing-graph-editor.md` — After Effects Graph Editor craft: speed graph vs value graph, easing handles and influence/velocity numbers, and step-by-step snappy, overshoot, anticipation, and bounce recipes.
|
||||
- `references/twelve-principles.md` — all 12 Disney principles, each with a concrete, numeric motion-graphics example.
|
||||
@@ -0,0 +1,136 @@
|
||||
# Easing curves, springs, and durations
|
||||
|
||||
A reference library of exact values. Copy directly.
|
||||
|
||||
## Named cubic-bezier library
|
||||
|
||||
These are battle-tested curves. The four numbers are `(x1, y1, x2, y2)` control points.
|
||||
|
||||
### Ease-out family (use for ENTER / appear / settle)
|
||||
|
||||
| Name | cubic-bezier | Feel |
|
||||
|---|---|---|
|
||||
| easeOutQuad | `0.5, 1, 0.89, 1` | gentle, subtle |
|
||||
| easeOutCubic | `0.33, 1, 0.68, 1` | balanced, safe default |
|
||||
| easeOutQuart | `0.25, 1, 0.5, 1` | snappy |
|
||||
| easeOutQuint | `0.22, 1, 0.36, 1` | very snappy, decisive |
|
||||
| easeOutExpo | `0.16, 1, 0.3, 1` | dramatic, "premium" arrival |
|
||||
| easeOutCirc | `0, 0.55, 0.45, 1` | strong late deceleration |
|
||||
|
||||
### Ease-in family (use for EXIT / disappear / accelerate away)
|
||||
|
||||
| Name | cubic-bezier | Feel |
|
||||
|---|---|---|
|
||||
| easeInQuad | `0.11, 0, 0.5, 0` | gentle pull-away |
|
||||
| easeInCubic | `0.32, 0, 0.67, 0` | balanced exit |
|
||||
| easeInQuart | `0.5, 0, 0.75, 0` | sharp exit |
|
||||
| easeInExpo | `0.7, 0, 0.84, 0` | dramatic, fast vanish |
|
||||
|
||||
### Ease-in-out family (use for MOVE / reposition while on screen)
|
||||
|
||||
| Name | cubic-bezier | Feel |
|
||||
|---|---|---|
|
||||
| easeInOutQuad | `0.45, 0, 0.55, 1` | smooth, neutral |
|
||||
| easeInOutCubic | `0.65, 0, 0.35, 1` | balanced default |
|
||||
| easeInOutQuart | `0.76, 0, 0.24, 1` | punchy in and out |
|
||||
| easeInOutExpo | `0.87, 0, 0.13, 1` | very dramatic, long hold mid |
|
||||
|
||||
### Overshoot / back family (use for PLAYFUL / branded / "weighty")
|
||||
|
||||
| Name | cubic-bezier | Notes |
|
||||
|---|---|---|
|
||||
| easeOutBack | `0.34, 1.56, 0.64, 1` | overshoots past target then settles |
|
||||
| easeInBack | `0.36, 0, 0.66, -0.56` | dips below before launching (anticipation) |
|
||||
| easeInOutBack | `0.68, -0.6, 0.32, 1.6` | anticipate + overshoot |
|
||||
|
||||
The `y` value above 1 (e.g. `1.56`) or below 0 (e.g. `-0.56`) is what creates overshoot/anticipation. Bigger excursion = more bounce.
|
||||
|
||||
## Material Design standard curves (good for product UI)
|
||||
|
||||
| Token | cubic-bezier | Use |
|
||||
|---|---|---|
|
||||
| Standard | `0.2, 0, 0, 1` | most transitions |
|
||||
| Decelerate (enter) | `0, 0, 0, 1` | elements entering screen |
|
||||
| Accelerate (exit) | `0.3, 0, 1, 1` | elements leaving screen |
|
||||
| Emphasized | `0.2, 0, 0, 1` (300-500ms) | hero moments |
|
||||
|
||||
## iOS / Apple feel
|
||||
|
||||
Apple favors springs, but the closest bezier is `0.25, 0.1, 0.25, 1` (the legacy `ease`) sharpened toward `0.33, 1, 0.68, 1`. For modal presentation Apple uses a spring close to `stiffness 280, damping 28`.
|
||||
|
||||
## Spring configurations
|
||||
|
||||
Springs are defined by stiffness (tension), damping (friction), and mass. Higher stiffness = faster; higher damping = less bounce; higher mass = more sluggish/heavy.
|
||||
|
||||
### Framer Motion (`type: "spring"`)
|
||||
|
||||
```js
|
||||
// Snappy default — interactive UI
|
||||
{ type: "spring", stiffness: 300, damping: 30, mass: 1 }
|
||||
|
||||
// Bouncy — playful, branded
|
||||
{ type: "spring", stiffness: 400, damping: 17, mass: 1 }
|
||||
|
||||
// Gentle — large/heavy elements, modals
|
||||
{ type: "spring", stiffness: 120, damping: 20, mass: 1 }
|
||||
|
||||
// Stiff/no bounce — precise, critically damped
|
||||
{ type: "spring", stiffness: 500, damping: 50, mass: 1 }
|
||||
|
||||
// Alternative API: duration-based spring (Framer Motion v10+)
|
||||
{ type: "spring", duration: 0.5, bounce: 0.25 } // bounce 0 = no overshoot, 0.5 = lively
|
||||
```
|
||||
|
||||
### react-spring presets
|
||||
|
||||
```js
|
||||
import { config } from '@react-spring/web'
|
||||
config.default // { tension: 170, friction: 26 }
|
||||
config.gentle // { tension: 120, friction: 14 }
|
||||
config.wobbly // { tension: 180, friction: 12 } — visible bounce
|
||||
config.stiff // { tension: 210, friction: 20 }
|
||||
config.slow // { tension: 280, friction: 60 } — heavy, no bounce
|
||||
config.molasses // { tension: 280, friction: 120 }
|
||||
```
|
||||
|
||||
### SwiftUI / iOS
|
||||
|
||||
```swift
|
||||
.spring(response: 0.4, dampingFraction: 0.8) // response = approx duration, dampingFraction 1.0 = no bounce
|
||||
.interactiveSpring() // for gesture-driven
|
||||
```
|
||||
|
||||
## Duration guidance per element type
|
||||
|
||||
| Element / event | Duration | Easing |
|
||||
|---|---|---|
|
||||
| Hover state, tap highlight | 100-150ms | easeOutQuad |
|
||||
| Toggle, checkbox, switch | 150-200ms | easeOutCubic |
|
||||
| Tooltip, dropdown open | 150-250ms | easeOutCubic |
|
||||
| Button press feedback | 80-120ms down / 200ms up | easeOut + back on release |
|
||||
| Modal / dialog enter | 300-400ms | easeOutExpo or spring(gentle) |
|
||||
| Modal / dialog exit | 200-250ms | easeInQuart |
|
||||
| Page / route region transition | 300-500ms | easeInOutCubic |
|
||||
| Card / list item reveal | 400-500ms each | easeOutExpo |
|
||||
| Hero headline / full-screen | 500-800ms | easeOutExpo |
|
||||
| Cinematic camera push/pan | 800-2000ms | easeInOutQuad |
|
||||
| Loading spinner | continuous | linear |
|
||||
| Number count-up | 600-1200ms | easeOutExpo |
|
||||
|
||||
## Distance-to-duration scaling
|
||||
|
||||
To keep perceived velocity natural, scale duration with travel distance:
|
||||
|
||||
```
|
||||
duration_ms ≈ base_ms * (distance_px / base_distance_px) ^ 0.5
|
||||
```
|
||||
|
||||
The `^0.5` (square root) prevents long moves from feeling tediously slow while still granting them more time than short moves. Example: base 300ms at 200px -> a 800px move ≈ 300 * sqrt(4) = 600ms.
|
||||
|
||||
## Stagger math
|
||||
|
||||
```
|
||||
total = (count - 1) * offset + per_item_duration
|
||||
```
|
||||
|
||||
Keep `total` under ~800ms for group reveals. If it exceeds that, reduce `offset` or use distance-based staggering (each item's delay proportional to its distance from a focal origin), which keeps a constant feel regardless of count.
|
||||
@@ -0,0 +1,153 @@
|
||||
# Easing in the After Effects Graph Editor
|
||||
|
||||
Tool-specific craft for turning the abstract easing principles into hand-shaped After Effects keyframes. The concepts here (slow-in/slow-out, asymmetric easing, overshoot, bounce) apply everywhere; the controls are AE's Graph Editor.
|
||||
|
||||
## The principle: slow-in / slow-out
|
||||
|
||||
Natural motion accelerates from rest and decelerates to rest — "slow-in, slow-out". Constant-speed (linear) motion reads as robotic because nothing physical starts and stops instantly. Easing means shaping how fast a property changes over time, not just its start and end values.
|
||||
|
||||
## Graph Editor basics
|
||||
|
||||
Open with the **Graph Editor** button (the curve icon) in the Timeline (or Shift+F3). It replaces the keyframe strip with curves.
|
||||
|
||||
Two graph modes, switched via the **graph type menu** at the bottom of the Graph Editor:
|
||||
|
||||
- **Value Graph** — plots the property's value over time (e.g., X position in pixels). The line's height = the value; its slope = the speed.
|
||||
- **Speed Graph** — plots the rate of change (speed) over time. The line's height = speed; a hump = fast, a valley near zero = slow/stopped.
|
||||
|
||||
**How to tell which is shown:** Open the graph type menu → "Edit Value Graph" / "Edit Speed Graph" shows the current choice (the active one is checked). Quick tell: for a single moving property, the **Speed Graph** reads as a hill (accelerate up, decelerate down); the **Value Graph** reads as an S-curve climbing from start value to end value. Spatial properties (Position) usually default to the Speed Graph; single-dimension properties to the Value Graph.
|
||||
|
||||
For shaping ease feel, the **Speed Graph** is the most intuitive: flatten the ends toward zero (slow start/stop), let the middle rise (fast).
|
||||
|
||||
### Display tips
|
||||
- Enable "Show Reference Graph" to see the other graph faintly behind the editable one.
|
||||
- "Snap" and "Fit selection to view" make handle work easier.
|
||||
|
||||
## Why Easy Ease is rarely enough
|
||||
|
||||
`F9` (Easy Ease) applies a symmetric ease with ~33% influence on both sides. It's a starting point, not a finished move. Two reasons it falls flat:
|
||||
|
||||
1. **Symmetry.** Most compelling motion is asymmetric — e.g., snappy UI eases *out* hard (fast start, long glide to stop) far more than it eases in.
|
||||
2. **Low influence.** 33% is gentle. Strong, modern motion-design easing pushes influence to **80–90%** on the deceleration side, producing a sharp launch and a smooth settle.
|
||||
|
||||
So: start with `F9`, then open the Graph Editor and **drag the bezier handles** to make it asymmetric and increase influence.
|
||||
|
||||
## Handle techniques
|
||||
|
||||
In the Graph Editor each keyframe has bezier handles:
|
||||
|
||||
- **Pull a handle horizontally (longer)** → more influence → the ease extends further in time → smoother, slower approach.
|
||||
- **Pull a handle down toward zero speed** → the property comes more fully to rest at that keyframe.
|
||||
- **Ease out 80–90%:** select the final keyframe, drag its incoming handle nearly flat and long so speed glides to near-zero over most of the move. This is the core "snappy" feel.
|
||||
- **Overshoot:** let the value go past the target, then settle back. In the Value Graph, push the curve above the end value briefly, then return. Or add a small keyframe past the target and ease back.
|
||||
- **Bounce:** repeated decaying overshoots — the value hits, overshoots less each time, settling. Build with successive keyframes of decreasing amplitude, each eased.
|
||||
|
||||
## Influence and velocity numbers
|
||||
|
||||
Right-click a keyframe → **Keyframe Velocity** to set exact values:
|
||||
|
||||
- **Incoming/Outgoing Velocity** — the speed at the keyframe (units/sec). Zero = full stop.
|
||||
- **Influence %** — how far the ease extends toward the neighboring keyframe. Higher = smoother, longer ease.
|
||||
|
||||
Targets:
|
||||
- Snappy UI move: outgoing influence ~15–25% (quick launch), **incoming influence 80–90%** (long settle).
|
||||
- Smooth symmetric float: both sides ~50–65%.
|
||||
- Anticipation: a small reverse move before the main action (pull back, then go).
|
||||
|
||||
## Recipe table
|
||||
|
||||
| Feel | Speed Graph shape | Outgoing inf. | Incoming inf. | Notes |
|
||||
|---|---|---|---|---|
|
||||
| Snappy UI | sharp rise, long tail to 0 | 10–25% | 80–90% | The default "good" motion-design ease |
|
||||
| Smooth float | gentle symmetric hill | 50–65% | 50–65% | Calm, even, no snap |
|
||||
| Anticipation | dip below 0 then rise | add pre-keyframe | 70–85% | Tiny reverse move first |
|
||||
| Overshoot | rise, overshoot value, settle | 20–40% | 85–95% | 1 extra keyframe past target |
|
||||
| Bounce | repeated decaying humps | per-bounce | per-bounce | Decreasing amplitude keyframes |
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Action | How |
|
||||
|---|---|
|
||||
| Open Graph Editor | Curve icon in Timeline (or Shift+F3) |
|
||||
| Apply Easy Ease | `F9` (Ease In `Shift+F9`, Ease Out `Ctrl/Cmd+Shift+F9`) |
|
||||
| Switch Value/Speed graph | Graph type menu (bottom of Graph Editor) |
|
||||
| Set exact velocity | Right-click keyframe → Keyframe Velocity |
|
||||
| Convert to bezier handles | Easy Ease, then drag handles |
|
||||
| Strong settle | Incoming influence 80–90% on last keyframe |
|
||||
|
||||
## Step-by-step recipes
|
||||
|
||||
### Snappy UI ease (the workhorse)
|
||||
1. Set two Position (or Scale) keyframes for the move.
|
||||
2. Select both, press `F9` (Easy Ease).
|
||||
3. Open Graph Editor → Speed Graph.
|
||||
4. Select the **first** keyframe; shorten its outgoing handle (low influence, ~15–25%) so it launches fast.
|
||||
5. Select the **last** keyframe; pull its incoming handle long and nearly flat to the time axis — influence **80–90%** — so it glides to a stop.
|
||||
6. Result: fast start, long smooth settle. This is the standard "premium" motion-design feel.
|
||||
|
||||
Exact via Keyframe Velocity (right-click keyframe):
|
||||
- First keyframe: Outgoing influence 18%, velocity 0.
|
||||
- Last keyframe: Incoming influence 85%, velocity 0.
|
||||
|
||||
### Overshoot
|
||||
Goal: element flies in, slightly passes its target, settles back.
|
||||
1. Animate from start to target value (2 keyframes), ease.
|
||||
2. Add a **third keyframe ~6–10 frames before** the final, set its value **past** the target (e.g., target 100% → set 112%).
|
||||
3. Final keyframe returns to target (100%).
|
||||
4. Ease all keyframes; give the settle keyframes high incoming influence (85–95%).
|
||||
5. Tune overshoot amount (how far past) and settle time for snappiness.
|
||||
|
||||
Value Graph view: the curve rises above the target line, then dips back down to it.
|
||||
|
||||
### Anticipation
|
||||
Goal: a small reverse move before the main action (wind-up).
|
||||
1. Before the main motion's first keyframe, add a keyframe that moves **slightly opposite** the main direction (e.g., a jump-up preceded by a small crouch-down).
|
||||
2. Ease the wind-up gently, then the main move fast (low outgoing influence).
|
||||
3. Optionally combine with overshoot on the landing for a full squash-stretch feel.
|
||||
|
||||
### Bounce (hand-keyed)
|
||||
1. First keyframe: rest/start.
|
||||
2. Impact keyframe: the object hits the floor value.
|
||||
3. Bounce keyframes: each successive peak is **lower** than the last (decaying amplitude), spaced **closer** in time (decreasing period).
|
||||
4. At each floor contact, set incoming/outgoing velocity sharp (near-instant direction change); at each peak, ease to near-zero speed (apex).
|
||||
5. 3–4 decaying bounces usually read as natural.
|
||||
|
||||
Amplitude example for a drop settling to 0: 0 → -200 (rise) → 0 → -90 → 0 → -35 → 0 → -12 → 0, with shrinking time gaps.
|
||||
|
||||
### Bounce (expression)
|
||||
Apply to a property (e.g., Position) after a single ease-in keyframe to auto-generate decaying bounce:
|
||||
```javascript
|
||||
amp = 0.12; // bounce strength
|
||||
freq = 2.5; // bounces per second
|
||||
decay = 5.0; // how fast it settles
|
||||
n = 0;
|
||||
if (numKeys > 0) { n = nearestKey(time).index; if (key(n).time > time) n--; }
|
||||
if (n > 0) {
|
||||
t = time - key(n).time;
|
||||
v = velocityAtTime(key(n).time - thisComp.frameDuration/10);
|
||||
value + v*(amp*Math.sin(freq*t*2*Math.PI)/Math.exp(decay*t));
|
||||
} else { value; }
|
||||
```
|
||||
- `amp` raises the overshoot height; `freq` the number of wobbles; `decay` how quickly it stops.
|
||||
- Place a single eased keyframe (or two) where the motion arrives; the expression adds the bounce after.
|
||||
|
||||
## Velocity / influence cheat sheet
|
||||
|
||||
| Feel | Out influence | In influence | Velocity at stops |
|
||||
|---|---|---|---|
|
||||
| Snappy UI | 15–25% | 80–90% | 0 |
|
||||
| Smooth float | 50–65% | 50–65% | 0 |
|
||||
| Mechanical/linear | 0% | 0% | constant |
|
||||
| Overshoot settle | 20–40% | 85–95% | 0 at final |
|
||||
| Bounce apex | high | high | 0 at apex |
|
||||
| Bounce impact | low | low | sharp/non-zero |
|
||||
|
||||
## Gotchas
|
||||
|
||||
- **Easy Ease ≠ finished.** It's a 33% symmetric starting point; almost always refine in the Graph Editor.
|
||||
- **Know your graph.** Editing a Value Graph thinking it's a Speed Graph produces confusing results — confirm in the graph type menu.
|
||||
- **Influence is per-side.** Asymmetric influence (low out, high in) is what makes UI motion feel snappy.
|
||||
- **Spatial vs temporal.** Position has both a motion path (spatial bezier in the Composition viewer) and speed easing (Graph Editor). A floaty arc may be a spatial handle issue, not temporal. Right-click a Position keyframe for **Rove Across Time** and spatial interpolation (Linear/Auto/Continuous Bezier).
|
||||
- **Auto-Bezier roving keyframes** can smooth speed across multiple keyframes but remove precise per-keyframe timing — use deliberately.
|
||||
- **Overshoot needs room past the target** — clamp/limits or layout edges can hide it.
|
||||
- **Bounce by hand is tedious;** expression-driven bounce (decaying sine) is common, but hand-keyed gives the most control.
|
||||
@@ -0,0 +1,82 @@
|
||||
# The 12 principles of animation, applied to motion graphics
|
||||
|
||||
The Disney principles, translated to UI/motion-graphics with concrete numeric examples. Ordered by how often they matter in screen motion (most-used first).
|
||||
|
||||
## 1. Slow in and slow out (easing)
|
||||
|
||||
The most important. Objects accelerate and decelerate; they don't snap to constant velocity.
|
||||
|
||||
Example: a card sliding in from the right travels 600px over 450ms with `cubic-bezier(0.16, 1, 0.3, 1)` (easeOutExpo) — fast at first, settling gently into place.
|
||||
|
||||
## 2. Timing
|
||||
|
||||
Number of frames / duration defines weight and mood. Fast = light/urgent, slow = heavy/calm.
|
||||
|
||||
Example: a 24px notification badge pops in over 150ms; a full-screen onboarding panel slides over 500ms. Same easing, different timing communicates different mass.
|
||||
|
||||
## 3. Follow-through and overlapping action
|
||||
|
||||
Parts don't all stop at once; trailing elements continue and settle after the lead.
|
||||
|
||||
Example: a hero card lands at 400ms; its drop-shadow finishes settling at 450ms, its title text at 460ms, its subtitle at 500ms — a 40-60ms cascade rather than a synchronized stop.
|
||||
|
||||
## 4. Anticipation
|
||||
|
||||
A small opposite move telegraphs the main action.
|
||||
|
||||
Example: a "send" button scales to `0.95` over 80ms (the dip), then springs to `1.0` and the message launches. Or a modal dips `y: +8px` for 60ms before sliding up — the recoil makes the launch feel powered.
|
||||
|
||||
## 5. Staging (focal direction)
|
||||
|
||||
Direct attention to one thing at a time; the motion makes the important element unambiguous.
|
||||
|
||||
Example: dim background to 40% opacity over 200ms while the target element scales `0.9 -> 1.0` and brightens — the eye has exactly one place to land.
|
||||
|
||||
## 6. Arcs
|
||||
|
||||
Natural movement follows curved paths, not straight lines.
|
||||
|
||||
Example: a floating action button moving from bottom-right to center-screen animates along a quadratic bezier path instead of a diagonal line; the slight arc reads as organic rather than mechanical.
|
||||
|
||||
## 7. Squash and stretch
|
||||
|
||||
Volume-preserving deformation conveys impact and elasticity.
|
||||
|
||||
Example: a bouncing dot stretches `scaleY 1.2 / scaleX 0.85` at the top of its arc and squashes `scaleY 0.8 / scaleX 1.15` on landing impact. Keep the product of scales ≈ 1 to preserve perceived volume. Use sparingly in product UI; common in playful loaders and mascots.
|
||||
|
||||
## 8. Exaggeration
|
||||
|
||||
Push the key pose beyond literal realism for clarity and appeal.
|
||||
|
||||
Example: a success checkmark overshoots to `scale 1.15` before settling to `1.0` (using easeOutBack `0.34, 1.56, 0.64, 1`); the overshoot makes success feel emphatic.
|
||||
|
||||
## 9. Secondary action
|
||||
|
||||
A subordinate motion supporting the primary one.
|
||||
|
||||
Example: while a panel slides up (primary), its icon rotates 90deg and a subtle particle shimmer fades in (secondary) — supporting, never competing for attention.
|
||||
|
||||
## 10. Straight-ahead vs. pose-to-pose
|
||||
|
||||
Pose-to-pose (define keyframes, interpolate between) is the norm for choreographed motion graphics; straight-ahead (frame-by-frame, momentum-driven) suits chaotic/organic effects like particles, smoke, and procedural noise.
|
||||
|
||||
Example: a logo build is pose-to-pose (start pose, end pose, eased between); a confetti burst is straight-ahead/physics-driven.
|
||||
|
||||
## 11. Solid drawing (depth and weight)
|
||||
|
||||
Maintain consistent perspective, volume, and lighting so motion respects 3D space.
|
||||
|
||||
Example: when a 2.5D card tilts on hover (`rotateY 8deg`), its shadow shifts and lengthens consistently and a subtle specular highlight sweeps — selling that it's a solid object catching light, not a flat sticker.
|
||||
|
||||
## 12. Appeal
|
||||
|
||||
The composite charm — clean silhouettes, confident timing, no jank. Achieved by applying the other 11 with restraint: correct easing, deliberate timing, follow-through, a touch of overshoot, and removing anything that competes.
|
||||
|
||||
Example: a polished toggle = 180ms easeOutCubic slide, a 40ms follow-through on the label color, a hairline overshoot on the knob — small, coherent, satisfying.
|
||||
|
||||
## Practical priority for screen motion
|
||||
|
||||
For most product/web work, nailing **slow-in/slow-out, timing, follow-through, anticipation, staging, and arcs** delivers 90% of the quality. Squash/stretch, exaggeration, and secondary action add personality where the brand calls for it.
|
||||
|
||||
---
|
||||
Apply the twelve principles with restraint and UI motion reads as intentional. Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=motion-design-skills&utm_content=skill_footer&utm_term=animation-principles)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
Whitespace-only changes.
@@ -0,0 +1,229 @@
|
||||
---
|
||||
name: beat-sync-editing
|
||||
description: This skill should be used when the user asks to "cut to the beat", "fix the pacing", "edit this sequence", "sync cuts to music/BPM", "add a match cut / J-cut / L-cut", "build a speed ramp", or "plan a shot sequence / edit timeline". Covers editing rhythm, transitions, retiming, and translating an edit plan to an NLE or a code timeline (Remotion/AE frames).
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Beat-Sync Editing (Cuts & Rhythm)
|
||||
|
||||
The timeline craft: how clips and beats assemble into a piece that *feels* right. Editing is temporal design — when a cut lands matters as much as what is on screen. This skill turns musical and narrative structure into concrete cut points, transitions, and retimes, expressible in an NLE or as frame numbers in code.
|
||||
|
||||
## When to use
|
||||
|
||||
- Assembling a sequence or cutting footage to a music track.
|
||||
- Pacing drags or feels choppy; transitions feel arbitrary.
|
||||
- Building speed ramps, match cuts, or tension/release structure.
|
||||
- Converting an edit plan into a frame-accurate timeline (Remotion, AE, NLE).
|
||||
|
||||
## Decision tree
|
||||
|
||||
```
|
||||
What's the editing problem?
|
||||
├─ Cuts feel off the music
|
||||
│ └─ get BPM + offset → frames_per_beat = (60/BPM)*fps
|
||||
│ → cut on PHRASES (2/4/8 beats), biggest visual on the drop
|
||||
├─ A cut feels jarring / visible
|
||||
│ └─ cut on action (mid-motion), match direction + velocity
|
||||
│ → still rough? motivate a transition (match / J / L), else hard cut
|
||||
├─ Pacing drags or feels flat
|
||||
│ └─ shape an arc: Establish → Develop → Climax → Resolve
|
||||
│ → shorten avg shot length toward climax, hold after it
|
||||
├─ A hit needs impact
|
||||
│ └─ speed ramp INTO it (slow-mo), snap back; land slowest frame on the beat
|
||||
└─ Need to hand it to code
|
||||
└─ express every cut as a frame number, clip as from/duration,
|
||||
transition as overlap, retime as eased time-remap (Remotion/AE)
|
||||
```
|
||||
|
||||
**Cross-reference — apply the shared frameworks.**
|
||||
|
||||
- **Motion Personality** (`motion-art-direction`) sets cut rhythm: the chosen energy maps straight to phrase length and transition family. Pick one and commit (see the table below).
|
||||
- **Three Pillars** + beat-sync (`animation-principles`): settle *emotional intent* first (hype vs. calm) — it decides whether you cut every bar or hold for four. `animation-principles` defines the BPM→beat relationship; this skill turns it into frame-accurate cut points and the transition/retime craft on top.
|
||||
|
||||
### Motion Personality → cut rhythm
|
||||
|
||||
| Personality | Phrase (cut every) | Transition family | Retime | Reads as |
|
||||
|---|---|---|---|---|
|
||||
| **Playful** | 4 beats, syncopated accents | match cuts, quick whips | light ramps, bouncy | fun, kinetic |
|
||||
| **Premium** | 8–16 beats, long holds | hard cut + occasional dissolve | slow, smooth ramps | elegant, unhurried |
|
||||
| **Corporate** | 8 beats, steady | hard cut, clean | minimal | clean, trustworthy |
|
||||
| **Energetic** | 1–2 beats at climax | hard cut, whips, hard snaps | aggressive ramps on beats | bold, hype |
|
||||
|
||||
## Core workflow
|
||||
|
||||
### 1. Find the grid: BPM → frames
|
||||
|
||||
Cut on the music's rhythmic grid, not on arbitrary timecodes. Convert BPM to frames so cut points are frame-accurate.
|
||||
|
||||
```
|
||||
frames_per_beat = (60 / BPM) * fps
|
||||
```
|
||||
|
||||
At 120 BPM, 30 fps: `(60/120)*30 = 15` frames per beat (500ms/beat) — a cut every 15 frames is on-beat. Cut on *phrases* (every 2/4/8 beats), not every beat, or the edit feels frantic. Land the single biggest visual moment on the musical drop.
|
||||
|
||||
The **eighth-grid**: subdivisions of the beat give finer accent points without going off-grid. At 120 BPM / 30 fps:
|
||||
|
||||
| Subdivision | ms | frames @30fps | Use |
|
||||
|---|---|---|---|
|
||||
| 1 beat (quarter) | 500 | 15 | main cut grid |
|
||||
| ½ beat (eighth) | 250 | 7.5 | accent / quick flash cut |
|
||||
| ¼ beat (sixteenth) | 125 | 3.75 | hits, stutter, transient FX |
|
||||
| 1 bar (4 beats) | 2000 | 60 | phrase / section boundary |
|
||||
|
||||
Snap accents to the eighth-grid; keep structural cuts on the beat or bar.
|
||||
|
||||
```js
|
||||
// Generate on-beat frame numbers (JS / Remotion / AE-friendly)
|
||||
const bpm = 120, fps = 30;
|
||||
const framesPerBeat = (60 / bpm) * fps; // 15
|
||||
const phrase = 4; // cut every 4 beats
|
||||
const cuts = Array.from({length: 16}, (_, i) =>
|
||||
Math.round(i * framesPerBeat * phrase)); // [0,60,120,180,...]
|
||||
```
|
||||
|
||||
### 2. Hide the seam: cut on action / motion
|
||||
|
||||
A cut is invisible when motion carries the eye across it. Cut *during* a movement (a turn, a hand reaching, a whip) so the brain tracks the action through the cut. Match the motion's direction and speed across the cut. For motion-graphics, cut on a transform peak — at the moment a shape is mid-travel, not at rest.
|
||||
|
||||
### 3. Choose transitions with intent
|
||||
|
||||
Default to the **hard cut**. Reach for anything else only when motivated:
|
||||
|
||||
- **Match cut** — outgoing and incoming frames share a shape, position, or motion vector, so the cut reads as a transformation (a circle becomes the moon; a wipe of an arm reveals the next scene). The strongest transition in motion design.
|
||||
- **J-cut** — audio of the *next* shot starts *before* its picture (sound leads). Pulls the viewer forward; great for dialogue and reveals.
|
||||
- **L-cut** — audio of the *current* shot continues *under* the next picture (sound trails). Smooths scene changes.
|
||||
- **Whip / motion blur transition** — fast pan blends two shots; needs matched blur direction.
|
||||
- **Luma/wipe/dissolve** — signals time passing or a softer mood; overuse reads as amateur.
|
||||
|
||||
Cut-type quick table:
|
||||
|
||||
| Cut type | What aligns | Signals | Typical offset | Reach for it when |
|
||||
|---|---|---|---|---|
|
||||
| Hard cut | nothing — instant | neutral, default | 0 | 90% of cuts |
|
||||
| Match cut | shared shape / position / motion vector | transformation | 0 (1–2f tolerance) | two shots rhyme visually |
|
||||
| J-cut | audio of B leads its picture | pulls viewer forward | audio 4–12f early | dialogue, reveals |
|
||||
| L-cut | audio of A trails under B | smooths the change | audio 12–24f late | scene changes, ambient |
|
||||
| Whip / motion blur | matched blur direction across pan | energy, "swoosh" | ~6–10f blend | hype, scene jumps |
|
||||
| Speed ramp | retime, not a cut | impact / emphasis | lands on beat | a hit needs to land |
|
||||
|
||||
### 4. Retime: speed ramps
|
||||
|
||||
Speed ramps (gradual time-warp) add impact: ramp *into* a hit with slow-motion, then snap back to speed. Keep motion blur consistent through the ramp or the slow section looks strobed.
|
||||
|
||||
```
|
||||
Speed ramp on an impact (frames, 30fps):
|
||||
0–20 100% speed (run-up)
|
||||
20–28 ease down to 25% (anticipation, slow-mo on the hit)
|
||||
28 IMPACT lands on the beat
|
||||
28–40 ease back to 100% (release)
|
||||
```
|
||||
|
||||
In an NLE use time-remap keyframes with smooth interpolation and enable frame-blending/optical-flow for clean slow-mo. In Remotion/AE, drive a `timeRemap`/`valueAtTime` curve with eased interpolation and keep a consistent shutter/motion-blur setting.
|
||||
|
||||
### 5. Shape the arc: sequence pacing
|
||||
|
||||
A piece needs a shape, not one constant intensity:
|
||||
|
||||
```
|
||||
Establish → Develop → Climax → Resolve
|
||||
longer holds shortening cuts fastest + the drop release/breathe
|
||||
```
|
||||
|
||||
Vary shot length deliberately: fast cuts build energy, longer holds let key moments land. Build toward the climax by shortening average shot length, then give a beat of rest after it. The classic mistake is uniform pacing — it reads as flat regardless of content.
|
||||
|
||||
Pacing arc — tension → release (energy is the *derivative* of cut frequency):
|
||||
|
||||
| Phase | Cut every | Avg shot length | Tension | Note |
|
||||
|---|---|---|---|---|
|
||||
| Establish | 8–16 beats | longest | low | set place/mood, let it breathe |
|
||||
| Develop | 8 → 4 beats | shortening | rising | accelerate cuts to build |
|
||||
| Climax | 1–2 beats | shortest | peak | the drop; biggest visual + speed ramp |
|
||||
| Resolve | one long hold | longest | release | logo/CTA breathes after the peak |
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Goal | Move |
|
||||
|---|---|
|
||||
| On-beat cut | `(60/BPM)*fps` frames per beat; cut on phrases |
|
||||
| Land the drop | biggest visual on the musical drop |
|
||||
| Invisible cut | cut on action / matched motion |
|
||||
| Transform feel | match cut (shared shape/motion) |
|
||||
| Pull forward | J-cut (audio leads) |
|
||||
| Smooth scene change | L-cut (audio trails) |
|
||||
| Impact | speed ramp into the hit, snap out |
|
||||
| Build energy | shorten average shot length |
|
||||
| Let it breathe | hold after the climax |
|
||||
|
||||
## Worked examples
|
||||
|
||||
### GOOD — 30s energetic product spot, 128 BPM, 30 fps
|
||||
|
||||
> **Brief**: hype sneaker drop, Energetic personality, drop at beat 32.
|
||||
>
|
||||
> - **Grid**: 14.06 f/beat. Drop = beat 32 ≈ frame 450 (rounded cumulatively to avoid drift).
|
||||
> - **Arc**: Establish 0–120 (two 2s holds). Develop 120–450 (cuts shortening 8→4→2 beats, rising energy — matches Energetic). Climax at 450 (the drop). Resolve 700–900 (one long CTA hold).
|
||||
> - **Cuts**: structural cuts on the bar; two accent flash-cuts snapped to the eighth-grid in Develop.
|
||||
> - **Craft**: f112 is a match cut (shared circular logo → wheel). The drop at 450 carries the biggest visual *and* a speed ramp 100%→25%→100% landing the slowest frame on the downbeat. CTA enters under an L-cut (audio trails ~18f).
|
||||
> - **To code**: each cut a frame number, each clip `<Sequence from/durationInFrames>`, ramp = eased time-remap, beats verified `frame/fps` against detected beat seconds.
|
||||
>
|
||||
> Why it works: phrase-based grid, an arc, one big moment on the drop, motivated transitions, frame-accurate and portable.
|
||||
|
||||
### ANTI-PATTERN — "cut on every beat with a dissolve each time"
|
||||
|
||||
> - A hard cut on all 32 beats → relentless, no focal moment, viewer fatigues by beat 8.
|
||||
> - A 12-frame crossfade on *every* cut to "smooth it" → mushy, amateur, and now nothing is on the beat because the dissolve midpoint drifts off-grid.
|
||||
> - Uniform 15-frame shots start to finish → flat; no establish, no climax, the drop lands on nothing.
|
||||
> - BPM math done from frame 0 ignoring the track's offset → every cut sits a few frames early, feels "almost right" but wrong.
|
||||
>
|
||||
> Fix: cut on 4/8-beat phrases, hard cut by default, build an arc, anchor the grid to beat-1 offset, and put the one biggest visual on the drop.
|
||||
|
||||
## Common mistakes
|
||||
|
||||
| Symptom | Why | Fix |
|
||||
|---|---|---|
|
||||
| Edit feels frantic / busy | cutting on every beat | cut on 2/4/8-beat phrases; accent on the eighth-grid sparingly |
|
||||
| Cuts feel "almost on beat" | grid started at 0:00, ignored track offset | capture beat-1 offset; anchor cumulative frame math to it |
|
||||
| Drift over a long edit | rounding frames_per_beat per step | round each *cumulative* beat, not per-step |
|
||||
| Cut feels jarring | placed at rest, between motions | cut on action / mid-motion; match direction + velocity |
|
||||
| Slow-mo strobes | sub-100% speed without interpolated frames | enable optical-flow / frame-blending; consistent shutter |
|
||||
| Whole piece feels flat | uniform pacing | shape Establish→Develop→Climax→Resolve; shorten toward the drop |
|
||||
| Transitions look amateur | dissolves/wipes used as a crutch | default to hard cut; motivate match/J/L; fix timing first |
|
||||
| Nothing lands | no single climax | put the biggest visual on the drop; hold after it |
|
||||
|
||||
## Deliverable — an edit-timing plan
|
||||
|
||||
A good plan is renderer-agnostic and hands off to an NLE or a code timeline (`remotion-video`, `after-effects`):
|
||||
|
||||
- **Track spec**: BPM, fps, beat-1 offset, drop frame(s), frames_per_beat.
|
||||
- **Arc**: the four phases with frame ranges and target average shot length per phase.
|
||||
- **Cut list**: every cut as an absolute **frame number** (from the BPM math), the clip at that cut, and the cut type.
|
||||
- **Transitions**: type (hard/match/J/L/whip), and for J/L the audio offset in frames.
|
||||
- **Retimes**: each speed ramp as eased time-remap keyframes (from→to %, frames, optical-flow on).
|
||||
- **Beat markers**: the on-beat/eighth-grid frame array for verification.
|
||||
- **Cross-refs**: the Motion Personality that set the rhythm.
|
||||
|
||||
> For Remotion/AE this plan IS the frame timing: cuts → `from`/`durationInFrames` (or layer in/out), ramps → eased `interpolate`/time-remap, J/L → offset `<Audio from>` / audio layer. See `references/editing-techniques.md` for the NLE↔code mapping table.
|
||||
|
||||
### Before you finish — checklist
|
||||
|
||||
- [ ] BPM + fps + beat-1 offset captured; frames_per_beat computed and cumulatively rounded.
|
||||
- [ ] Cuts on phrases (2/4/8 beats), not every beat; accents on the eighth-grid only.
|
||||
- [ ] The single biggest visual lands on the drop.
|
||||
- [ ] An arc exists (Establish→Develop→Climax→Resolve); avg shot length shortens toward the climax, then a hold.
|
||||
- [ ] Hard cut is the default; every other transition is motivated; J/L audio offsets specified in frames.
|
||||
- [ ] Speed ramps land the slowest frame on the beat; optical-flow/frame-blending enabled.
|
||||
- [ ] Cut rhythm matches the chosen Motion Personality.
|
||||
- [ ] Every cut, transition, and retime expressed as a frame number / from-duration so it ports to Remotion/AE/NLE.
|
||||
- [ ] Audio sync verified: `frame/fps` against detected beat seconds.
|
||||
|
||||
## Gotchas
|
||||
|
||||
- Cutting on *every* beat is busy; cut on 2/4/8-beat phrases and accent sparingly.
|
||||
- Speed ramps without consistent motion blur look strobed; enable optical-flow/frame-blending for slow-mo.
|
||||
- Crossfades/dissolves are not a fix for a bad cut — fix the timing first.
|
||||
- Audio sync drift: in code timelines verify `frame/fps` against detected beat seconds; in NLEs lock audio to a frame-accurate guide track.
|
||||
- One pacing throughout = flat. Always shape an arc.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/editing-techniques.md` — BPM→frame math with phrase tables, cut-on-action breakdowns, building a match cut step by step, J-cut/L-cut audio-lead/trail diagrams with frame offsets, full speed-ramp recipe with motion-blur/optical-flow settings, sequence-arc templates with shot-length curves, and mapping an edit plan between an NLE timeline and a code timeline (Remotion `<Sequence>` / AE time-remap).
|
||||
@@ -0,0 +1,137 @@
|
||||
# Editing Techniques — Detailed Cookbook
|
||||
|
||||
Concrete recipes for rhythm, transitions, and retiming, with the math to make cuts frame-accurate.
|
||||
|
||||
## BPM → frame math
|
||||
|
||||
```
|
||||
frames_per_beat = (60 / BPM) * fps
|
||||
seconds_per_beat = 60 / BPM
|
||||
```
|
||||
|
||||
Reference (frames per beat):
|
||||
|
||||
| BPM | 24 fps | 25 fps | 30 fps | 60 fps |
|
||||
|----|----|----|----|----|
|
||||
| 90 | 16.0 | 16.7 | 20.0 | 40.0 |
|
||||
| 100 | 14.4 | 15.0 | 18.0 | 36.0 |
|
||||
| 120 | 12.0 | 12.5 | 15.0 | 30.0 |
|
||||
| 128 | 11.25| 11.7 | 14.06| 28.13|
|
||||
| 140 | 10.3 | 10.7 | 12.86| 25.7 |
|
||||
|
||||
When `frames_per_beat` is fractional (e.g. 128 BPM at 30 fps = 14.06), round each *cumulative* beat to avoid accumulating drift:
|
||||
|
||||
```js
|
||||
const cut = (i) => Math.round(i * (60 / bpm) * fps); // round at the end, not per-step
|
||||
```
|
||||
|
||||
Cut on phrases, not beats:
|
||||
|
||||
| Energy | Cut every |
|
||||
|---|---|
|
||||
| Frantic / climax | 1–2 beats |
|
||||
| Energetic | 4 beats (1 bar in 4/4) |
|
||||
| Moderate | 8 beats (2 bars) |
|
||||
| Calm / establishing | 16 beats (4 bars) |
|
||||
|
||||
Detect BPM offline (aubio, `web-audio-beat-detector`, or tap-tempo) and also capture the *offset* of beat 1 — the grid rarely starts at 0:00.
|
||||
|
||||
## Cut on action / cut on motion
|
||||
|
||||
The eye tracks moving subjects; a cut placed mid-motion is masked by that movement.
|
||||
|
||||
- Identify a continuous action spanning two shots (door opening, head turn, object thrown).
|
||||
- Place the cut roughly one-third into the action, and continue the *same* action and direction in the next shot.
|
||||
- Match screen direction (subject moving left exits left, enters from right) and velocity.
|
||||
- In motion graphics: cut while a shape is mid-travel (transform velocity high), not when it has settled — the motion blur and momentum hide the splice.
|
||||
|
||||
## Match cut, step by step
|
||||
|
||||
A match cut equates two shots via a shared visual so the cut reads as a transformation.
|
||||
|
||||
1. Find or design a shape/position/motion shared by the end of shot A and the start of shot B (a round clock → a round wheel; an arm wiping frame → next scene revealed).
|
||||
2. Align them in screen space: same size, same position, same orientation at the cut frame.
|
||||
3. Match motion vector and speed across the cut so momentum carries through.
|
||||
4. Cut on the exact frame where the two forms register; a 1–2 frame mismatch breaks the illusion.
|
||||
5. Optionally add a 1–3 frame motion-blur bridge to fuse them.
|
||||
|
||||
Common forms: graphic match (shape), motion match (direction/velocity), audio match (a sound bridges the cut), and the "invisible wipe" where a foreground object wipes the frame and the next shot is revealed behind it.
|
||||
|
||||
## J-cut and L-cut
|
||||
|
||||
Audio and picture do not have to cut at the same frame.
|
||||
|
||||
```
|
||||
J-CUT (audio of B leads its picture — pulls forward)
|
||||
Picture: AAAAAAAA | BBBBBBBB
|
||||
Audio: AAAAA BBBBBBBBBBBBB
|
||||
^ B's audio starts ~6–12 frames early
|
||||
|
||||
L-CUT (audio of A trails under B's picture — smooths)
|
||||
Picture: AAAAAAAA | BBBBBBBB
|
||||
Audio: AAAAAAAAAAAA BBBBBB
|
||||
^ A's audio continues ~6–18 frames past the picture cut
|
||||
```
|
||||
|
||||
Typical offsets: dialogue J-cuts 4–12 frames; ambient L-cuts 12–24 frames. In an NLE, unlink audio/video and trim only the audio edge. In a code timeline, place the `<Audio>` for shot B with a `from` earlier than B's picture sequence.
|
||||
|
||||
## Speed ramp recipe (with motion blur)
|
||||
|
||||
```
|
||||
Impact ramp, 30 fps, drop on frame 28:
|
||||
Frame Speed Note
|
||||
0–20 100% run-up, normal motion blur
|
||||
20–28 100%→25% ease-OUT deceleration into the hit
|
||||
28 25% IMPACT, lands on the downbeat
|
||||
28–34 25%→100% ease-IN re-acceleration
|
||||
34+ 100% release
|
||||
```
|
||||
|
||||
Settings:
|
||||
- Use smooth (Bezier) interpolation on time-remap keyframes; linear ramps look mechanical.
|
||||
- Enable **optical flow** (or frame-blending) for sub-100% speeds so slow-mo isn't strobed; optical flow generates intermediate frames.
|
||||
- Keep motion-blur shutter consistent across the ramp — if the source has motion blur, the slow section keeps it; if synthetic, hold a constant shutter angle (e.g. 180°).
|
||||
- Land the slowest frame *on the beat*; the ramp is felt, the impact is heard.
|
||||
|
||||
In AE: enable time remapping, keyframe the source time, ease the keyframes, set frame blending to Pixel Motion. In Remotion: drive a remapped source time via eased `interpolate` and keep effects deterministic.
|
||||
|
||||
## Sequence arc templates
|
||||
|
||||
```
|
||||
30s social spot (≈ 900 frames @30fps):
|
||||
Establish 0–120 2 shots, ~2s each (longer holds)
|
||||
Develop 120–540 shots shortening 1.5s → 0.8s, rising energy
|
||||
Climax 540–720 fastest cuts 0.3–0.5s, the drop at ~600
|
||||
Resolve 720–900 one long hold, logo/CTA breathes
|
||||
```
|
||||
|
||||
Plot average shot length as a curve: high at the start, descending to the climax, then a single long shot to release. Energy is the *derivative* of cut frequency — accelerate cuts to build, decelerate to land.
|
||||
|
||||
## NLE timeline vs code timeline
|
||||
|
||||
| Concept | NLE | Remotion | After Effects |
|
||||
|---|---|---|---|
|
||||
| Clip placement | drag to track at TC | `<Sequence from durationInFrames>` | layer in/out points |
|
||||
| Back-to-back shots | adjacent clips | `<Series>` | trimmed layers |
|
||||
| Crossfade | overlap + dissolve | overlap `Series` + opacity interp | opacity keyframes / overlap |
|
||||
| Speed ramp | time-remap keyframes | eased source-time `interpolate` | time remap + frame blending |
|
||||
| J/L cut | unlink + trim audio | `<Audio from>` ≠ picture seq | offset audio layer |
|
||||
| Beat marker | timeline markers | beat array `frame/fps` | comp markers / expression |
|
||||
|
||||
To port an edit plan: express each cut as a frame number (from the BPM math), each clip as `from`/`duration`, each transition as an overlap, and each retime as an eased time-remap. The plan is renderer-agnostic; only the placement syntax differs.
|
||||
|
||||
## Example edit plan (portable)
|
||||
|
||||
```
|
||||
Track: 128 BPM, 30fps → 14.06 f/beat, drop at beat 32 (frame 450)
|
||||
Cut on 4-beat phrases (≈56 frames):
|
||||
f0 ShotA establish, hold
|
||||
f56 ShotB cut on action (subject turns)
|
||||
f112 ShotC match cut from B (shared circle)
|
||||
f168 ... cuts shorten toward climax
|
||||
f450 DROP biggest visual; speed ramp 100→25→100 around f450
|
||||
f700 CTA long hold, L-cut audio trails in
|
||||
```
|
||||
|
||||
---
|
||||
Cut on the beat and ramp with purpose and the edit moves with the music. Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=motion-design-skills&utm_content=skill_footer&utm_term=beat-sync-editing)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -1,24 +0,0 @@
|
||||
# Caption Animation
|
||||
|
||||
Turn a transcript or voiceover into word-timed, scroll-stopping captions: each word pops in on the syllable, the active word highlights, and the text stays glued to the narration. Build the kind of captions that drive watch-time on TikTok, Reels, and Shorts — readable, on-beat, and inside the safe area. For short-form creators burning open captions onto vertical video.
|
||||
|
||||
## When it activates
|
||||
|
||||
- "Add animated captions," "make TikTok captions," or "karaoke captions."
|
||||
- "Word-by-word subtitles," "auto subtitles," or "highlight the active word."
|
||||
- "Burn in captions to a video" or "sync captions to a voiceover/narration."
|
||||
- "Turn an SRT/transcript into animated text."
|
||||
|
||||
## Example prompts
|
||||
|
||||
- "Turn this Whisper transcript into karaoke captions with a per-word pop."
|
||||
- "Restyle these auto subtitles into a branded, Hormozi-style reveal."
|
||||
- "Burn word-level captions onto my voiceover and export a sidecar SRT."
|
||||
|
||||
## What's inside
|
||||
|
||||
- `SKILL.md` — the transcribe → normalize → page → animate → place → export pipeline.
|
||||
- `references/word-timed-captions.md` — deeper guidance on word-level timing and animation.
|
||||
|
||||
---
|
||||
Part of **[Short-Form Video Skills](../../)** · Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=tiktok-video-skills&utm_content=skill_footer&utm_term=caption-animation)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -1,177 +0,0 @@
|
||||
---
|
||||
name: caption-animation
|
||||
description: This skill should be used when the user asks to "add animated captions", "make TikTok captions", "karaoke captions", "word-by-word subtitles", "auto subtitles", "highlight the active word", "burn in captions to a video", "sync captions to a voiceover/narration", or "turn an SRT/transcript into animated text". Covers word-level timing from Whisper, per-word pop/scale-in, active-word highlight, 9:16 safe-area placement, readable type, and burn-in vs sidecar SRT/VTT.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Caption Animation
|
||||
|
||||
Turn a transcript or voiceover into word-timed, scroll-stopping captions: each word pops in on the syllable, the active word highlights, and the text stays glued to the narration. Build the kind of captions that drive watch-time on TikTok, Reels, and Shorts — readable, on-beat, and inside the safe area.
|
||||
|
||||
## When to use
|
||||
|
||||
- Word-by-word or karaoke captions on short-form vertical video (9:16).
|
||||
- Burning open captions onto a voiceover/narration track from a transcript.
|
||||
- Turning a Whisper/SRT transcript into animated text in Remotion or on the web.
|
||||
- Restyling auto-generated subtitles into a branded, "Hormozi-style" reveal.
|
||||
|
||||
## The pipeline
|
||||
|
||||
| Stage | Job | Tool |
|
||||
|---|---|---|
|
||||
| Transcribe | Audio → word-level timestamps | Whisper (`@remotion/install-whisper-cpp`), AssemblyAI |
|
||||
| Normalize | → `{ text, startMs, endMs }[]` tokens | `@remotion/captions` `Caption` type |
|
||||
| Page | Group words into 1–4 word "pages" | `createTikTokStyleCaptions()` |
|
||||
| Animate | Per-word pop + active-word highlight | Remotion `spring()` / CSS |
|
||||
| Place | Safe-area, readable type | layout rules below |
|
||||
| Export | Burn-in (MP4) and/or sidecar SRT/VTT | Remotion render / file emit |
|
||||
|
||||
## Two non-negotiable rules
|
||||
|
||||
1. **Word-level timing, not line-level.** Karaoke reads as magic only when each word lands on the syllable. Always transcribe to word timestamps; never fake them by splitting a line evenly over its duration — drift is instantly visible.
|
||||
2. **Readability beats style.** Captions are read on a phone, in sunlight, muted. Bold sans-serif, heavy stroke or shadow, high contrast first; decoration second. A caption nobody can read is decoration, not a caption.
|
||||
|
||||
## Word timing from a transcript
|
||||
|
||||
Whisper emits word-level timestamps directly. Normalize every source into one flat token shape so the renderer never cares where the words came from:
|
||||
|
||||
```ts
|
||||
// Caption token — the one shape everything downstream consumes
|
||||
type Token = { text: string; startMs: number; endMs: number };
|
||||
|
||||
// Parse an SRT cue block "00:00:01,200 --> 00:00:01,640" into ms
|
||||
const toMs = (t: string) => {
|
||||
const [h, m, rest] = t.split(":");
|
||||
const [s, ms] = rest.split(",");
|
||||
return ((+h * 60 + +m) * 60 + +s) * 1000 + +ms;
|
||||
};
|
||||
```
|
||||
|
||||
If only sentence-level cues exist, re-transcribe for word timing — do not interpolate. Whisper's `medium.en` is ~2x faster than `large` and accurate enough for clean voice audio.
|
||||
|
||||
## Per-word pop-in (Remotion)
|
||||
|
||||
Drive each word's scale and opacity off a spring anchored to its own `startMs`. The micro-overshoot is what makes it feel "alive".
|
||||
|
||||
```tsx
|
||||
import { useCurrentFrame, useVideoConfig, spring, interpolate } from "remotion";
|
||||
|
||||
const Word: React.FC<{ token: Token; active: boolean }> = ({ token, active }) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps } = useVideoConfig();
|
||||
const enter = (token.startMs / 1000) * fps; // word's own entrance frame
|
||||
const p = spring({ frame: frame - enter, fps, config: { damping: 12, mass: 0.6 } });
|
||||
const scale = interpolate(p, [0, 1], [0.6, 1]); // pop from 60% → 100% (overshoots)
|
||||
return (
|
||||
<span style={{
|
||||
display: "inline-block",
|
||||
transform: `scale(${scale})`,
|
||||
opacity: interpolate(p, [0, 1], [0, 1]),
|
||||
color: active ? "#FFE45E" : "#FFFFFF", // active-word highlight
|
||||
transition: "color 80ms linear",
|
||||
}}>{token.text} </span>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
Compute `active` by testing `frame` against each token's `[startMs, endMs]`. Keep the highlight a single accent color or a filled "box" behind the live word — never animate every word's color at once.
|
||||
|
||||
## Pages, not walls of text
|
||||
|
||||
Show 1–4 words at a time. Use Remotion's `createTikTokStyleCaptions()` (or group manually) and tune `combineTokensWithinMilliseconds`: ~200–500ms for true word-by-word, ~1000–1500ms for short readable phrases.
|
||||
|
||||
```ts
|
||||
import { createTikTokStyleCaptions } from "@remotion/captions";
|
||||
const { pages } = createTikTokStyleCaptions({
|
||||
captions, combineTokensWithinMilliseconds: 1200, // ~2–3 words per page
|
||||
});
|
||||
```
|
||||
|
||||
## Web/CSS pop (no Remotion)
|
||||
|
||||
For a DOM/live player, give each word its own `animation-delay` equal to `startMs`:
|
||||
|
||||
```css
|
||||
.word { display:inline-block; opacity:0; animation: pop .26s cubic-bezier(.34,1.56,.64,1) forwards; }
|
||||
@keyframes pop { from { opacity:0; transform:translateY(.18em) scale(.7) } to { opacity:1; transform:none } }
|
||||
.word.active { color:#FFE45E; }
|
||||
```
|
||||
```js
|
||||
words.forEach(w => { const el = mk(w.text); el.style.animationDelay = `${w.startMs}ms`; track.append(el); });
|
||||
```
|
||||
|
||||
## Readable type and safe-area placement (9:16, 1080×1920)
|
||||
|
||||
| Property | Value | Why |
|
||||
|---|---|---|
|
||||
| Font | Bold/ExtraBold sans (Montserrat, Inter, Helvetica) | Reads on small, busy screens |
|
||||
| Size | 56–80px (≈8% of frame height), min 45px | Legible muted on a phone |
|
||||
| Stroke | 2–6px solid black outline | Survives any background |
|
||||
| Shadow | Soft drop shadow as backup to stroke | Separation on bright frames |
|
||||
| Case | Uppercase or sentence; high contrast fill | Punch + scannability |
|
||||
| Vertical pos | Center band, ~62–70% down | Above the UI, below the action |
|
||||
| Bottom safe | Keep clear of bottom ~280px / 15% | Avoids caption/CTA/audio UI |
|
||||
| Side safe | Keep within center 80% width | Avoids right-rail icons |
|
||||
|
||||
Place captions in the center band — not pinned to the very bottom, where the platform UI (username, audio, buttons) lives. Pad text to a max of ~2 lines.
|
||||
|
||||
## Sync to a voiceover / narration
|
||||
|
||||
The transcript already carries the timing of the exact audio. Lay that same audio under the composition and keep timestamps in the audio's timebase — captions and voice stay locked with zero manual nudging. If the VO is re-recorded, re-transcribe; never hand-shift offsets.
|
||||
|
||||
## Burn-in vs sidecar SRT/VTT
|
||||
|
||||
| | Burn-in (open) | Sidecar SRT/VTT (closed) |
|
||||
|---|---|---|
|
||||
| Where | Pixels in the MP4 | Separate `.srt`/`.vtt` file |
|
||||
| Social (TikTok/Reels) | Required — guaranteed, styleable | Often ignored by the platform |
|
||||
| Accessibility (ADA/WCAG) | Does not satisfy alone | Required (toggleable) |
|
||||
| Best practice | Burn animated captions for social | Also ship a sidecar for the web/SEO |
|
||||
|
||||
Emit both: burn the animated captions for social, and write a plain SRT/VTT from the same tokens for accessible/SEO playback.
|
||||
|
||||
## Output checklist
|
||||
|
||||
- Word-level timestamps (real, from transcription) — no even-split fakery.
|
||||
- Per-word pop with micro-overshoot; one accent color for the active word.
|
||||
- 1–4 words per page; never a wall of text.
|
||||
- Bold sans, 56–80px, 2–6px stroke + shadow, high contrast.
|
||||
- Center band, clear of bottom ~280px and side rails.
|
||||
- Audio laid under composition; timestamps in the audio timebase.
|
||||
- Ship burn-in for social and a sidecar SRT/VTT for accessibility.
|
||||
|
||||
## Deliver & verify (rendered stills → MP4)
|
||||
|
||||
> **Packaged helper** (`scripts/`): tile your stills with `scripts/contact-sheet.sh sheet.png f-hook.png f-mid.png f-end.png`, then assert the encode with `scripts/probe-mp4.sh out.mp4 [WxH] [fps]`. See `scripts/README.md`.
|
||||
|
||||
Captions ship as a Remotion composition (`<Composition>` + zod `schema` + `defaultProps`) — all word motion frame-driven off `useCurrentFrame()`, never `Date.now()` / `Math.random()` / timers. Deliverable = `out/*.mp4` (burned-in) + the project + the sidecar SRT/VTT. 9:16 vertical (1080×1920) is the default.
|
||||
|
||||
**Verify loop — render stills → inspect → encode.** Word timing is the thing that breaks; check it at exact frames before you spend an encode.
|
||||
|
||||
```bash
|
||||
# Stills at start / a sampled active-word frame / end — WITH SHIPPED PROPS (real tokens + audio)
|
||||
npx remotion still Captions out/f-start.png --frame=0 --props='{"captionsSrc":"vo.json"}'
|
||||
npx remotion still Captions out/f-mid.png --frame=90 --props='{"captionsSrc":"vo.json"}'
|
||||
npx remotion still Captions out/f-end.png --frame=N --props='{"captionsSrc":"vo.json"}' # N = durationInFrames-1
|
||||
|
||||
# Inspect each PNG:
|
||||
# - the word highlighted at frame 90 is the word whose [startMs,endMs] contains 90/fps (no drift)
|
||||
# - burn-in legible: bold sans, stroke+shadow holds, no clipping
|
||||
# - 9:16 safe area: caption sits in the center band, clear of top ~12% and bottom ~20-35% (captions/CTA/audio UI) and the right action rail
|
||||
|
||||
npx remotion render Captions out/captions.mp4 --props='{"captionsSrc":"vo.json"}' # encode once stills are right
|
||||
npx remotion render Captions out/demo.gif --codec=gif # README proof clip
|
||||
```
|
||||
|
||||
Use `npx remotion compositions` to read `durationInFrames`/`fps` and pick the active-word + end frames.
|
||||
|
||||
**Before you finish:**
|
||||
1. Stills render cleanly at frame 0, a mid active-word frame, and last — no missing font/audio.
|
||||
2. The correct word is highlighted at the sampled frame (frame/fps lands inside its token); no even-split fakery.
|
||||
3. Burn-in is legible (stroke+shadow) and the caption is fully inside the 9:16 safe area at every checked frame.
|
||||
4. Frame-driven only — no `Date.now()` / `Math.random()` / timers.
|
||||
5. Shipped props (real tokens, not just `defaultProps`) render correctly; MP4 + sidecar SRT/VTT emitted, GIF optional.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/word-timed-captions.md` — end-to-end build: Whisper transcription and the `Caption` type, a full SRT parser, Remotion word-timed component with active-highlight, manual paging, an SRT/VTT emitter, per-platform safe-area maps, and a readable-type spec sheet.
|
||||
@@ -1,254 +0,0 @@
|
||||
# Word-Timed Captions — End-to-End Build
|
||||
|
||||
A complete, runnable path from audio/transcript to animated burn-in captions plus a sidecar SRT/VTT. Code targets Remotion for burn-in and plain JS for the web; the data shapes are framework-agnostic.
|
||||
|
||||
## 1. The one data shape
|
||||
|
||||
Everything normalizes to a flat array of word tokens. Get to this shape and the renderer never cares about the source.
|
||||
|
||||
```ts
|
||||
export type Token = { text: string; startMs: number; endMs: number };
|
||||
export type Page = { startMs: number; endMs: number; tokens: Token[] };
|
||||
```
|
||||
|
||||
Remotion's `@remotion/captions` uses a compatible `Caption` type:
|
||||
|
||||
```ts
|
||||
type Caption = { text: string; startMs: number; endMs: number; timestampMs: number | null; confidence: number | null };
|
||||
```
|
||||
|
||||
## 2. Transcribe to word-level timing (Whisper)
|
||||
|
||||
Word timestamps are non-negotiable for karaoke. Whisper produces them natively.
|
||||
|
||||
```bash
|
||||
npx @remotion/install-whisper-cpp # one-time: installs whisper.cpp + a model
|
||||
```
|
||||
|
||||
```ts
|
||||
import { transcribe, toCaptions } from "@remotion/install-whisper-cpp";
|
||||
|
||||
const { transcription } = await transcribe({
|
||||
inputPath: "voiceover.wav", // 16kHz mono WAV is ideal for whisper.cpp
|
||||
model: "medium.en", // medium.en ≈ 2x faster than large, accurate on clean voice
|
||||
tokenLevelTimestamps: true, // <-- the key flag: per-word, not per-line
|
||||
whisperPath: "./whisper.cpp",
|
||||
});
|
||||
const { captions } = toCaptions({ transcription }); // → Caption[]
|
||||
```
|
||||
|
||||
Tips:
|
||||
- Feed clean voice audio (denoise/normalize first). Background music wrecks word boundaries.
|
||||
- `large` is most accurate but 2–3x real-time on a laptop. Use `medium.en` unless accuracy is critical.
|
||||
- Cloud alternatives (AssemblyAI, Deepgram, OpenAI Whisper API) also return word timing — map their `words[]` into `Token` the same way.
|
||||
|
||||
Map any token shape into `Token`:
|
||||
|
||||
```ts
|
||||
const toTokens = (caps: Caption[]): Token[] =>
|
||||
caps.map(c => ({ text: c.text.trim(), startMs: c.startMs, endMs: c.endMs }));
|
||||
```
|
||||
|
||||
## 3. Parse an existing SRT into tokens
|
||||
|
||||
When handed an SRT, parse cues; if cues are sentence-level, re-transcribe for word timing rather than interpolating.
|
||||
|
||||
```ts
|
||||
const toMs = (t: string) => {
|
||||
const [h, m, rest] = t.trim().split(":");
|
||||
const [s, ms] = rest.split(",");
|
||||
return ((+h * 60 + +m) * 60 + +s) * 1000 + +ms;
|
||||
};
|
||||
|
||||
export function parseSrt(src: string): Token[] {
|
||||
return src.trim().split(/\n\s*\n/).map(block => {
|
||||
const lines = block.split("\n");
|
||||
const time = lines.find(l => l.includes("-->"))!;
|
||||
const [a, b] = time.split("-->");
|
||||
const text = lines.slice(lines.indexOf(time) + 1).join(" ").trim();
|
||||
return { text, startMs: toMs(a), endMs: toMs(b) };
|
||||
});
|
||||
}
|
||||
```
|
||||
|
||||
## 4. Page the tokens (1–4 words at a time)
|
||||
|
||||
Either use Remotion's helper or group manually. Manual grouping by gap + max-words keeps full control:
|
||||
|
||||
```ts
|
||||
export function paginate(tokens: Token[], maxWords = 3, maxGapMs = 1200): Page[] {
|
||||
const pages: Page[] = [];
|
||||
let cur: Token[] = [];
|
||||
for (const t of tokens) {
|
||||
const span = cur.length ? t.endMs - cur[0].startMs : 0;
|
||||
if (cur.length >= maxWords || span > maxGapMs) {
|
||||
if (cur.length) pages.push(pageOf(cur)); cur = [];
|
||||
}
|
||||
cur.push(t);
|
||||
}
|
||||
if (cur.length) pages.push(pageOf(cur));
|
||||
return pages;
|
||||
}
|
||||
const pageOf = (ts: Token[]): Page => ({ startMs: ts[0].startMs, endMs: ts.at(-1)!.endMs, tokens: ts });
|
||||
```
|
||||
|
||||
Remotion equivalent:
|
||||
|
||||
```ts
|
||||
import { createTikTokStyleCaptions } from "@remotion/captions";
|
||||
const { pages } = createTikTokStyleCaptions({ captions, combineTokensWithinMilliseconds: 1200 });
|
||||
// 200–500ms → word-by-word; 1000–1500ms → short readable phrases
|
||||
```
|
||||
|
||||
## 5. Remotion component: word-timed with active highlight
|
||||
|
||||
```tsx
|
||||
import { AbsoluteFill, Audio, Sequence, useCurrentFrame, useVideoConfig,
|
||||
spring, interpolate, staticFile } from "remotion";
|
||||
|
||||
const CAPTION_STYLE: React.CSSProperties = {
|
||||
fontFamily: "Montserrat, Inter, sans-serif",
|
||||
fontWeight: 800,
|
||||
fontSize: 72, // ~8% of 1920px height
|
||||
color: "#fff",
|
||||
textAlign: "center",
|
||||
textTransform: "uppercase",
|
||||
WebkitTextStroke: "5px #000", // heavy outline survives any background
|
||||
paintOrder: "stroke fill", // stroke behind fill, not clipping glyphs
|
||||
textShadow: "0 4px 16px rgba(0,0,0,.55)",
|
||||
lineHeight: 1.1,
|
||||
padding: "0 80px", // keep inside center 80% width
|
||||
};
|
||||
|
||||
const Word: React.FC<{ token: Token; active: boolean }> = ({ token, active }) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps } = useVideoConfig();
|
||||
const enter = (token.startMs / 1000) * fps;
|
||||
const p = spring({ frame: frame - enter, fps, config: { damping: 12, mass: 0.6 } });
|
||||
return (
|
||||
<span style={{
|
||||
display: "inline-block",
|
||||
transform: `scale(${interpolate(p, [0, 1], [0.6, 1])})`,
|
||||
opacity: interpolate(p, [0, 1], [0, 1]),
|
||||
color: active ? "#FFE45E" : undefined,
|
||||
transition: "color 80ms linear",
|
||||
}}>{token.text} </span>
|
||||
);
|
||||
};
|
||||
|
||||
export const Captions: React.FC<{ pages: Page[]; audio: string }> = ({ pages, audio }) => {
|
||||
const { fps } = useVideoConfig();
|
||||
const f = (ms: number) => Math.round((ms / 1000) * fps);
|
||||
const frame = useCurrentFrame();
|
||||
return (
|
||||
<AbsoluteFill>
|
||||
<Audio src={staticFile(audio)} /> {/* same audio the transcript came from */}
|
||||
{pages.map((pg, i) => (
|
||||
<Sequence key={i} from={f(pg.startMs)} durationInFrames={Math.max(1, f(pg.endMs - pg.startMs))}>
|
||||
<AbsoluteFill style={{ justifyContent: "flex-start", alignItems: "center", paddingTop: "62%" }}>
|
||||
<div style={CAPTION_STYLE}>
|
||||
{pg.tokens.map((t, j) => {
|
||||
const now = (frame / fps) * 1000;
|
||||
return <Word key={j} token={t} active={now >= t.startMs && now < t.endMs} />;
|
||||
})}
|
||||
</div>
|
||||
</AbsoluteFill>
|
||||
</Sequence>
|
||||
))}
|
||||
</AbsoluteFill>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
Because the `<Audio>` and the timestamps share one timebase, voice and captions stay locked with no manual nudging. Re-record the VO → re-transcribe → done.
|
||||
|
||||
## 6. Web/CSS path (live DOM player)
|
||||
|
||||
```js
|
||||
function renderTrack(tokens, mountEl) {
|
||||
tokens.forEach(t => {
|
||||
const el = document.createElement("span");
|
||||
el.className = "word";
|
||||
el.textContent = t.text + " ";
|
||||
el.style.animationDelay = `${t.startMs}ms`; // each word pops on its own start
|
||||
mountEl.append(el);
|
||||
});
|
||||
}
|
||||
// drive .active off the media element's currentTime
|
||||
audio.addEventListener("timeupdate", () => {
|
||||
const ms = audio.currentTime * 1000;
|
||||
[...track.children].forEach((el, i) =>
|
||||
el.classList.toggle("active", ms >= tokens[i].startMs && ms < tokens[i].endMs));
|
||||
});
|
||||
```
|
||||
|
||||
```css
|
||||
.cap { font: 800 72px/1.1 Montserrat, sans-serif; text-transform: uppercase;
|
||||
color:#fff; -webkit-text-stroke:5px #000; paint-order:stroke fill;
|
||||
text-shadow:0 4px 16px rgba(0,0,0,.55); text-align:center; }
|
||||
.word { display:inline-block; opacity:0; animation: pop .26s cubic-bezier(.34,1.56,.64,1) forwards; }
|
||||
.word.active { color:#FFE45E; }
|
||||
@keyframes pop { from { opacity:0; transform:translateY(.18em) scale(.7) } to { opacity:1; transform:none } }
|
||||
```
|
||||
|
||||
## 7. Emit a sidecar SRT / VTT from the same tokens
|
||||
|
||||
Burn-in alone does not satisfy ADA/WCAG and many platforms ignore uploaded tracks — so ship both. Group tokens into readable cues and serialize:
|
||||
|
||||
```ts
|
||||
const pad = (n: number, w = 2) => String(n).padStart(w, "0");
|
||||
const stamp = (ms: number, sep: string) => {
|
||||
const h = Math.floor(ms / 3.6e6), m = Math.floor(ms % 3.6e6 / 6e4),
|
||||
s = Math.floor(ms % 6e4 / 1000), x = ms % 1000;
|
||||
return `${pad(h)}:${pad(m)}:${pad(s)}${sep}${pad(x, 3)}`;
|
||||
};
|
||||
|
||||
export const toSrt = (pages: Page[]) => pages.map((p, i) =>
|
||||
`${i + 1}\n${stamp(p.startMs, ",")} --> ${stamp(p.endMs, ",")}\n${p.tokens.map(t => t.text).join(" ")}`
|
||||
).join("\n\n") + "\n";
|
||||
|
||||
export const toVtt = (pages: Page[]) => "WEBVTT\n\n" + pages.map(p =>
|
||||
`${stamp(p.startMs, ".")} --> ${stamp(p.endMs, ".")}\n${p.tokens.map(t => t.text).join(" ")}`
|
||||
).join("\n\n") + "\n";
|
||||
```
|
||||
|
||||
Use SRT for the broadest platform support; VTT when styling/positioning cues or serving on the web.
|
||||
|
||||
## 8. Per-platform safe areas (1080×1920, 9:16)
|
||||
|
||||
| Platform | Bottom keep-clear | Top keep-clear | Side keep-clear | Notes |
|
||||
|---|---|---|---|---|
|
||||
| TikTok | ~340px | ~120px | ~110px right rail | Captions, audio title, action buttons crowd the bottom-right |
|
||||
| Instagram Reels | ~280px | ~130px | ~60px | Username/CTA bar at bottom |
|
||||
| YouTube Shorts | ~240px | ~100px | ~120px right rail | Title + channel at bottom-left |
|
||||
| Facebook Reels | ~300px | ~130px | ~80px | Similar to Reels |
|
||||
|
||||
Safe rule across all: place caption baseline at ~62–70% of frame height, keep text within the center 80% width, and never let it enter the bottom ~15%.
|
||||
|
||||
## 9. Readable-type spec sheet
|
||||
|
||||
| Property | Recommended | Floor |
|
||||
|---|---|---|
|
||||
| Weight | 800 (ExtraBold) | 700 |
|
||||
| Size @1080p | 56–80px | 45px |
|
||||
| Stroke | 4–6px solid #000 | 2px |
|
||||
| Shadow | `0 4px 16px rgba(0,0,0,.55)` | soft 1px |
|
||||
| Active accent | one color (e.g. #FFE45E) or filled box | — |
|
||||
| Lines on screen | 1–2 | 2 max |
|
||||
| Words per page | 1–4 | 5 max |
|
||||
|
||||
Set `paint-order: stroke fill` so a thick stroke sits behind the glyph instead of eating it. Test on a bright frame and a busy frame; if it survives both, it ships.
|
||||
|
||||
## 10. Render
|
||||
|
||||
```bash
|
||||
npx remotion render src/index.ts Captions out/captions.mp4 \
|
||||
--props='{"pages": ..., "audio": "voiceover.wav"}'
|
||||
```
|
||||
|
||||
A 10-minute video with animated captions can take 30–60 min to render locally; render long-form in segments, or render on Remotion Lambda for parallel speed.
|
||||
|
||||
---
|
||||
## Built by the team behind iart.ai
|
||||
|
||||
This skill is part of an open motion-graphics collection from iart.ai — the AI motion agent that turns data, scripts, and designs into editable motion graphics (Remotion → MP4). If you'd rather not hand-build this, iart.ai can auto-time captions to a narration track and one-click burn them in from a template — change the text/audio and re-export. → [iart.ai](https://iart.ai/?utm_source=github&utm_medium=reference&utm_campaign=tiktok-video-skills&utm_content=ref_footer&utm_term=caption-animation)
|
||||
Whitespace-only changes.
@@ -0,0 +1,175 @@
|
||||
---
|
||||
name: color-motion
|
||||
description: This skill should be used when the user asks to "build a color palette for a motion piece", "make a gradient background", "animate a color transition", "interpolate colors smoothly", "fix muddy/gray gradients", "convert colors to OKLCH", "color-grade a video for mood", "fix washed-out After Effects renders", "gamma shift", "Rec.709 vs sRGB", "colors shifted after rendering", or "match preview to final video". Covers palette construction, gradients, perceptual interpolation (OKLCH/Lab), grading order, and video/render color management.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Color in Motion
|
||||
|
||||
Choose, pair, and animate color for motion design: building restrained palettes, constructing gradients that read as premium, interpolating colors in perceptually-uniform spaces, and grading a piece for mood and consistency. Produces palettes with exact OKLCH/hex values, gradient CSS/code, and transition parameters.
|
||||
|
||||
## When to use
|
||||
|
||||
- Picking or pairing a palette for a motion or video piece.
|
||||
- Building gradient backgrounds (mesh, noise, animated drift).
|
||||
- Animating color transitions (button states, theme switch, loaders).
|
||||
- Grading a video for mood, or fixing dull/gray gradient interpolation.
|
||||
- Fixing washed-out, gamma-shifted, or platform-inconsistent After Effects renders, and choosing correct per-platform export/color tags.
|
||||
|
||||
## Core techniques
|
||||
|
||||
### 1. Build a restrained palette
|
||||
|
||||
Limit hues. A strong motion palette is **1 primary + 1 accent + 2-3 neutrals**. Derive contrast and depth from lightness and chroma, not from adding more hues. Extra hues fight for attention and read as amateur.
|
||||
|
||||
In dark scenes, make the accent pop with **low-to-moderate chroma and high lightness** rather than maximum saturation — a blown-out saturated accent vibrates and looks cheap on dark backgrounds.
|
||||
|
||||
Use OKLCH to reason about color: `oklch(L C H)` where L is perceptual lightness (0-1), C is chroma (0 ≈ 0.37 max), H is hue angle (0-360). Holding L and C constant while rotating H gives hues of genuinely equal visual weight — impossible to do reliably in HSL.
|
||||
|
||||
```
|
||||
Primary: oklch(0.62 0.19 264) /* confident blue */
|
||||
Accent: oklch(0.78 0.17 50) /* warm amber, lighter so it pops */
|
||||
Neutral-0: oklch(0.97 0.005 264) /* near-white, faint hue tint */
|
||||
Neutral-1: oklch(0.55 0.01 264)
|
||||
Neutral-2: oklch(0.18 0.015 264) /* near-black, same hue family */
|
||||
```
|
||||
|
||||
Tinting neutrals slightly toward the primary hue (tiny chroma) unifies the frame; pure gray neutrals next to colored elements look disconnected.
|
||||
|
||||
### 2. Interpolate in OKLCH or Lab — never raw sRGB
|
||||
|
||||
Linearly interpolating two colors in sRGB passes through a desaturated, often gray or muddy midpoint (blue->yellow goes through gray; blue->red through muddy purple). Perceptual spaces (OKLCH, OKLab, CIE Lab) keep midpoints vivid and lightness even.
|
||||
|
||||
CSS gradients support this natively:
|
||||
|
||||
```css
|
||||
/* Vivid, even midpoint */
|
||||
background: linear-gradient(in oklch, #2b6cff, #ffb02e);
|
||||
|
||||
/* Control hue path around the wheel for two-stop gradients */
|
||||
background: linear-gradient(in oklch longer hue, #2b6cff, #ffb02e);
|
||||
background: conic-gradient(in oklch, red, blue, red);
|
||||
```
|
||||
|
||||
For animated transitions interpolate with a library (see references) and emit per-frame colors, or for two-color CSS transitions register a typed custom property so the browser interpolates in the chosen space.
|
||||
|
||||
### 3. Gradients that read as premium
|
||||
|
||||
- **Mesh / multi-radial**: stack several `radial-gradient`s at different positions over a base color. Soft, organic, expensive-looking.
|
||||
- **Add grain**: a faint noise overlay (SVG `feTurbulence` or a tiled PNG at ~3-5% opacity) kills banding and adds texture.
|
||||
- **Animate slowly**: drift gradient positions over 8-20s with `ease-in-out alternate`. Slow is premium; fast gradient motion looks like a screensaver.
|
||||
|
||||
```css
|
||||
.mesh {
|
||||
background-color: #0b0f1a;
|
||||
background-image:
|
||||
radial-gradient(at 20% 25%, oklch(0.62 0.19 264 / 0.7), transparent 50%),
|
||||
radial-gradient(at 80% 30%, oklch(0.70 0.16 320 / 0.6), transparent 55%),
|
||||
radial-gradient(at 50% 85%, oklch(0.78 0.17 50 / 0.5), transparent 50%);
|
||||
animation: drift 16s ease-in-out infinite alternate;
|
||||
}
|
||||
@keyframes drift {
|
||||
to { background-position: 8% -6%, -8% 6%, 4% 8%; }
|
||||
}
|
||||
```
|
||||
|
||||
### 4. Color transition animation
|
||||
|
||||
Inline timing essentials so this skill stands alone: enter/highlight with ease-out (`cubic-bezier(0.16, 1, 0.3, 1)`), state changes 150-300ms, theme/full-screen washes 400-800ms. Use `linear` only for continuous loops.
|
||||
|
||||
```css
|
||||
@property --c { syntax: "<color>"; initial-value: #2b6cff; inherits: false; }
|
||||
.btn { background: var(--c); transition: --c 200ms cubic-bezier(0.16,1,0.3,1); }
|
||||
.btn:hover { --c: #ffb02e; }
|
||||
```
|
||||
|
||||
`@property` registration is what makes the color actually interpolate (unregistered custom properties jump instantly).
|
||||
|
||||
### 5. Grade for mood — correct order
|
||||
|
||||
Apply grading operations in this order; reordering changes the result:
|
||||
|
||||
1. **Exposure / white balance** — fix overall brightness and neutralize/intentionally set the cast first.
|
||||
2. **Contrast** — set black and white points; expand or compress the tonal range.
|
||||
3. **Midtones / saturation** — adjust overall saturation and midtone color.
|
||||
4. **Split-tone** — tint shadows and highlights separately. The classic cinematic look is **teal shadows + orange highlights** (complementary, flatters skin). Keep it subtle.
|
||||
5. **Unify** — a final overall tint/LUT and a subtle vignette to bind the frame.
|
||||
|
||||
In code, grade by chaining CSS filters or a fragment shader; for video, an actual LUT or grading panel. Always grade the whole piece consistently so shots cut together.
|
||||
|
||||
### 6. Video / render color management
|
||||
|
||||
When a render looks correct in the After Effects comp viewer but washed-out, too bright, milky, or gamma-shifted in QuickTime, a browser, or across macOS/Windows, the cause is a gamma/transfer mismatch across four stages — AE working space, the render's embedded tag, the player's interpretation, and the display/OS gamma — not the artwork.
|
||||
|
||||
Core rules:
|
||||
|
||||
- **Set an explicit working space.** File > Project Settings > Color tab: Working Space = **Rec.709 Gamma 2.4** for SDR video (or **sRGB** for UI/stills); "None" disables color management and the viewer lies about the final look. Set Depth to **16 bpc** for grades/gradients to avoid banding.
|
||||
- **Rec.709 vs sRGB.** They share primaries but differ in transfer (sRGB ~2.2 piecewise vs Rec.709 display 2.4 / BT.1886). Using the wrong one shifts midtone brightness — Rec.709 for video, sRGB for screenshots/PNG/UI.
|
||||
- **Embed a color tag on export.** The shift across macOS/Windows/players comes from untagged files; each player guesses its own default. Tag Rec.709 (or sRGB) explicitly via Render Queue > **Output Module > Color Management** (do not leave "Preserve RGB" when a shift occurs).
|
||||
- **The QuickTime gamma bug.** QuickTime/AVFoundation apply an extra ~1.96 gamma, so a file can look fine in QuickTime but dark/contrasty in Chrome. **Never grade or approve to QuickTime** — Chrome, Premiere, or Resolve is the correct reference for web delivery.
|
||||
- **Match preview to final.** Toggle "Use Display Color Management" in the comp viewer while reviewing, and verify the export numerically: place a 50% grey (RGB 128) patch, render, re-sample — a value that rises (128 → 150+) confirms a washed-out gamma shift. Confirm the embedded tag with `ffprobe` (`color_primaries`/`color_transfer`/`color_space` = bt709).
|
||||
|
||||
Per-platform export at a glance: **ProRes 422 HQ/4444** tagged Rec.709 for editing masters; **H.264 (MP4)** tagged Rec.709 for web/`<video>`/social (verify in Chrome); ProRes 4444 + alpha → PNG (sRGB) for transparency; ProRes 4444 / HEVC 10-bit Rec.2020 PQ/HLG for HDR (needs 32 bpc float).
|
||||
|
||||
See `references/color-management.md` for the full diagnosis decision tree, exact menu paths, per-platform export table, the QuickTime workarounds, per-OS/player gamma values, the numeric test-patch method, `ffprobe`/`ffmpeg` tag commands, limited-vs-full-range pitfalls, and the pre-export checklist.
|
||||
|
||||
## Deliver & verify (standalone HTML)
|
||||
|
||||
> **Packaged helper** (`scripts/`): `scripts/seek-shot.sh anim.html 0 1.5 3` freezes the `?t=N` harness and screenshots each moment; `scripts/contact-sheet.sh sheet.png frame-*.png` tiles them for one-glance review. See `scripts/README.md`.
|
||||
|
||||
For a web palette/gradient/color-transition deliverable, ship **one `.html` file that opens directly in a browser** — no build step. CSS handles the color work; the file is self-contained.
|
||||
|
||||
**Output contract:**
|
||||
- One `.html` file: your markup + the gradient/transition in inline `<style>` (CDN only if a JS interpolation lib is needed).
|
||||
- One animation driver — a CSS `@keyframes` drift or a single `@property` transition; nothing competing.
|
||||
- Include the freeze harness below so any moment can be screenshotted deterministically.
|
||||
|
||||
**Freeze harness — pin a CSS-animation frame for screenshots.** `?t=N` sets a negative `animation-delay` and pauses, so the gradient/transition lands on the exact frame at `N` seconds.
|
||||
|
||||
```html
|
||||
<script>
|
||||
const t = new URLSearchParams(location.search).get("t");
|
||||
if (t !== null) {
|
||||
document.querySelectorAll(".bg, .btn").forEach(el => { // animated elements
|
||||
el.style.animationDelay = (-parseFloat(t)) + "s";
|
||||
el.style.animationPlayState = "paused";
|
||||
});
|
||||
}
|
||||
window.__ready = true; // ready signal for headless wait
|
||||
</script>
|
||||
```
|
||||
|
||||
**Verify loop — render → freeze → screenshot → check:**
|
||||
1. Open the file frozen at start / mid / end: `…/color.html?t=0`, `?t=<period/2>`, `?t=<period>`.
|
||||
2. Screenshot each frozen frame.
|
||||
3. Check **fidelity** (matches the palette/brief) and **color artifacts** — at each frozen frame confirm the gradient/transition has **no banding and no muddy or gray midpoint** (the OKLCH/Lab path stays vivid). Add grain if banding appears.
|
||||
4. **Iterate:** if banding or a muddy/gray midpoint shows up, adjust the OKLCH/Lab chroma & lightness (or add grain), re-render the **same** `?t=N`, and re-screenshot to confirm the fix — loop until every frozen frame is clean.
|
||||
|
||||
```bash
|
||||
npx playwright screenshot --wait-for-timeout=500 "file://$PWD/color.html?t=8" frame-mid.png
|
||||
```
|
||||
|
||||
**Before you finish:**
|
||||
1. Opens standalone in a browser — no console errors, no missing CDN.
|
||||
2. One color driver; `?t=N` freezes the exact frame correctly.
|
||||
3. Screenshotted at start / mid / end — no banding, no muddy/gray midpoints, lightness even.
|
||||
4. `prefers-reduced-motion` honored (drift paused on a static, readable frame).
|
||||
5. Interpolation is in OKLCH/Lab (gradients/`@property`), never raw sRGB.
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Need | Do |
|
||||
|---|---|
|
||||
| Palette size | 1 primary + 1 accent + 2-3 neutrals |
|
||||
| Equal-weight hues | hold L,C in OKLCH, vary H |
|
||||
| Vivid gradient | `linear-gradient(in oklch, ...)` |
|
||||
| Smooth color anim | interpolate in OKLCH/Lab, register via `@property` |
|
||||
| Premium bg | stacked radials + grain + 8-20s slow drift |
|
||||
| Cinematic grade | teal shadows / orange highlights, subtle |
|
||||
| Fix washed-out AE render | tag Rec.709 in Output Module > Color Management; verify in Chrome |
|
||||
| Avoid | raw-sRGB interpolation (muddy midpoints); grading to QuickTime |
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/palettes-and-interpolation.md` — palette construction rules with OKLCH values and named recipes, gradient cookbook (mesh, conic, grain), culori and d3-interpolate code for perceptual interpolation and scales, and the full grading-order checklist with filter/shader examples.
|
||||
- `references/color-management.md` — full video/render color management: diagnosis decision tree, AE project/output color settings with exact menu paths, per-platform export table, QuickTime gamma-bug workarounds, per-OS/player gamma values, the numeric test-patch method, `ffprobe`/`ffmpeg` tag verification, limited-vs-full-range pitfalls, and the pre-export checklist.
|
||||
@@ -0,0 +1,165 @@
|
||||
# Video / render color management (After Effects)
|
||||
|
||||
How to diagnose and prevent the most common After Effects color complaint: a render that looks correct inside the AE composition viewer but appears washed-out, too bright, milky, or gamma-shifted when played in QuickTime Player, uploaded to a browser, or moved between macOS and Windows. The root cause is almost always a mismatch in how gamma / transfer functions are interpreted across AE, the codec metadata, the OS player, and the display pipeline — not a problem with the artwork.
|
||||
|
||||
## Mental model: where shifts come from
|
||||
|
||||
A pixel value travels through four stages, each of which can apply or assume a gamma. A shift is a mismatch between two of them:
|
||||
|
||||
1. **AE project working space** — the math AE uses internally (linear vs a tagged display space). Set in File > Project Settings > Color tab.
|
||||
2. **The render's embedded tag** — what transfer/primaries metadata the codec writes (or omits) into the file.
|
||||
3. **The player's interpretation** — QuickTime, Chrome, VLC, and Windows Media each guess differently when the tag is missing or when they apply ColorSync.
|
||||
4. **The display + OS gamma** — macOS historically assumed ~1.8, now ~2.2/2.4-ish via ColorSync; the long-standing "QuickTime gamma bug" comes from QuickTime/AVFoundation applying an extra gamma curve (~1.96) that Chrome and most editors do not.
|
||||
|
||||
Key fact: most web video is **Rec.709**, whose camera/encode side is roughly gamma 2.2-ish but whose *standard display* gamma is **2.4** (BT.1886). macOS QuickTime effectively renders nearer **1.96**, which is why a file can look fine in QuickTime but slightly dark/contrasty in a browser, or vice-versa. The fix is never "eyeball it in QuickTime" — QuickTime is the least trustworthy reference.
|
||||
|
||||
## Decision tree: which problem is this?
|
||||
|
||||
```
|
||||
Render looks DIFFERENT from comp preview?
|
||||
├─ Looks washed-out / milky / brighter, low contrast
|
||||
│ └─ Almost always: untagged or mistagged H.264/ProRes +
|
||||
│ player applying its own gamma (QuickTime gamma bug).
|
||||
│ → Use ProRes for masters; tag Rec.709; verify in
|
||||
│ Chrome, NOT QuickTime.
|
||||
├─ Looks fine in QuickTime, WRONG (dark/contrasty) in Chrome
|
||||
│ └─ QuickTime is applying ~1.96 gamma; Chrome honors the
|
||||
│ 2.4/sRGB pipeline. The browser is the correct reference
|
||||
│ for web delivery.
|
||||
├─ Looks fine on macOS, wrong on Windows (or reverse)
|
||||
│ └─ Missing color tag → each OS assumes a default. Embed an
|
||||
│ explicit Rec.709 (or sRGB) tag on export.
|
||||
└─ Banding / posterization in gradients
|
||||
└─ 8-bit project on heavy grades. Set Project > Depth to
|
||||
16 bpc and/or enable Linearize Working Space.
|
||||
```
|
||||
|
||||
## Project Color Settings (exact menu paths)
|
||||
|
||||
Open **File > Project Settings… > Color** tab.
|
||||
|
||||
- **Depth**: set to **16 bits per channel** for grades/gradients (8 bpc is fine for simple cuts; 32 bpc float only for HDR/heavy linear comps).
|
||||
- **Working Space**: For standard web/SDR delivery choose **Rec.709 Gamma 2.4** (or **sRGB IEC61966-2.1** if the deliverable is screenshots/UI). Setting any working space (not "None") turns AE color management ON, so the viewer now shows display-referred color consistently.
|
||||
- **Linearize Working Space** / **Blend Colors Using 1.0 Gamma**: enable when doing glows, blurs, and additive composites so blending math is physically correct. This changes the look of feathered edges and screen blends — enable it early, not at the end.
|
||||
- **Compensate for Scene-referred Profiles**: leave default unless mixing log/cineon footage.
|
||||
|
||||
To preview what a viewer/OS will actually see, use the composition viewer's **bottom toolbar > "Use Display Color Management"** toggle, and the eyedropper/Info panel to read numeric RGB values rather than trusting the on-screen brightness.
|
||||
|
||||
## Export settings per platform
|
||||
|
||||
Prefer the **Render Queue** (Composition > Add to Render Queue) for ProRes masters, or **File > Export > Add to Adobe Media Encoder Queue** for H.264 web files.
|
||||
|
||||
| Target | Codec / format | Color tag to embed | Notes |
|
||||
|---|---|---|---|
|
||||
| Editing master (FCP/Premiere/Resolve) | **ProRes 422 HQ / 4444** (QuickTime) | Rec.709 | Highest fidelity; avoids H.264 gamma guessing |
|
||||
| Web / `<video>` / social | **H.264** (MP4) via Media Encoder | **Rec.709**, tagged | Set Match Source then override profile; verify in Chrome |
|
||||
| YouTube / Vimeo upload | H.264 high bitrate or ProRes | Rec.709 | Platforms re-encode; correct tag prevents their guess |
|
||||
| Transparency for web | **PNG / ProRes 4444 + alpha** then convert | sRGB | Browsers treat PNG as sRGB |
|
||||
| HDR | ProRes 4444 / HEVC 10-bit | Rec.2020 PQ/HLG | Requires 32 bpc float project |
|
||||
|
||||
In **Adobe Media Encoder**, the QuickTime/ProRes color tag is governed by the AE **Output Module > Color Management** sub-tab. In the Render Queue, click the **Output Module** name > **Color Management** tab and set the output profile to **Rec.709 Gamma 2.4** (do not leave "Preserve RGB" if a shift is occurring).
|
||||
|
||||
## The QuickTime gamma bug — workarounds
|
||||
|
||||
QuickTime Player and any AVFoundation-based player can apply an extra ~1.96 gamma to ProRes/H.264, making files look darker/contrastier than they do in Chrome, Premiere, or Resolve. It is a *player* problem, not a render problem.
|
||||
|
||||
1. **Never grade or approve to QuickTime.** Use Chrome, Premiere, or Resolve as the reference, since those match how the web and other editors display the file.
|
||||
2. For ProRes masters that must look right in QuickTime too, **embed the Rec.709 (Gamma 2.4) tag** via Output Module > Color Management. Correctly tagged ProRes is interpreted consistently by modern macOS.
|
||||
3. If a legacy pipeline still shifts, render a **second 1:1 reference still** (Composition > Save Frame As > File) as PNG and compare numeric values — pixels, not perception.
|
||||
4. Avoid the old "nudge gamma by 1.08 to cancel QuickTime" hacks for new work; they bake a wrong curve that breaks on every correct player. Only use a compensating curve if delivering *exclusively* to a known-broken legacy QuickTime target.
|
||||
|
||||
## Per-player / per-OS gamma reference
|
||||
|
||||
| Environment | Effective transfer assumption | Notes |
|
||||
|---|---|---|
|
||||
| Rec.709 standard display (BT.1886) | **2.4** | The correct broadcast/web display gamma |
|
||||
| sRGB display | **~2.2** (piecewise) | UI, screenshots, PNG |
|
||||
| macOS QuickTime / AVFoundation | **~1.96** | Source of the "QuickTime gamma bug"; brighter/contrast-shifted vs browser |
|
||||
| Chrome / Firefox `<video>` | honors 2.4 / sRGB pipeline | Correct reference for web delivery |
|
||||
| macOS classic (pre-10.6) | 1.8 | Legacy only; ignore for new work |
|
||||
| Windows default | 2.2 | Assumed when file is untagged |
|
||||
|
||||
Takeaway: the *only* reliable cross-environment guarantee is an **explicit embedded tag**. Without it, every player above falls back to its own default and they disagree.
|
||||
|
||||
## Numeric test-patch method (perception-free)
|
||||
|
||||
Eyes adapt; numbers do not. To prove whether a shift exists:
|
||||
|
||||
1. In the comp, add a Solid (Layer > New > Solid) filled with **50% grey: RGB 128,128,128** (or 188,188,188 for a known mid-light patch).
|
||||
2. Note the value AE reports in the **Info panel** when hovering with managed display on.
|
||||
3. Render the file with the intended export settings.
|
||||
4. Re-import the rendered file into a fresh comp (or open in an app with a pixel readout) and sample the same patch.
|
||||
5. If the sampled value differs from 128 by more than ~2-3 levels, a gamma transform is being applied somewhere in the chain. A washed-out look corresponds to the value rising (e.g. 128 → 150+).
|
||||
|
||||
This isolates whether the shift is in the *render* (values changed in the file) vs the *playback* (file values correct, player re-curving).
|
||||
|
||||
## Verify the embedded tag with ffprobe
|
||||
|
||||
```bash
|
||||
ffprobe -v error -select_streams v:0 \
|
||||
-show_entries stream=color_transfer,color_primaries,color_space,color_range \
|
||||
-of default=nk=0 export.mp4
|
||||
```
|
||||
|
||||
Desired for SDR web (Rec.709):
|
||||
|
||||
```
|
||||
color_space=bt709
|
||||
color_primaries=bt709
|
||||
color_transfer=bt709
|
||||
color_range=tv # (limited) for video; pc/full for some web pipelines
|
||||
```
|
||||
|
||||
If `color_transfer=unknown` or fields are absent, the file is untagged → re-export with the tag, or stamp it without re-encoding:
|
||||
|
||||
```bash
|
||||
ffmpeg -i export.mp4 -c copy \
|
||||
-color_primaries bt709 -color_trc bt709 -colorspace bt709 \
|
||||
export_tagged.mp4
|
||||
```
|
||||
|
||||
Note: `-c copy` only writes container/stream metadata; it does not fix pixels that were already baked wrong — use it to *tag*, not to *correct*.
|
||||
|
||||
## Limited vs full range pitfall
|
||||
|
||||
H.264 for video is typically **limited/TV range** (luma 16-235). If AE/AME outputs full-range pixels but the file is tagged limited (or vice versa), blacks crush or wash. Symptoms: milky blacks = full data read as limited; crushed blacks = limited data read as full. Keep range consistent end-to-end and tag it (`color_range`).
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Symptom | Likely cause | Fix |
|
||||
|---|---|---|
|
||||
| Washed-out / milky blacks | Untagged H.264, player adds gamma | Tag Rec.709 in Output Module > Color Management |
|
||||
| Fine in QuickTime, dark in Chrome | QuickTime ~1.96 gamma | Trust Chrome; QuickTime is wrong for web |
|
||||
| Differs macOS vs Windows | No embedded color tag | Embed explicit Rec.709/sRGB tag |
|
||||
| Banding in gradients | 8 bpc project | Project Settings > Depth = 16 bpc |
|
||||
| Glows/blurs look wrong | Non-linear blending | Enable Linearize Working Space |
|
||||
| Screenshots/UI look off | Wrong working space | Use sRGB working space, not Rec.709 |
|
||||
|
||||
## Gotchas
|
||||
|
||||
- "Working Space = None" disables AE color management entirely — the viewer then lies about the final look. Always set an explicit space for managed delivery.
|
||||
- Toggling Linearize Working Space late changes the look of every feather, glow, and screen blend; decide at project start.
|
||||
- AME "Match Source" can silently drop the color tag; verify the exported file's tag (e.g. with `ffprobe -show_streams`, look for `color_transfer`/`color_primaries`).
|
||||
- ProRes is not magic — an untagged ProRes still shifts. The *tag* is what fixes consistency, not the codec.
|
||||
- sRGB and Rec.709 share primaries but differ in transfer (sRGB ~2.2 piecewise vs Rec.709 display 2.4); using the wrong one shifts midtone brightness.
|
||||
|
||||
## Pre-export checklist
|
||||
|
||||
- [ ] Project Settings > Color: Working Space set to the deliverable's space (Rec.709 Gamma 2.4 for SDR video; sRGB for UI/stills).
|
||||
- [ ] Project Settings > Color: Depth = 16 bpc (or 32 for HDR/heavy linear).
|
||||
- [ ] Linearize Working Space decision made at project start (on for glow/blur-heavy comps).
|
||||
- [ ] Display Color Management toggle ON while reviewing, so the viewer reflects final look.
|
||||
- [ ] 50% grey test patch placed and noted before render.
|
||||
- [ ] Output Module > Color Management: explicit output profile = Rec.709 Gamma 2.4 (not "Preserve RGB" when a shift exists).
|
||||
- [ ] Codec chosen per target (ProRes 422 HQ/4444 master, H.264 web).
|
||||
- [ ] Reviewed final in **Chrome / Premiere / Resolve**, not QuickTime.
|
||||
- [ ] `ffprobe` confirms color_primaries/transfer/space = bt709 (or sRGB target) and color_range consistent.
|
||||
- [ ] Re-imported render and re-sampled the grey patch; value matches within ~2-3 levels.
|
||||
|
||||
## When the client only has QuickTime
|
||||
|
||||
If the approval environment is genuinely locked to macOS QuickTime and the file must look right *there*:
|
||||
|
||||
1. Deliver correctly tagged ProRes (modern macOS QuickTime respects Rec.709 tags fairly well).
|
||||
2. If a residual shift remains and cannot be avoided, produce a *separate* QuickTime-targeted file with a measured compensating gamma adjustment — and clearly label it "QuickTime-only; do not use for web." Never let this compensated file become the web master.
|
||||
3. Educate: provide the same file opened in Chrome side-by-side so the difference is attributed to the player, not the work.
|
||||
@@ -0,0 +1,180 @@
|
||||
# Palettes, gradients, interpolation, and grading
|
||||
|
||||
Exact recipes and runnable code. OKLCH notation: `oklch(L C H)`, L in 0-1 perceptual lightness, C chroma (0 to ~0.37), H hue degrees.
|
||||
|
||||
## Palette construction rules
|
||||
|
||||
1. **Count**: 1 primary, 1 accent, 2-3 neutrals. Resist adding hues.
|
||||
2. **Contrast from lightness**: build a 5-9 step ramp by varying L while holding C and H roughly constant, nudging chroma down at the extremes (very light/dark colors can't hold high chroma).
|
||||
3. **Accent placement**: accent hue should sit 90-180deg from the primary on the OKLCH wheel for clean separation; complementary (~180deg) is boldest, analogous (~30deg) is calmest.
|
||||
4. **Neutral tinting**: give neutrals a tiny chroma (0.005-0.02) in the primary's hue family so the frame feels unified, not assembled from separate kits.
|
||||
5. **Dark-scene accents**: high L (0.7-0.85), moderate C (0.12-0.18). Avoid max chroma — it vibrates on dark backgrounds.
|
||||
|
||||
### Generating a tonal ramp (hold hue, vary lightness)
|
||||
|
||||
```
|
||||
oklch(0.97 0.03 264) // 50
|
||||
oklch(0.92 0.06 264) // 100
|
||||
oklch(0.84 0.10 264) // 200
|
||||
oklch(0.74 0.14 264) // 300
|
||||
oklch(0.66 0.18 264) // 400
|
||||
oklch(0.58 0.19 264) // 500 (base)
|
||||
oklch(0.50 0.17 264) // 600
|
||||
oklch(0.40 0.14 264) // 700
|
||||
oklch(0.28 0.10 264) // 800
|
||||
oklch(0.18 0.06 264) // 900
|
||||
```
|
||||
|
||||
Note chroma rises toward the mid-lightness band and falls at both ends — this matches what's physically representable and looks natural.
|
||||
|
||||
### Named starter palettes (OKLCH)
|
||||
|
||||
```
|
||||
// "Midnight product" — confident, techy
|
||||
primary oklch(0.62 0.19 264) accent oklch(0.78 0.17 50)
|
||||
bg oklch(0.16 0.02 264) fg oklch(0.97 0.005 264)
|
||||
|
||||
// "Warm editorial" — premium, soft
|
||||
primary oklch(0.55 0.12 25) accent oklch(0.80 0.10 90)
|
||||
bg oklch(0.96 0.01 60) fg oklch(0.25 0.03 25)
|
||||
|
||||
// "Neon dark" — energetic, bold
|
||||
primary oklch(0.72 0.20 150) accent oklch(0.68 0.24 330)
|
||||
bg oklch(0.14 0.03 280) fg oklch(0.95 0.02 150)
|
||||
```
|
||||
|
||||
## Gradient cookbook
|
||||
|
||||
### Two-stop, vivid (OKLCH)
|
||||
|
||||
```css
|
||||
background: linear-gradient(135deg in oklch, #2b6cff, #ffb02e);
|
||||
/* control the hue arc explicitly */
|
||||
background: linear-gradient(in oklch shorter hue, oklch(0.6 0.2 264), oklch(0.8 0.17 50));
|
||||
```
|
||||
|
||||
`shorter hue` / `longer hue` / `increasing hue` / `decreasing hue` pick the direction around the wheel — `longer hue` produces rainbow sweeps, `shorter` the direct path.
|
||||
|
||||
### Mesh background (stacked radials)
|
||||
|
||||
```css
|
||||
.mesh {
|
||||
background-color: #0b0f1a;
|
||||
background-image:
|
||||
radial-gradient(at 18% 22%, oklch(0.62 0.19 264 / 0.7), transparent 50%),
|
||||
radial-gradient(at 82% 28%, oklch(0.70 0.16 320 / 0.6), transparent 55%),
|
||||
radial-gradient(at 55% 88%, oklch(0.78 0.17 50 / 0.5), transparent 50%),
|
||||
radial-gradient(at 40% 50%, oklch(0.55 0.14 200 / 0.4), transparent 60%);
|
||||
}
|
||||
```
|
||||
|
||||
### Animated drift
|
||||
|
||||
```css
|
||||
.mesh { background-size: 140% 140%; animation: drift 16s ease-in-out infinite alternate; }
|
||||
@keyframes drift {
|
||||
0% { background-position: 0% 0%; }
|
||||
100% { background-position: 12% -8%; }
|
||||
}
|
||||
```
|
||||
|
||||
### Conic
|
||||
|
||||
```css
|
||||
background: conic-gradient(in oklch from 90deg, #2b6cff, #b02eff, #ffb02e, #2b6cff);
|
||||
```
|
||||
|
||||
### Grain overlay (kills banding)
|
||||
|
||||
```html
|
||||
<svg width="0" height="0">
|
||||
<filter id="grain"><feTurbulence type="fractalNoise" baseFrequency="0.8" numOctaves="2"/></filter>
|
||||
</svg>
|
||||
```
|
||||
```css
|
||||
.grain::after {
|
||||
content: ""; position: absolute; inset: 0; pointer-events: none;
|
||||
filter: url(#grain); opacity: 0.04; mix-blend-mode: overlay;
|
||||
}
|
||||
```
|
||||
|
||||
## Perceptual interpolation in code
|
||||
|
||||
### culori (JS) — interpolate and sample
|
||||
|
||||
```js
|
||||
import { interpolate, formatHex, formatCss } from "culori";
|
||||
|
||||
// Interpolate in OKLCH (vivid midpoints, even lightness)
|
||||
const mix = interpolate(["#2b6cff", "#ffb02e"], "oklch");
|
||||
formatHex(mix(0.5)); // vivid mid, not gray
|
||||
|
||||
// Build N stops for an animated transition
|
||||
const stops = Array.from({ length: 60 }, (_, i) => formatCss(mix(i / 59)));
|
||||
|
||||
// Control hue direction
|
||||
import { fixupHueLonger } from "culori";
|
||||
const rainbow = interpolate(["#2b6cff", "#ffb02e"], "oklch", {
|
||||
h: { use: fixupHueLonger },
|
||||
});
|
||||
```
|
||||
|
||||
### d3-interpolate / d3-scale-chromatic
|
||||
|
||||
```js
|
||||
import { interpolateLab, interpolateHcl, quantize } from "d3-interpolate";
|
||||
|
||||
const lab = interpolateLab("#2b6cff", "#ffb02e"); // perceptual, smooth lightness
|
||||
lab(0.5); // mid color
|
||||
const ramp = quantize(interpolateHcl("#2b6cff", "#ffb02e"), 8); // 8 even stops
|
||||
```
|
||||
|
||||
`interpolateLab` and `interpolateHcl` avoid the sRGB gray-midpoint problem; `interpolateRgb` does not — only use RGB interpolation for near-identical colors.
|
||||
|
||||
### Animating a color with @property (CSS, browser interpolates)
|
||||
|
||||
```css
|
||||
@property --from { syntax: "<color>"; initial-value: oklch(0.62 0.19 264); inherits: false; }
|
||||
.swatch { background: var(--from); transition: --from 400ms cubic-bezier(0.16,1,0.3,1); }
|
||||
.swatch.active { --from: oklch(0.78 0.17 50); }
|
||||
```
|
||||
|
||||
Without `@property` registration the color snaps instead of interpolating. The browser interpolates registered `<color>` properties in OKLab by default — perceptually clean for free.
|
||||
|
||||
## Grading order checklist
|
||||
|
||||
Apply in this exact order; reordering changes the result.
|
||||
|
||||
1. **Exposure / white balance** — set overall brightness, neutralize or intentionally tint the cast.
|
||||
2. **Contrast** — set black/white points, expand or compress tonal range.
|
||||
3. **Midtones / saturation** — overall saturation and midtone color shift.
|
||||
4. **Split-tone** — tint shadows and highlights separately. Cinematic default: teal shadows (`hue ~190`), orange highlights (`hue ~40`), kept subtle.
|
||||
5. **Unify** — final overall tint / LUT and a gentle vignette.
|
||||
|
||||
### CSS-filter grade (quick, for web video/canvas)
|
||||
|
||||
```css
|
||||
.graded {
|
||||
filter:
|
||||
brightness(1.02) /* 1: exposure */
|
||||
contrast(1.12) /* 2: contrast */
|
||||
saturate(1.08); /* 3: saturation */
|
||||
}
|
||||
/* 4 split-tone: overlay a teal->orange gradient at low opacity with mix-blend-mode */
|
||||
```
|
||||
|
||||
### Fragment-shader split-tone (GLSL sketch)
|
||||
|
||||
```glsl
|
||||
vec3 graded = color;
|
||||
float lum = dot(graded, vec3(0.2126, 0.7152, 0.0722));
|
||||
vec3 shadowTint = vec3(0.0, 0.45, 0.5); // teal
|
||||
vec3 highlightTint = vec3(1.0, 0.6, 0.2); // orange
|
||||
graded = mix(graded, graded * shadowTint, (1.0 - lum) * 0.12);
|
||||
graded = mix(graded, graded + highlightTint * 0.06, lum);
|
||||
```
|
||||
|
||||
Keep split-tone strength under ~0.15; over-tinting reads as a cheap Instagram filter. Grade the whole piece with the same chain so cuts match.
|
||||
|
||||
---
|
||||
Interpolate color in perceptual space and gradients shift without muddy mids. Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=motion-design-skills&utm_content=skill_footer&utm_term=color-motion)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -1,25 +0,0 @@
|
||||
# Countdown Video
|
||||
|
||||
Make a countdown that ends exactly when it should. The whole craft is one idea: derive the remaining time from the frame (or a monotonic timestamp), then format and animate that number — never count down with a `setInterval`. Covers livestream "starting soon" screens, launch and flash-sale countdowns, and looping backgrounds. For creators who need a timer that lands on zero on the exact frame.
|
||||
|
||||
## When it activates
|
||||
|
||||
- "Make a countdown timer video" or "create a countdown animation."
|
||||
- "Build a stream starting soon countdown" or "do an event/launch/sale countdown."
|
||||
- "Animate a mm:ss or dd:hh:mm:ss timer."
|
||||
- "Add digit roll/flip transitions to a counter" or "loop a countdown background."
|
||||
|
||||
## Example prompts
|
||||
|
||||
- "Build a 10-minute 'be right back' livestream countdown."
|
||||
- "Make a dd:hh:mm:ss launch countdown to a fixed target date."
|
||||
- "Add flip transitions to an mm:ss timer that hits zero on the exact frame."
|
||||
|
||||
## What's inside
|
||||
|
||||
- `SKILL.md` — frame-accurate (drift-free) timers, time formatting, and a configurable duration prop.
|
||||
- `references/countdown-cookbook.md` — recipes for common countdown scenarios.
|
||||
- `references/digit-flip.md` — digit flip/roll transition techniques.
|
||||
|
||||
---
|
||||
Part of **[Short-Form Video Skills](../../)** · Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=tiktok-video-skills&utm_content=skill_footer&utm_term=countdown-video)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -1,152 +0,0 @@
|
||||
---
|
||||
name: countdown-video
|
||||
description: This skill should be used when the user asks to "make a countdown timer video", "create a countdown animation", "build a stream starting soon countdown", "do an event/launch/sale countdown", "animate a mm:ss or dd:hh:mm:ss timer", "add digit roll/flip transitions to a counter", or "loop a countdown background". Covers frame-accurate (drift-free) timers, digit flip/roll motion, time formatting, and a configurable duration prop.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Countdown Video
|
||||
|
||||
Make a countdown that ends exactly when it should. The whole craft is one idea: derive the remaining time from the frame (or a monotonic timestamp), then format and animate that number — never count down with a `setInterval`. Covers livestream "starting soon" screens, launch/sale countdowns, and looping backgrounds.
|
||||
|
||||
## When to use
|
||||
|
||||
- "Starting soon" / "be right back" livestream screens with a 5–10 min timer.
|
||||
- Launch and flash-sale countdowns (dd:hh:mm:ss down to a target date; mm:ss for urgency).
|
||||
- Any animated counter where the number must land on zero on the exact frame.
|
||||
|
||||
## The one rule: drive the number from the clock, never a tick
|
||||
|
||||
A `setInterval(fn, 1000)` countdown **drifts** — the callback fires *at least* 1000 ms later, never exactly, and stacks up when the tab is throttled or a render frame is slow. Reports of ~1 second lost per minute are common. The fix is the same for rendered video and live web: compute remaining time as a pure function of an authoritative clock, so a late frame self-corrects on the next frame instead of accumulating error.
|
||||
|
||||
| Context | Authoritative clock | Remaining time |
|
||||
|---|---|---|
|
||||
| Remotion render | `useCurrentFrame()` | `total − frame/fps` |
|
||||
| Web (live page) | `performance.now()` or `Date.now()` | `targetMs − now` |
|
||||
| To a fixed date | `Date.now()` | `targetTimestamp − Date.now()` |
|
||||
|
||||
Never store remaining time in state and decrement it. Recompute it every frame.
|
||||
|
||||
## Frame-accurate countdown (Remotion → MP4)
|
||||
|
||||
`useCurrentFrame()` is deterministic: frame N is always rendered at exactly `N/fps` seconds. Subtract from the total and the timer is drift-free by construction.
|
||||
|
||||
```tsx
|
||||
import { useCurrentFrame, useVideoConfig } from "remotion";
|
||||
|
||||
export const Countdown: React.FC<{ durationInSeconds: number }> = ({
|
||||
durationInSeconds,
|
||||
}) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps } = useVideoConfig();
|
||||
|
||||
// remaining seconds, clamped at 0 — ceil so "1" shows for its full second
|
||||
const elapsed = frame / fps;
|
||||
const remaining = Math.max(0, Math.ceil(durationInSeconds - elapsed));
|
||||
|
||||
return <div style={{ fontVariantNumeric: "tabular-nums" }}>{formatMMSS(remaining)}</div>;
|
||||
};
|
||||
```
|
||||
|
||||
`durationInSeconds` is a **prop**, not a constant — the same composition renders a 5-minute stream timer or a 30-second sale timer by changing one input. Set the composition's `durationInFrames` to `durationInSeconds * fps` so the video is exactly as long as the countdown.
|
||||
|
||||
## Web timer (live page, requestAnimationFrame)
|
||||
|
||||
For a live page, drive an rAF loop from `performance.now()` and only repaint when the displayed second changes (no work 60×/sec for a 1 Hz number):
|
||||
|
||||
```js
|
||||
const start = performance.now();
|
||||
let lastShown = -1;
|
||||
function tick(now) {
|
||||
const remaining = Math.max(0, Math.ceil(duration - (now - start) / 1000));
|
||||
if (remaining !== lastShown) { render(remaining); lastShown = remaining; }
|
||||
if (remaining > 0) requestAnimationFrame(tick);
|
||||
}
|
||||
requestAnimationFrame(tick);
|
||||
```
|
||||
|
||||
Counting down to a fixed date is the same idea with `Date.now()`: `Math.max(0, target - Date.now())`. See `references/countdown-cookbook.md` for the full date-target version with timezone handling.
|
||||
|
||||
## Time formatting
|
||||
|
||||
Break a remaining-seconds integer into fields with division + modulo, then zero-pad. Pick the format to fit the duration — never show `00d 00h 05m` when minutes is all that matters.
|
||||
|
||||
```js
|
||||
const pad = (n) => String(n).padStart(2, "0");
|
||||
const formatMMSS = (s) => `${pad(Math.floor(s / 60))}:${pad(s % 60)}`;
|
||||
const formatDHMS = (s) => {
|
||||
const d = Math.floor(s / 86400), h = Math.floor((s % 86400) / 3600);
|
||||
const m = Math.floor((s % 3600) / 60), sec = s % 60;
|
||||
return `${pad(d)}:${pad(h)}:${pad(m)}:${pad(sec)}`;
|
||||
};
|
||||
```
|
||||
|
||||
| Format | Use | Drop fields when |
|
||||
|---|---|---|
|
||||
| `mm:ss` | stream "starting soon", short timers (<1h) | — |
|
||||
| `hh:mm:ss` | flash sale, same-day event | hide `dd` while 0 |
|
||||
| `dd:hh:mm:ss` | launch / multi-day countdown | hide leading 0 units for cleaner read |
|
||||
|
||||
Always set `font-variant-numeric: tabular-nums` (or a monospace digit font) so digits keep equal width and the layout does not jitter as numbers change.
|
||||
|
||||
## Digit roll / flip transitions
|
||||
|
||||
A flat number swap reads as cheap; the rolling "odometer" look sells the countdown. Render each digit as a vertical strip 0–9 and translate it by `-digit × digitHeight`. Animate the *position*, not the text content, so the change is a glide, not a pop.
|
||||
|
||||
```css
|
||||
.reel { height: 1em; overflow: hidden; }
|
||||
.reel .col { transition: transform .45s cubic-bezier(.22,1,.36,1); }
|
||||
/* col holds stacked 0..9; shift by -value em */
|
||||
```
|
||||
|
||||
In Remotion, drive the strip offset off the frame instead of a CSS transition (so it is deterministic). The flip-card style (top half folds down over the bottom) and the per-digit reel are both detailed, with runnable components, in `references/digit-flip.md`.
|
||||
|
||||
## Looping background
|
||||
|
||||
Stream screens hold for minutes, so the background must loop seamlessly. Build motion on a cycle that divides evenly into the timeline (e.g. a gradient or particle drift whose period is `loopFrames`), and read it as `frame % loopFrames` so there is no visible seam at the wrap. Keep it low-contrast and slow — the number is the subject; the background is ambience. Recipes in `references/countdown-cookbook.md`.
|
||||
|
||||
## Output checklist
|
||||
|
||||
- Remaining time recomputed from `useCurrentFrame()` / `performance.now()` every frame — no `setInterval` decrement.
|
||||
- Timer hits `00:00` on the exact intended frame; `durationInFrames === durationInSeconds × fps`.
|
||||
- Duration is an input prop, not a hardcoded constant.
|
||||
- Digits use `tabular-nums`; layout does not shift as numbers change.
|
||||
- Format matches the span (mm:ss vs dd:hh:mm:ss); leading zero units dropped when distracting.
|
||||
- Background loop period divides the timeline evenly; no seam at the wrap.
|
||||
|
||||
## Deliver & verify (rendered stills → MP4)
|
||||
|
||||
> **Packaged helper** (`scripts/`): tile your stills with `scripts/contact-sheet.sh sheet.png f-hook.png f-mid.png f-end.png`, then assert the encode with `scripts/probe-mp4.sh out.mp4 [WxH] [fps]`. See `scripts/README.md`.
|
||||
|
||||
The countdown ships as a Remotion composition (`<Composition>` + zod `schema` + `defaultProps`, `durationInSeconds` a prop) — the timer is a pure function of `useCurrentFrame()`, never `setInterval` / `Date.now()` / `Math.random()`. Set `durationInFrames = durationInSeconds * fps`. Deliverable = `out/*.mp4` + the project. 9:16 vertical (1080×1920) is the default.
|
||||
|
||||
**Verify loop — render stills → inspect → encode.** The whole point is frame-accuracy, so check the *digit value* at exact frames before encoding.
|
||||
|
||||
```bash
|
||||
# Stills at start / mid / end — WITH SHIPPED PROPS (the real duration you'll render)
|
||||
npx remotion still Countdown out/f-start.png --frame=0 --props='{"durationInSeconds":300}'
|
||||
npx remotion still Countdown out/f-mid.png --frame=N --props='{"durationInSeconds":300}'
|
||||
npx remotion still Countdown out/f-end.png --frame=L --props='{"durationInSeconds":300}' # L = durationInFrames-1
|
||||
|
||||
# Inspect each PNG — the digits must read EXACTLY the frame/fps math:
|
||||
# - at frame N, displayed = ceil(durationInSeconds - N/fps) (e.g. 300s @ 30fps, frame 4500 → 150s → "02:30")
|
||||
# - last frame reads 00:00 (or the end-state hold); no drift, no off-by-one second
|
||||
# - tabular-nums: layout does not jitter between frames
|
||||
# - 9:16 safe area: the timer + brand block sit clear of top ~12% and bottom ~20-35% UI and the right action rail
|
||||
|
||||
npx remotion render Countdown out/countdown.mp4 --props='{"durationInSeconds":300}' # encode once digits verify
|
||||
npx remotion render Countdown out/demo.gif --codec=gif # README proof clip
|
||||
```
|
||||
|
||||
Use `npx remotion compositions` to confirm `durationInFrames`/`fps`, then compute the frame for any second you want to check.
|
||||
|
||||
**Before you finish:**
|
||||
1. Stills render cleanly at frame 0, mid, and last — no errors.
|
||||
2. The digit value at frame N equals `ceil(durationInSeconds - N/fps)` exactly; last frame hits 00:00 on the intended frame (no drift / off-by-one).
|
||||
3. Timer + brand stay inside the 9:16 safe area at every checked frame.
|
||||
4. Frame-driven only — no `setInterval` / `Date.now()` / timers; `tabular-nums` so layout holds.
|
||||
5. Shipped `durationInSeconds` (not just `defaultProps`) renders correctly; MP4 encoded, GIF optional.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/countdown-cookbook.md` — complete runnable Remotion `<Countdown>` (configurable duration, mm:ss / dd:hh:mm:ss, end-state hold), the count-down-to-a-fixed-date web component with timezone handling, a "starting soon" stream-screen layout (brand block + social handles + looping gradient background), and the seamless-loop background pattern.
|
||||
- `references/digit-flip.md` — two deterministic digit-animation components: the odometer reel (vertical 0–9 strip driven by frame) and the flip-card (top-half fold) effect, plus when each suits launch vs sale vs stream.
|
||||
@@ -1,205 +0,0 @@
|
||||
# Countdown Cookbook
|
||||
|
||||
Complete, runnable recipes. Every value is derived from an authoritative clock (frame or timestamp) so nothing drifts.
|
||||
|
||||
## 1. Configurable Remotion countdown (mm:ss + dd:hh:mm:ss)
|
||||
|
||||
A single composition that handles any duration and either format. Drop into `src/Countdown.tsx`.
|
||||
|
||||
```tsx
|
||||
import React from "react";
|
||||
import { useCurrentFrame, useVideoConfig, AbsoluteFill } from "remotion";
|
||||
|
||||
const pad = (n: number) => String(n).padStart(2, "0");
|
||||
|
||||
const formatTime = (totalSeconds: number, mode: "mm:ss" | "dd:hh:mm:ss") => {
|
||||
const s = Math.max(0, totalSeconds);
|
||||
if (mode === "mm:ss") {
|
||||
return `${pad(Math.floor(s / 60))}:${pad(s % 60)}`;
|
||||
}
|
||||
const d = Math.floor(s / 86400);
|
||||
const h = Math.floor((s % 86400) / 3600);
|
||||
const m = Math.floor((s % 3600) / 60);
|
||||
const sec = s % 60;
|
||||
// drop leading day field once it is zero, for a cleaner read
|
||||
return d > 0
|
||||
? `${pad(d)}d ${pad(h)}:${pad(m)}:${pad(sec)}`
|
||||
: `${pad(h)}:${pad(m)}:${pad(sec)}`;
|
||||
};
|
||||
|
||||
export type CountdownProps = {
|
||||
durationInSeconds: number;
|
||||
format?: "mm:ss" | "dd:hh:mm:ss";
|
||||
label?: string;
|
||||
};
|
||||
|
||||
export const Countdown: React.FC<CountdownProps> = ({
|
||||
durationInSeconds,
|
||||
format = "mm:ss",
|
||||
label = "",
|
||||
}) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps } = useVideoConfig();
|
||||
|
||||
// ceil so the last whole second ("1") is shown for its full duration,
|
||||
// and the display reaches "0" exactly on the final frame.
|
||||
const remaining = Math.max(0, Math.ceil(durationInSeconds - frame / fps));
|
||||
const ended = remaining === 0;
|
||||
|
||||
return (
|
||||
<AbsoluteFill
|
||||
style={{
|
||||
justifyContent: "center",
|
||||
alignItems: "center",
|
||||
background: "#0b0d17",
|
||||
color: "#fff",
|
||||
fontFamily: "Inter, system-ui, sans-serif",
|
||||
}}
|
||||
>
|
||||
{label && (
|
||||
<div style={{ fontSize: 36, opacity: 0.7, marginBottom: 16, letterSpacing: 4 }}>
|
||||
{label.toUpperCase()}
|
||||
</div>
|
||||
)}
|
||||
<div
|
||||
style={{
|
||||
fontSize: 180,
|
||||
fontWeight: 800,
|
||||
fontVariantNumeric: "tabular-nums", // equal-width digits → no jitter
|
||||
lineHeight: 1,
|
||||
color: ended ? "#ff4d4d" : "#fff",
|
||||
}}
|
||||
>
|
||||
{ended ? formatTime(0, format) : formatTime(remaining, format)}
|
||||
</div>
|
||||
</AbsoluteFill>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
Register it so `durationInFrames` matches the countdown exactly — the video ends when the timer does:
|
||||
|
||||
```tsx
|
||||
import { Composition } from "remotion";
|
||||
import { Countdown } from "./Countdown";
|
||||
|
||||
const FPS = 30;
|
||||
const DURATION_SECONDS = 300; // 5-minute "starting soon" timer
|
||||
|
||||
export const RemotionRoot = () => (
|
||||
<Composition
|
||||
id="Countdown"
|
||||
component={Countdown}
|
||||
durationInFrames={DURATION_SECONDS * FPS}
|
||||
fps={FPS}
|
||||
width={1920}
|
||||
height={1080}
|
||||
defaultProps={{ durationInSeconds: DURATION_SECONDS, format: "mm:ss", label: "Starting soon" }}
|
||||
/>
|
||||
);
|
||||
```
|
||||
|
||||
Render any duration/format without touching code — duration is a prop:
|
||||
|
||||
```bash
|
||||
npx remotion render Countdown out/sale.mp4 \
|
||||
--props='{"durationInSeconds":86400,"format":"dd:hh:mm:ss","label":"Sale ends in"}'
|
||||
```
|
||||
|
||||
To hold a final state (e.g. "WE'RE LIVE") instead of ending on `00:00`, extend `durationInFrames` past the countdown and branch on `ended`.
|
||||
|
||||
## 2. Count down to a fixed date (live web component)
|
||||
|
||||
For a real-time page that targets a calendar moment. Drift-free because remaining time is `target − Date.now()` recomputed each frame, never a decrement.
|
||||
|
||||
```js
|
||||
// targetISO must include an offset, e.g. "2026-07-01T09:00:00-07:00",
|
||||
// so the deadline is unambiguous regardless of the viewer's timezone.
|
||||
function startCountdown(targetISO, el) {
|
||||
const target = new Date(targetISO).getTime();
|
||||
const pad = (n) => String(n).padStart(2, "0");
|
||||
let lastShown = -1;
|
||||
|
||||
function tick() {
|
||||
const remainingMs = Math.max(0, target - Date.now());
|
||||
const s = Math.floor(remainingMs / 1000);
|
||||
if (s !== lastShown) {
|
||||
const d = Math.floor(s / 86400);
|
||||
const h = Math.floor((s % 86400) / 3600);
|
||||
const m = Math.floor((s % 3600) / 60);
|
||||
el.textContent = `${pad(d)}:${pad(h)}:${pad(m)}:${pad(s % 60)}`;
|
||||
lastShown = s;
|
||||
}
|
||||
if (remainingMs > 0) requestAnimationFrame(tick);
|
||||
else el.textContent = "00:00:00:00"; // or swap to an "It's live" state
|
||||
}
|
||||
requestAnimationFrame(tick);
|
||||
}
|
||||
```
|
||||
|
||||
Timezone notes:
|
||||
- Bake the offset into the target string (`-07:00`, `Z`). A bare `"2026-07-01T09:00:00"` is parsed as the viewer's *local* time and every viewer sees a different deadline.
|
||||
- To show "ends 9am in *your* timezone everywhere," instead pick the wall-clock time per locale and let each viewer's `Date` resolve it — but that is a different deadline per region. Decide which you want before building.
|
||||
|
||||
## 3. "Starting soon" stream screen
|
||||
|
||||
Livestream lead-in screens hold 5–10 minutes: a clean brand block, social handles, and a calm looping background, with the timer as the focal point. Keep it low-clutter — chat and music carry the wait; the screen just signals "we're about to start."
|
||||
|
||||
```tsx
|
||||
import { AbsoluteFill, useCurrentFrame, useVideoConfig, interpolate } from "remotion";
|
||||
|
||||
export const StartingSoon: React.FC<{ durationInSeconds: number; handle: string; brand: string }> = ({
|
||||
durationInSeconds,
|
||||
handle,
|
||||
brand,
|
||||
}) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps, durationInFrames } = useVideoConfig();
|
||||
const remaining = Math.max(0, Math.ceil(durationInSeconds - frame / fps));
|
||||
const pad = (n: number) => String(n).padStart(2, "0");
|
||||
|
||||
// seamless looping gradient: angle cycles over loopFrames with no seam
|
||||
const loopFrames = fps * 8;
|
||||
const angle = interpolate(frame % loopFrames, [0, loopFrames], [0, 360]);
|
||||
|
||||
return (
|
||||
<AbsoluteFill
|
||||
style={{
|
||||
background: `linear-gradient(${angle}deg, #141a3a, #0b0d17 60%)`,
|
||||
color: "#fff",
|
||||
fontFamily: "Inter, system-ui, sans-serif",
|
||||
justifyContent: "center",
|
||||
alignItems: "center",
|
||||
}}
|
||||
>
|
||||
<div style={{ fontSize: 64, fontWeight: 800, marginBottom: 8 }}>{brand}</div>
|
||||
<div style={{ fontSize: 28, opacity: 0.6, letterSpacing: 6, marginBottom: 40 }}>
|
||||
STARTING SOON
|
||||
</div>
|
||||
<div style={{ fontSize: 200, fontWeight: 800, fontVariantNumeric: "tabular-nums", lineHeight: 1 }}>
|
||||
{pad(Math.floor(remaining / 60))}:{pad(remaining % 60)}
|
||||
</div>
|
||||
<div style={{ position: "absolute", bottom: 60, fontSize: 30, opacity: 0.7 }}>{handle}</div>
|
||||
</AbsoluteFill>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
Export as MP4 and add it as a Media/Video source in OBS, or use a looping playout for a browser source. Common durations: 300s (5 min) or 600s (10 min) at 1920×1080.
|
||||
|
||||
## 4. Seamless looping background
|
||||
|
||||
Any motion whose period divides the timeline evenly wraps without a seam. Read animation phase as `frame % loopFrames` so the value at the wrap equals the value at frame 0.
|
||||
|
||||
```tsx
|
||||
const loopFrames = fps * 6; // one full cycle every 6 seconds
|
||||
const phase = (frame % loopFrames) / loopFrames; // 0 → 1, repeats
|
||||
const driftY = Math.sin(phase * Math.PI * 2) * 40; // returns to start each cycle
|
||||
```
|
||||
|
||||
Use sine/cosine (inherently periodic) or `interpolate(frame % loopFrames, [0, loopFrames], [a, a])` with matching start/end so there is no jump. Keep contrast low and motion slow — the number is the subject.
|
||||
|
||||
---
|
||||
## Built by the team behind iart.ai
|
||||
|
||||
This skill is part of an open motion-graphics collection from iart.ai — the AI motion agent that turns data, scripts, and designs into editable motion graphics (Remotion → MP4). If you'd rather not hand-build this, iart.ai can render countdown videos from a template with a set time and brand — change the text/data and re-export. → [iart.ai](https://iart.ai/?utm_source=github&utm_medium=reference&utm_campaign=tiktok-video-skills&utm_content=ref_footer&utm_term=countdown-video)
|
||||
@@ -1,146 +0,0 @@
|
||||
# Digit Flip & Roll Transitions
|
||||
|
||||
Two ways to animate a digit changing. Both are driven by the frame (deterministic) so they render identically every time — no CSS transition clock that the renderer can disagree with.
|
||||
|
||||
Pick by tone:
|
||||
|
||||
| Effect | Feel | Best for |
|
||||
|---|---|---|
|
||||
| Odometer reel | smooth, continuous glide | launch countdowns, premium/minimal timers |
|
||||
| Flip card | mechanical "snap", split-flap board | sale urgency, retro/airport-board style |
|
||||
|
||||
## Shared idea
|
||||
|
||||
A single displayed digit is a stack of glyphs `0..9` laid out vertically. Showing digit `d` means translating the stack so glyph `d` sits in the window: `translateY(-d × digitHeight)`. Animate the *offset*, not the text, and you get a roll. The trick for a countdown is that the offset is continuous: between whole seconds the digit eases from its old value toward its new one.
|
||||
|
||||
## Odometer reel (Remotion, frame-driven)
|
||||
|
||||
```tsx
|
||||
import React from "react";
|
||||
import { useCurrentFrame, useVideoConfig, interpolate, Easing } from "remotion";
|
||||
|
||||
const DIGIT_H = 160; // px per glyph
|
||||
const ROLL_FRAMES = 12; // how long a single digit roll takes
|
||||
|
||||
// Animated single digit: rolls from `prev` to `value` over the last ROLL_FRAMES
|
||||
// of the current second.
|
||||
const Reel: React.FC<{ value: number; prev: number; secondProgress: number }> = ({
|
||||
value,
|
||||
prev,
|
||||
secondProgress,
|
||||
}) => {
|
||||
// secondProgress: 0→1 across the current whole second.
|
||||
// Start the roll near the end of the second so the number "settles" on change.
|
||||
const t = interpolate(secondProgress, [0.7, 1], [0, 1], {
|
||||
extrapolateLeft: "clamp",
|
||||
extrapolateRight: "clamp",
|
||||
easing: Easing.bezier(0.22, 1, 0.36, 1),
|
||||
});
|
||||
// continuous position between prev and value
|
||||
const shown = prev + (value - prev) * t;
|
||||
const offset = -shown * DIGIT_H;
|
||||
|
||||
return (
|
||||
<div style={{ height: DIGIT_H, width: 110, overflow: "hidden", position: "relative" }}>
|
||||
<div style={{ transform: `translateY(${offset}px)` }}>
|
||||
{[0, 1, 2, 3, 4, 5, 6, 7, 8, 9].map((n) => (
|
||||
<div
|
||||
key={n}
|
||||
style={{
|
||||
height: DIGIT_H,
|
||||
lineHeight: `${DIGIT_H}px`,
|
||||
textAlign: "center",
|
||||
fontSize: 140,
|
||||
fontWeight: 800,
|
||||
fontVariantNumeric: "tabular-nums",
|
||||
}}
|
||||
>
|
||||
{n}
|
||||
</div>
|
||||
))}
|
||||
</div>
|
||||
</div>
|
||||
);
|
||||
};
|
||||
|
||||
// Renders an mm:ss reel where each digit rolls.
|
||||
export const ReelClock: React.FC<{ durationInSeconds: number }> = ({ durationInSeconds }) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps } = useVideoConfig();
|
||||
|
||||
const elapsed = frame / fps;
|
||||
const remaining = Math.max(0, Math.ceil(durationInSeconds - elapsed));
|
||||
const prevRemaining = Math.max(0, remaining + 1); // value one second ago
|
||||
const secondProgress = elapsed % 1; // 0→1 within the current second
|
||||
|
||||
const digits = (s: number) =>
|
||||
[Math.floor(s / 600) % 10, Math.floor(s / 60) % 10, Math.floor(s / 10) % 10, s % 10];
|
||||
|
||||
const cur = digits(remaining);
|
||||
const prev = digits(prevRemaining);
|
||||
|
||||
return (
|
||||
<div style={{ display: "flex", gap: 8, color: "#fff", background: "#0b0d17", padding: 40 }}>
|
||||
{cur.map((d, i) => (
|
||||
<React.Fragment key={i}>
|
||||
{i === 2 && <div style={{ fontSize: 140, fontWeight: 800 }}>:</div>}
|
||||
<Reel value={d} prev={prev[i]} secondProgress={secondProgress} />
|
||||
</React.Fragment>
|
||||
))}
|
||||
</div>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
Notes:
|
||||
- Each digit only rolls when it actually changes; a digit whose `prev === value` sits still because `shown` stays constant.
|
||||
- A `9 → 0` rollover rolls *backward* through the strip (the cheap-looking part). To make tens/units roll the natural way on a carry, render an extra `0` glyph below the `9` and clamp — covered as an exercise; for most timers the backward roll at the tens boundary is unnoticeable at speed.
|
||||
- `ROLL_FRAMES` is unused above because the roll is tied to `secondProgress`; expose it instead if a fixed-length roll is preferred.
|
||||
|
||||
## Flip card (split-flap)
|
||||
|
||||
The top half of the old digit folds down over a static bottom half, revealing the new digit. Pure CSS keyframe version — fine for a live web page; for Remotion, drive the `rotateX` off `secondProgress` the same way as the reel so it stays deterministic.
|
||||
|
||||
```css
|
||||
.flip { position: relative; width: 110px; height: 160px; perspective: 600px; }
|
||||
.flip .card {
|
||||
position: absolute; inset: 0;
|
||||
display: flex; align-items: center; justify-content: center;
|
||||
font-size: 140px; font-weight: 800; font-variant-numeric: tabular-nums;
|
||||
backface-visibility: hidden;
|
||||
}
|
||||
.flip .top { clip-path: inset(0 0 50% 0); transform-origin: bottom; }
|
||||
.flip .bottom { clip-path: inset(50% 0 0 0); }
|
||||
/* on change, animate the leaf folding down */
|
||||
.flip .leaf {
|
||||
position: absolute; inset: 0; transform-origin: bottom;
|
||||
animation: fold .45s cubic-bezier(.36,0,.66,-0.2) forwards;
|
||||
}
|
||||
@keyframes fold {
|
||||
0% { transform: rotateX(0deg); }
|
||||
100% { transform: rotateX(-90deg); }
|
||||
}
|
||||
```
|
||||
|
||||
```js
|
||||
// swap the digit and replay the fold whenever the value changes
|
||||
function setFlip(el, newDigit) {
|
||||
if (el.dataset.v === String(newDigit)) return; // no change → no flip
|
||||
el.dataset.v = String(newDigit);
|
||||
el.querySelector(".top").textContent = newDigit;
|
||||
el.querySelector(".bottom").textContent = newDigit;
|
||||
const leaf = el.querySelector(".leaf");
|
||||
leaf.textContent = newDigit;
|
||||
leaf.style.animation = "none";
|
||||
void leaf.offsetWidth; // reflow → restart the animation
|
||||
leaf.style.animation = "";
|
||||
}
|
||||
```
|
||||
|
||||
Drive `setFlip` from the same `requestAnimationFrame` + `performance.now()` loop in the cookbook (only call it when the displayed second changes), so the flips fire on real time and never drift.
|
||||
|
||||
## Choosing
|
||||
|
||||
- **Launch / premium:** odometer reel, slow ease, large tabular digits, muted palette.
|
||||
- **Flash sale:** flip card for the mechanical urgency, or reel with a faster roll and a hot accent color on `mm:ss`.
|
||||
- **Stream "starting soon":** either works, but keep it calm — a single reel on `mm:ss` reads better than a busy split-flap for a 5–10 minute hold.
|
||||
Whitespace-only changes.
@@ -0,0 +1,148 @@
|
||||
---
|
||||
name: logo-animation
|
||||
description: This skill should be used when the user asks to "animate our logo", "make a logo intro/stinger", "build a brand reveal", "create an app splash animation", "loop a loader logo", "draw on our SVG mark", "morph an icon into the wordmark", or "sync the logo to a sound-logo". Covers web (SVG/Lottie), video, and app-splash delivery.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Logo Animation
|
||||
|
||||
Bring a brand mark to life with one clear idea, on-brand geometry, and a crisp settle. Animate for web (SVG/Lottie), video intros/outros, or app splash screens, and always ship a static end-frame fallback.
|
||||
|
||||
## When to use
|
||||
|
||||
- Build an intro stinger, sign-off, app splash, or loader idle loop.
|
||||
- Reveal a brand mark on a landing hero or sizzle reel.
|
||||
- Draw on an SVG mark, wipe/build it in piece by piece, or morph an icon into a wordmark.
|
||||
- Sync the final settle of the mark to a sound-logo beat.
|
||||
|
||||
## Core principle
|
||||
|
||||
Pick ONE technique per logo — draw OR build OR morph — not all at once. Reveals run 0.8–2.5s. Never distort brand geometry; scale and fade are safe, skew and squash usually are not. Respect clearspace (keep the mark's minimum margin clear of motion debris). Always provide a `prefers-reduced-motion` path that shows the final mark with no motion.
|
||||
|
||||
## Technique map
|
||||
|
||||
| Technique | Best for | Mechanism |
|
||||
|---|---|---|
|
||||
| Stroke draw-on | Line marks, monograms, signatures | `stroke-dashoffset` 1→0 |
|
||||
| Mask wipe / reveal | Filled shapes, gradients | animate a `<clipPath>` / `<mask>` rect |
|
||||
| Build-on (stagger) | Multi-piece marks, grids | per-group opacity+transform, staggered |
|
||||
| Morph | Icon ↔ wordmark, shape→logo | GSAP MorphSVG or matched path interpolation |
|
||||
| Idle loop | Loaders, ambient splash | seamless cyclic transform (rotate/pulse) |
|
||||
| Wordmark | Logotypes | per-letter mask/translate stagger |
|
||||
|
||||
## Quick start: SVG stroke draw-on
|
||||
|
||||
The fastest on-brand web reveal. Each path draws itself, then fills settle in.
|
||||
|
||||
```html
|
||||
<svg viewBox="0 0 200 80" id="logo">
|
||||
<path class="ink" d="M10 60 L40 20 L70 60" fill="none"
|
||||
stroke="#111" stroke-width="6" stroke-linecap="round"
|
||||
pathLength="1" stroke-dasharray="1" stroke-dashoffset="1"/>
|
||||
</svg>
|
||||
<style>
|
||||
@keyframes draw { to { stroke-dashoffset: 0; } }
|
||||
#logo .ink { animation: draw 1.1s cubic-bezier(.65,0,.35,1) forwards; }
|
||||
@media (prefers-reduced-motion: reduce) {
|
||||
#logo .ink { animation: none; stroke-dashoffset: 0; }
|
||||
}
|
||||
</style>
|
||||
```
|
||||
|
||||
`pathLength="1"` normalizes every path to length 1, so one `stroke-dasharray`/`offset` value works regardless of geometry. Stagger multiple strokes with `animation-delay`.
|
||||
|
||||
## Quick start: build-on with stagger (GSAP)
|
||||
|
||||
```js
|
||||
import gsap from "gsap";
|
||||
// each <g class="piece"> is one mark component
|
||||
gsap.from(".piece", {
|
||||
opacity: 0, scale: 0.85, transformOrigin: "50% 50%",
|
||||
duration: 0.5, ease: "back.out(1.6)", stagger: 0.08,
|
||||
onComplete: () => gsap.to("#mark", { scale: 1, duration: 0.2, ease: "power2.out" })
|
||||
});
|
||||
```
|
||||
|
||||
Group order matters: reveal the base/container first, then accents, then the wordmark last so the eye lands on the name.
|
||||
|
||||
## Sound-logo sync
|
||||
|
||||
When audio exists, the brand's final settle frame must land exactly on the sonic accent (the "ding"/whoosh peak). Decide the audio first, find the peak timecode, then time the reveal so the mark snaps to full opacity/scale on that frame. A 1.0s reveal landing a 1.2s mnemonic feels premium; a settle that floats free of the sound feels cheap. Add a tiny overshoot-and-settle (scale 1.04→1.00 over ~0.15s) on the beat for impact.
|
||||
|
||||
## Delivery targets
|
||||
|
||||
- **Web, interactive / hero**: inline SVG + CSS or GSAP. Smallest payload, crisp at any size, scriptable.
|
||||
- **Product / app, designer-made**: Lottie (`.json` exported from After Effects via Bodymovin). Use `lottie-web` or `@lottiefiles/dotlottie-web`; set `loop`, `autoplay`, and a poster fallback frame.
|
||||
- **Video intro/outro**: Remotion (React) or After Effects render to ProRes/H.264. Pre-render at 2x for crispness.
|
||||
|
||||
```js
|
||||
// Lottie delivery with reduced-motion guard
|
||||
import { DotLottie } from "@lottiefiles/dotlottie-web";
|
||||
const reduce = matchMedia("(prefers-reduced-motion: reduce)").matches;
|
||||
const dl = new DotLottie({
|
||||
canvas: document.querySelector("#logo-canvas"),
|
||||
src: "/logo.lottie", autoplay: !reduce, loop: false,
|
||||
});
|
||||
if (reduce) dl.addEventListener("load", () => dl.setFrame(dl.totalFrames - 1));
|
||||
```
|
||||
|
||||
## Output checklist
|
||||
|
||||
- Animated mark in the target format (SVG/Lottie/video).
|
||||
- Static final-frame fallback (SVG or PNG) for reduced-motion and posters.
|
||||
- One clear technique, reveal ≤2.5s, brand geometry intact, clearspace respected.
|
||||
- Sound-logo settle aligned to the audio peak (if audio present).
|
||||
|
||||
## Deliver & verify (standalone HTML)
|
||||
|
||||
> **Packaged helper** (`scripts/`): `scripts/seek-shot.sh anim.html 0 1.5 3` freezes the `?t=N` harness and screenshots each moment; `scripts/contact-sheet.sh sheet.png frame-*.png` tiles them for one-glance review. See `scripts/README.md`.
|
||||
|
||||
Logo animation spans web/SVG/Lottie, video, and app splash. **For web/SVG/Lottie delivery use this standalone-HTML loop; for a video logo sting, render via the remotion-video verify loop instead.**
|
||||
|
||||
For web/SVG/Lottie, ship **one `.html` file that opens directly in a browser** — inline SVG markup plus the reveal (CSS or one CDN GSAP `<script>`), self-contained. One file is the right tier.
|
||||
|
||||
**Output contract:**
|
||||
- One `.html` file: the inline SVG (or Lottie canvas) + the reveal in inline `<style>`/`<script>`.
|
||||
- One animation driver — the CSS `@keyframes` draw/build, or a single GSAP timeline; not both.
|
||||
- Include the freeze harness below, matched to the technique, so any moment can be screenshotted deterministically.
|
||||
|
||||
**Freeze harness — pin a frame for screenshots.** Match the mechanism to the technique:
|
||||
|
||||
```html
|
||||
<script>
|
||||
const t = new URLSearchParams(location.search).get("t");
|
||||
if (t !== null) {
|
||||
const T = parseFloat(t);
|
||||
// CSS stroke-draw / build-on:
|
||||
document.querySelectorAll("#logo *").forEach(el => {
|
||||
el.style.animationDelay = (-T) + "s";
|
||||
el.style.animationPlayState = "paused";
|
||||
});
|
||||
// GSAP timeline instead? → tl.pause(); tl.seek(T);
|
||||
// SMIL <animate>? → const s = document.querySelector("svg"); s.pauseAnimations(); s.setCurrentTime(T);
|
||||
// Lottie? → dl.setFrame(Math.round(T * fps));
|
||||
}
|
||||
window.__ready = true; // ready signal for headless wait
|
||||
</script>
|
||||
```
|
||||
|
||||
**Verify loop — render → freeze → screenshot → check:**
|
||||
1. Open the file frozen at start / mid / end of the reveal (≤2.5s): `…/logo.html?t=0`, `?t=<dur/2>`, `?t=<dur>`.
|
||||
2. Screenshot each frozen frame.
|
||||
3. Check **fidelity** (matches the mark, one clear technique) and **artifacts** — brand geometry undistorted, clearspace clear of motion debris, no half-drawn strokes left at the end frame, the static final frame is the true logo.
|
||||
4. **Iterate:** if any check fails (e.g. a half-drawn stroke at `?t=<dur>` or distorted geometry mid-reveal), fix the stroke-dashoffset/duration/easing, re-render the **same** `?t=N`, and re-screenshot to confirm the fix — loop until every frame is clean.
|
||||
|
||||
```bash
|
||||
npx playwright screenshot --wait-for-timeout=500 "file://$PWD/logo.html?t=1.1" frame-mid.png
|
||||
```
|
||||
|
||||
**Before you finish:**
|
||||
1. Opens standalone in a browser — no console errors, no missing CDN.
|
||||
2. One technique/driver; `?t=N` freezes the exact frame correctly.
|
||||
3. Screenshotted at start / mid / end — geometry intact, clearspace respected, final frame is the clean mark.
|
||||
4. `prefers-reduced-motion` honored (final static mark shown, no motion) + static end-frame fallback shipped.
|
||||
5. Reveal ≤2.5s; if audio present, the settle lands on the sound-logo peak.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/logo-patterns.md` — runnable code for every technique: stroke draw-on, mask wipe build-on, GSAP MorphSVG morph, seamless idle loop, Lottie delivery, plus timing tables (0.8–2.5s), brand clearspace rules, and reduced-motion fallbacks.
|
||||
@@ -0,0 +1,199 @@
|
||||
# Logo Animation Patterns — Cookbook
|
||||
|
||||
Self-contained, runnable recipes for every logo technique. Pick ONE per logo.
|
||||
|
||||
## Timing guidance
|
||||
|
||||
| Beat | Duration | Easing |
|
||||
|---|---|---|
|
||||
| Stroke draw-on (per path) | 0.6–1.2s | `cubic-bezier(.65,0,.35,1)` (easeInOutCubic) |
|
||||
| Mask wipe reveal | 0.5–0.9s | `cubic-bezier(.22,1,.36,1)` (easeOutQuint) |
|
||||
| Build-on per piece | 0.4–0.6s, stagger 0.06–0.10s | `back.out(1.5)` |
|
||||
| Morph | 0.6–1.0s | `power2.inOut` |
|
||||
| Final settle / overshoot | 0.12–0.20s | scale 1.04→1.00, `power2.out` |
|
||||
| Full reveal envelope | 0.8–2.5s total | — |
|
||||
| Idle loop cycle | 1.5–4s, infinite | `sine.inOut` / `linear` for spin |
|
||||
|
||||
Rule of thumb: longer marks (wordmarks, detailed monograms) sit near 2.0–2.5s; simple icons near 0.8–1.4s. Hold the settled mark at least 0.4s before any sign-off cut.
|
||||
|
||||
## Brand clearspace & geometry rules
|
||||
|
||||
- **Clearspace**: keep a margin equal to the mark's cap-height (or the brand's defined unit) clear of moving particles, wipes, or overshoot. Motion must not visually crop into the safe margin.
|
||||
- **No distortion**: scale uniformly, fade, translate, rotate-on-axis, and mask are safe. Avoid non-uniform `scaleX≠scaleY`, skew, and squash on the final mark — they read as off-brand. A squash-and-settle is acceptable ONLY as a sub-0.2s micro-overshoot that returns to 1.0.
|
||||
- **Color**: never animate brand color away from spec mid-reveal. Reveal in brand color, or grayscale→brand as an intentional, approved beat.
|
||||
- **Lockup integrity**: animate the symbol and wordmark with a clear hierarchy (symbol leads, wordmark follows) but land both in the exact locked relationship.
|
||||
|
||||
## 1. Stroke draw-on (SVG + CSS)
|
||||
|
||||
```html
|
||||
<svg viewBox="0 0 240 80" id="mark">
|
||||
<path class="ink" pathLength="1" fill="none" stroke="#0A84FF"
|
||||
stroke-width="7" stroke-linecap="round" stroke-linejoin="round"
|
||||
d="M12 64 L48 16 L84 64 M120 16 L120 64 M156 16 H204"/>
|
||||
</svg>
|
||||
<style>
|
||||
#mark .ink {
|
||||
stroke-dasharray: 1; stroke-dashoffset: 1;
|
||||
animation: draw 1.1s cubic-bezier(.65,0,.35,1) forwards;
|
||||
}
|
||||
@keyframes draw { to { stroke-dashoffset: 0; } }
|
||||
@media (prefers-reduced-motion: reduce){
|
||||
#mark .ink { animation: none; stroke-dashoffset: 0; }
|
||||
}
|
||||
</style>
|
||||
```
|
||||
|
||||
Draw-then-fill (line settles into a solid mark):
|
||||
|
||||
```css
|
||||
#mark .ink { fill: transparent; }
|
||||
@keyframes fill { to { fill: #0A84FF; stroke-width: 0; } }
|
||||
#mark .ink { animation: draw 1.1s cubic-bezier(.65,0,.35,1) forwards,
|
||||
fill .3s ease .9s forwards; }
|
||||
```
|
||||
|
||||
GSAP equivalent with per-path stagger via DrawSVGPlugin:
|
||||
|
||||
```js
|
||||
import gsap from "gsap";
|
||||
import { DrawSVGPlugin } from "gsap/DrawSVGPlugin";
|
||||
gsap.registerPlugin(DrawSVGPlugin);
|
||||
gsap.fromTo("#mark .ink",
|
||||
{ drawSVG: "0%" },
|
||||
{ drawSVG: "100%", duration: 1.0, ease: "power2.inOut", stagger: 0.15 });
|
||||
```
|
||||
|
||||
## 2. Mask wipe / build-on (clipPath)
|
||||
|
||||
Reveal a filled mark by sweeping a clip rect across it.
|
||||
|
||||
```html
|
||||
<svg viewBox="0 0 200 80">
|
||||
<defs>
|
||||
<clipPath id="wipe"><rect id="wiper" x="0" y="0" width="0" height="80"/></clipPath>
|
||||
</defs>
|
||||
<g clip-path="url(#wipe)">
|
||||
<!-- the full filled logo art here -->
|
||||
<path d="M10 10 H190 V70 H10 Z" fill="#111"/>
|
||||
</g>
|
||||
</svg>
|
||||
<style>
|
||||
#wiper { animation: wipe .8s cubic-bezier(.22,1,.36,1) forwards; }
|
||||
@keyframes wipe { to { width: 200px; } }
|
||||
@media (prefers-reduced-motion: reduce){ #wiper{ animation:none; width:200px; } }
|
||||
</style>
|
||||
```
|
||||
|
||||
Directional variants: animate `y`/`height` for a top-down build, or use a radial `<clipPath>` with a growing `<circle r>` for an iris reveal.
|
||||
|
||||
## 3. Build-on with stagger (GSAP timeline)
|
||||
|
||||
```js
|
||||
import gsap from "gsap";
|
||||
const tl = gsap.timeline({ defaults: { ease: "back.out(1.6)" } });
|
||||
tl.from("#base", { scale: .8, opacity: 0, transformOrigin: "50% 50%", duration: .5 })
|
||||
.from(".accent", { scale: .7, opacity: 0, transformOrigin: "50% 50%",
|
||||
duration: .45, stagger: 0.08 }, "-=0.2")
|
||||
.from(".word > *", { yPercent: 120, opacity: 0, duration: .4,
|
||||
stagger: 0.04, ease: "power3.out" }, "-=0.15")
|
||||
.to("#mark", { scale: 1.04, duration: .12, ease: "power2.out" })
|
||||
.to("#mark", { scale: 1.0, duration: .14, ease: "power2.inOut" });
|
||||
```
|
||||
|
||||
## 4. Morph (GSAP MorphSVG)
|
||||
|
||||
Morph requires two paths; MorphSVG handles differing point counts.
|
||||
|
||||
```js
|
||||
import gsap from "gsap";
|
||||
import { MorphSVGPlugin } from "gsap/MorphSVGPlugin";
|
||||
gsap.registerPlugin(MorphSVGPlugin);
|
||||
// #shape: a simple start shape (e.g. circle path) -> #logoPath: the brand mark
|
||||
gsap.to("#shape", {
|
||||
duration: .9, ease: "power2.inOut",
|
||||
morphSVG: { shape: "#logoPath", shapeIndex: "auto" }
|
||||
});
|
||||
```
|
||||
|
||||
No-plugin fallback (Flubber for interpolation + requestAnimationFrame):
|
||||
|
||||
```js
|
||||
import { interpolate } from "flubber";
|
||||
const from = "M10,10 ...", to = "M0,0 ...";
|
||||
const lerp = interpolate(from, to, { maxSegmentLength: 2 });
|
||||
const path = document.querySelector("#shape");
|
||||
const dur = 900, t0 = performance.now();
|
||||
(function frame(now){
|
||||
const k = Math.min(1, (now - t0) / dur);
|
||||
const e = k < .5 ? 2*k*k : 1 - Math.pow(-2*k+2, 2)/2; // easeInOutQuad
|
||||
path.setAttribute("d", lerp(e));
|
||||
if (k < 1) requestAnimationFrame(frame);
|
||||
})(t0);
|
||||
```
|
||||
|
||||
## 5. Seamless idle loop (loader)
|
||||
|
||||
Loop must have identical start and end frames — design the cycle so frame 0 == frame N.
|
||||
|
||||
```css
|
||||
#loader { animation: spin 1.4s linear infinite, pulse 2.8s ease-in-out infinite; }
|
||||
@keyframes spin { to { transform: rotate(360deg); } }
|
||||
@keyframes pulse { 0%,100%{ opacity:.85 } 50%{ opacity:1 } }
|
||||
@media (prefers-reduced-motion: reduce){
|
||||
#loader { animation: pulse 2.8s ease-in-out infinite; } /* drop rotation */
|
||||
}
|
||||
```
|
||||
|
||||
GSAP `yoyo` loop for a breathing mark:
|
||||
|
||||
```js
|
||||
gsap.to("#mark", { scale: 1.03, duration: 1.4, ease: "sine.inOut",
|
||||
yoyo: true, repeat: -1 });
|
||||
```
|
||||
|
||||
## 6. Lottie delivery (designer-made)
|
||||
|
||||
Export from After Effects via Bodymovin to `.json`, or compressed `.lottie`.
|
||||
|
||||
```html
|
||||
<canvas id="logo-canvas" width="512" height="512"></canvas>
|
||||
<script type="module">
|
||||
import { DotLottie } from "https://cdn.jsdelivr.net/npm/@lottiefiles/dotlottie-web/+esm";
|
||||
const reduce = matchMedia("(prefers-reduced-motion: reduce)").matches;
|
||||
const dl = new DotLottie({
|
||||
canvas: document.querySelector("#logo-canvas"),
|
||||
src: "/logo.lottie", autoplay: !reduce, loop: false, speed: 1,
|
||||
});
|
||||
// Reduced motion: jump to the final settled frame, no playback.
|
||||
if (reduce) dl.addEventListener("load", () => dl.setFrame(dl.totalFrames - 1));
|
||||
// Sound-logo sync: fire audio when the settle frame is reached.
|
||||
dl.addEventListener("frame", ({ currentFrame }) => {
|
||||
if (currentFrame >= SETTLE_FRAME && !audioFired) { audio.play(); audioFired = true; }
|
||||
});
|
||||
</script>
|
||||
```
|
||||
|
||||
Classic `lottie-web` (SVG renderer, sharpest for vector logos):
|
||||
|
||||
```js
|
||||
import lottie from "lottie-web";
|
||||
const anim = lottie.loadAnimation({
|
||||
container: document.getElementById("logo"),
|
||||
renderer: "svg", loop: false, autoplay: !reduce, path: "/logo.json",
|
||||
});
|
||||
if (reduce) anim.goToAndStop(anim.totalFrames - 1, true);
|
||||
```
|
||||
|
||||
Lottie tips: keep layers/effects Bodymovin-supported (no native blurs/expressions unless baked), name layers cleanly, and request a poster PNG of the final frame from the designer for the static fallback.
|
||||
|
||||
## 7. Reduced-motion fallback (always ship)
|
||||
|
||||
Every delivery must degrade to the static, fully-revealed mark:
|
||||
|
||||
- **CSS/SVG**: `@media (prefers-reduced-motion: reduce)` sets the final state (`stroke-dashoffset:0`, `width:full`, `opacity:1`) with `animation:none`.
|
||||
- **Lottie**: `goToAndStop(totalFrames-1)` / `setFrame(last)`.
|
||||
- **Video**: serve a poster image of the last frame; only autoplay motion when motion is allowed.
|
||||
- **JS gate**: branch on `matchMedia("(prefers-reduced-motion: reduce)").matches` before building any timeline.
|
||||
|
||||
---
|
||||
Match the reveal to the delivery target and a logo lands across web, video, and app. Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=motion-design-skills&utm_content=skill_footer&utm_term=logo-animation)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -1,24 +0,0 @@
|
||||
# Lower Thirds
|
||||
|
||||
Build a clean, branded lower third — the name/role caption that lives in the bottom band of the frame — that animates in, holds long enough to read, and exits without stealing the shot. The craft is restraint: a crisp reveal, a strong type hierarchy, correct safe-area placement, and one template that drives a whole roster of speakers. For creators producing interviews, panels, podcasts, and broadcast-style titles.
|
||||
|
||||
## When it activates
|
||||
|
||||
- "Make an animated lower third" or "add lower thirds."
|
||||
- "Create a name tag/name lower third" or "do a speaker name overlay."
|
||||
- "Build broadcast titles" or "animate a title/name/role caption."
|
||||
- "Template lower thirds across a list of names."
|
||||
|
||||
## Example prompts
|
||||
|
||||
- "Add a lower third with the speaker's name and role, branded to my colors."
|
||||
- "Build a staggered slide-in name plate for a podcast guest."
|
||||
- "Template one lower-third design across a roster of 12 panelists."
|
||||
|
||||
## What's inside
|
||||
|
||||
- `SKILL.md` — enter/exit animation, in/out timing and dwell, safe-area conventions, and brand styling.
|
||||
- `references/lower-third-component.md` — a reusable lower-third component.
|
||||
|
||||
---
|
||||
Part of **[Short-Form Video Skills](../../)** · Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=tiktok-video-skills&utm_content=skill_footer&utm_term=lower-thirds)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -1,146 +0,0 @@
|
||||
---
|
||||
name: lower-thirds
|
||||
description: This skill should be used when the user asks to "make an animated lower third", "add lower thirds", "create a name tag/name lower third", "build broadcast titles", "animate a title/name/role caption", "do a speaker name overlay", or "template lower thirds across a list of names". Covers enter/exit animation (slide/wipe/mask reveal), in/out timing & dwell, safe-area & broadcast conventions, name-vs-role type hierarchy, brand styling, and rendering one design across many names.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Lower Thirds
|
||||
|
||||
Build a clean, branded lower third — the name/role caption that lives in the bottom band of the frame — that animates in, holds long enough to read, and exits without stealing the shot. The craft is restraint: a crisp reveal, a strong type hierarchy, correct safe-area placement, and one template that drives a whole roster of speakers.
|
||||
|
||||
## When to use
|
||||
|
||||
- Speaker/interview name tags, panel rosters, talking-head captions.
|
||||
- Broadcast-style titles, news straps, podcast/webinar name plates.
|
||||
- Templating one design across a list of names/titles → many clips (the payoff).
|
||||
|
||||
## Anatomy
|
||||
|
||||
A lower third is two text lines plus a graphic bed, placed in the lower band:
|
||||
|
||||
| Element | Role | Type treatment |
|
||||
|---|---|---|
|
||||
| **Name** (primary) | Who | Largest, heaviest, tightest tracking |
|
||||
| **Role/title** (secondary) | What/where | ~45–60% of name size, lighter weight, often a brand accent color |
|
||||
| **Bar / panel / accent** | Reads against any background | Solid or 70–85% opacity; an accent rule ties to brand |
|
||||
|
||||
Keep it to two lines. A name line + role line is the whole job — extra lines erode the hierarchy and crowd the safe area.
|
||||
|
||||
## Three rules that prevent cheap-looking titles
|
||||
|
||||
1. **Never hard-cut a title on or off.** A bare appear/disappear reads as a mistake. Even a 12-frame fade or slide makes it feel produced.
|
||||
2. **Stagger, don't slam.** Animate the bar first, then the name, then the role, each offset by 3–6 frames. Everything arriving at once feels flat; a short cascade feels designed.
|
||||
3. **Hold long enough to read twice.** Roughly 3–5s on screen (read it aloud twice). Too short and it's missed; too long and it nags. Animate in/out, never mid-segment.
|
||||
|
||||
## Enter & exit animations
|
||||
|
||||
Pick one motion language and reuse it. The most reliable, premium-reading reveals:
|
||||
|
||||
| Reveal | Feel | How |
|
||||
|---|---|---|
|
||||
| **Slide + fade** | Clean, default | Translate in ~24–40px while fading; ease-out on enter |
|
||||
| **Wipe bar** | Broadcast | A bar scales/sweeps in (`scaleX`/clip), text fades behind it |
|
||||
| **Mask reveal** | Premium | Clip text to a moving rectangle so letters "wipe on" |
|
||||
|
||||
Use asymmetric easing: a snappy `ease-out` (`cubic-bezier(.22,1,.36,1)`) on enter, a quicker `ease-in` on exit. Exit faster than you entered (~⅔ the duration) so the title gets out of the way.
|
||||
|
||||
```css
|
||||
/* Mask reveal: text wipes on left→right, no plugin needed */
|
||||
.lt-name { clip-path: inset(0 100% 0 0); animation: wipe .5s cubic-bezier(.22,1,.36,1) forwards; }
|
||||
@keyframes wipe { to { clip-path: inset(0 0 0 0); } }
|
||||
```
|
||||
|
||||
## In/out timing & dwell
|
||||
|
||||
Think in three phases. Budget frames, not vibes (at 30fps):
|
||||
|
||||
| Phase | Budget | Note |
|
||||
|---|---|---|
|
||||
| Enter | 12–18f (0.4–0.6s) | Staggered: bar → name → role |
|
||||
| Hold | 90–150f (3–5s) | The readable dwell; scale with line length |
|
||||
| Exit | 8–12f (0.27–0.4s) | Faster than enter |
|
||||
|
||||
In code, drive every state from the current frame (or a paused timeline), not wall-clock time — a video renderer demands deterministic frames. Express enter/hold/exit as a single progress curve so the same component works live in a browser and frame-by-frame in a renderer.
|
||||
|
||||
## Safe area & broadcast conventions
|
||||
|
||||
The lower third must survive every screen and platform UI. Compose inside the **title-safe area** (inner ~90% / margin 5%) and respect platform chrome:
|
||||
|
||||
| Frame | Anchor | Avoid |
|
||||
|---|---|---|
|
||||
| 16:9 (1920×1080) | Left edge at ~5% margin; baseline ~80–85% down | Outer 5% (title-safe), action-safe outer 10% |
|
||||
| 9:16 (1080×1920) | Same left margin; **lift the band up** | Bottom ~18% (captions/UI), top ~12% |
|
||||
| 1:1 (1080×1080) | Lower band, center-left | Outer 5% |
|
||||
|
||||
Anchor to the **left** and grow rightward so names of different lengths stay pinned and never re-center awkwardly. Keep the role line within the same left margin as the name. For 9:16, do not park the title at the very bottom — platform captions and buttons sit there.
|
||||
|
||||
## Type hierarchy & brand styling
|
||||
|
||||
Hierarchy is the whole design. Make the name unmistakably dominant:
|
||||
|
||||
- **Size contrast** ≥ 1.6× (name vs role). If unsure, push the name bigger, not the role.
|
||||
- **Weight + color**: name in bold/near-white; role in a lighter weight and a **brand accent** color (one accent, used consistently).
|
||||
- **Contrast for legibility**: text rides over unknown footage. Use a solid/semi-opaque bar, a subtle scrim, or a soft text shadow so it reads over both bright and dark plates.
|
||||
- **Brand lock**: pin font, accent color, bar style, and corner radius in one theme object so every title across a series matches.
|
||||
|
||||
## Templating one design across a list
|
||||
|
||||
The reason to build this in code: change the text/data, re-export — no re-animating. Make name/role/theme **props**, never hardcoded, then map a roster to instances.
|
||||
|
||||
```jsx
|
||||
const ROSTER = [
|
||||
{ name: "Ada Lovelace", role: "Founder, Analytical Engines" },
|
||||
{ name: "Grace Hopper", role: "Rear Admiral, US Navy" },
|
||||
{ name: "Alan Turing", role: "Mathematician" },
|
||||
];
|
||||
// One <LowerThird> design; map the list → many instances (sequence them in a renderer)
|
||||
{ROSTER.map((p, i) => <LowerThird key={p.name} {...p} theme={BRAND} startFrame={i * 150} />)}
|
||||
```
|
||||
|
||||
In a renderer, make the roster an input prop and render once per row (`--props`) to emit one branded clip per person — same animation, same brand, only the text changes. See `references/lower-third-component.md` for the full runnable component and batch script.
|
||||
|
||||
## Output checklist
|
||||
|
||||
- Two lines only: dominant name, secondary role (≥1.6× size contrast).
|
||||
- Animated in and out (no hard cut); enter staggered bar→name→role; exit faster.
|
||||
- Hold 3–5s (readable twice); motion only at the ends, never mid-segment.
|
||||
- Left-anchored inside title-safe; 9:16 band lifted above the bottom-18% UI zone.
|
||||
- Text legible over any plate (bar/scrim/shadow); one brand accent throughout.
|
||||
- Name/role/theme are props — one template renders the whole roster.
|
||||
|
||||
## Deliver & verify (rendered stills → MP4)
|
||||
|
||||
> **Packaged helper** (`scripts/`): tile your stills with `scripts/contact-sheet.sh sheet.png f-hook.png f-mid.png f-end.png`, then assert the encode with `scripts/probe-mp4.sh out.mp4 [WxH] [fps]`. See `scripts/README.md`.
|
||||
|
||||
The lower third ships as a Remotion composition (`<Composition>` + zod `schema` + `defaultProps`, name/role/theme as props) — enter/dwell/exit all driven by `useCurrentFrame()`, never `Date.now()` / `Math.random()` / timers. Deliverable = `out/*.mp4` + the project (re-render per roster row). 9:16 vertical (1080×1920) is the default.
|
||||
|
||||
**Verify loop — render stills → inspect → encode.** Sample the three phases so you catch a stagger or safe-area bug before encoding (or batching the whole roster).
|
||||
|
||||
```bash
|
||||
# Stills at enter / dwell / exit — WITH SHIPPED PROPS (a real roster row, not defaults)
|
||||
npx remotion still LowerThird out/f-enter.png --frame=10 --props='{"name":"Ada Lovelace","role":"Founder, Analytical Engines"}'
|
||||
npx remotion still LowerThird out/f-dwell.png --frame=75 --props='{"name":"Ada Lovelace","role":"Founder, Analytical Engines"}'
|
||||
npx remotion still LowerThird out/f-exit.png --frame=N --props='{"name":"Ada Lovelace","role":"Founder, Analytical Engines"}'
|
||||
|
||||
# Inspect each PNG — focus the DWELL frame (fully on screen):
|
||||
# - hierarchy holds: name dominant (>=1.6x role), role in brand accent; both fully revealed, not mid-wipe
|
||||
# - brand lock: font, accent color, bar/scrim correct; legible over the plate
|
||||
# - 9:16 safe area: band LIFTED above the bottom ~20-35% UI zone, clear of top ~12% and the right action rail; left-anchored inside title-safe
|
||||
# - enter/exit frames show partial reveal (animation actually runs), not a hard cut
|
||||
|
||||
npx remotion render LowerThird out/lower-third.mp4 --props='{"name":"Ada Lovelace","role":"Founder, Analytical Engines"}'
|
||||
npx remotion render LowerThird out/demo.gif --codec=gif # README proof clip
|
||||
```
|
||||
|
||||
**Batch roster**: verify ONE row via stills *before* rendering all names — catch a layout/overflow bug once, not N times. Use `npx remotion compositions` for `durationInFrames`/`fps` to pick the dwell + exit frames.
|
||||
|
||||
**Before you finish:**
|
||||
1. Stills render cleanly at enter, dwell, and exit — no missing font/brand assets.
|
||||
2. At the dwell frame: name dominant over role, brand accent + bar correct, fully (not partly) revealed.
|
||||
3. The band is inside the 9:16 safe area — lifted above the bottom UI zone, clear of top and right rail — at every checked frame.
|
||||
4. Frame-driven only — no `Date.now()` / `Math.random()` / timers; enter/exit show motion, never a hard cut.
|
||||
5. A real roster row (not `defaultProps`) renders correctly; MP4 encoded, GIF optional.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/lower-third-component.md` — a complete runnable lower-third: a Remotion component with staggered spring enter, readable hold, and faster exit driven purely by frame; a typed theme object for brand lock; the list→many-instances sequencing pattern; a per-row batch render script; plus a dependency-free CSS/HTML version of the same design with mask-reveal and wipe-bar variants.
|
||||
@@ -1,268 +0,0 @@
|
||||
# Lower-Third Component — runnable code
|
||||
|
||||
A complete, production-grade lower third you can drop in. Two paths:
|
||||
|
||||
1. **Remotion** (frame-deterministic, renders to MP4, batches over a roster).
|
||||
2. **CSS/HTML** (no dependencies, for the web or a quick preview) with mask-reveal and wipe-bar variants.
|
||||
|
||||
Both share the same design language: a left-anchored bar, a dominant name, a brand-accent role line, a staggered enter, a readable hold, and a faster exit.
|
||||
|
||||
---
|
||||
|
||||
## 1. Remotion component (props-driven, frame-deterministic)
|
||||
|
||||
The animation is a pure function of `frame`, so it renders identically every time and stays in sync with the renderer. Enter staggers the bar, name, then role; exit is computed from the end of the clip so the title leaves faster than it arrived.
|
||||
|
||||
```tsx
|
||||
// LowerThird.tsx
|
||||
import {
|
||||
AbsoluteFill, Sequence, interpolate, spring,
|
||||
useCurrentFrame, useVideoConfig,
|
||||
} from "remotion";
|
||||
|
||||
// --- Brand lock: change once, every title matches -------------------------
|
||||
export type Theme = {
|
||||
font: string;
|
||||
bar: string; // panel/bar color
|
||||
name: string; // primary text color
|
||||
role: string; // accent color for the role line
|
||||
radius: number;
|
||||
marginPct: number; // left/safe margin as % of width
|
||||
};
|
||||
export const BRAND: Theme = {
|
||||
font: "Inter, system-ui, sans-serif",
|
||||
bar: "rgba(12,14,22,0.86)",
|
||||
name: "#FFFFFF",
|
||||
role: "#6EE7F0", // single brand accent
|
||||
radius: 10,
|
||||
marginPct: 5,
|
||||
};
|
||||
|
||||
export type LowerThirdProps = {
|
||||
name: string;
|
||||
role: string;
|
||||
theme?: Theme;
|
||||
holdSeconds?: number; // readable dwell, default 4s
|
||||
};
|
||||
|
||||
// Staggered enter via spring; exit via a quick fade/slide near the end.
|
||||
const useEnterExit = (delayFrames: number, holdSeconds: number) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps, durationInFrames } = useVideoConfig();
|
||||
|
||||
// ENTER: spring 0→1, offset by delayFrames (the cascade)
|
||||
const enter = spring({
|
||||
frame: frame - delayFrames,
|
||||
fps,
|
||||
config: { damping: 200, stiffness: 180, mass: 0.6 },
|
||||
});
|
||||
|
||||
// EXIT: faster than enter — last 10 frames of the clip
|
||||
const exitStart = durationInFrames - 10;
|
||||
const exit = interpolate(frame, [exitStart, durationInFrames], [1, 0], {
|
||||
extrapolateLeft: "clamp",
|
||||
extrapolateRight: "clamp",
|
||||
});
|
||||
|
||||
const opacity = enter * exit;
|
||||
const x = interpolate(enter, [0, 1], [-32, 0]); // slide in from left
|
||||
return { opacity, x };
|
||||
};
|
||||
|
||||
const Line: React.FC<{
|
||||
delay: number; holdSeconds: number; children: React.ReactNode;
|
||||
size: number; weight: number; color: string; tracking?: number;
|
||||
}> = ({ delay, holdSeconds, children, size, weight, color, tracking = 0 }) => {
|
||||
const { opacity, x } = useEnterExit(delay, holdSeconds);
|
||||
return (
|
||||
<div style={{
|
||||
opacity,
|
||||
transform: `translateX(${x}px)`,
|
||||
fontSize: size, fontWeight: weight, color,
|
||||
letterSpacing: tracking, lineHeight: 1.15, whiteSpace: "nowrap",
|
||||
}}>
|
||||
{children}
|
||||
</div>
|
||||
);
|
||||
};
|
||||
|
||||
export const LowerThird: React.FC<LowerThirdProps> = ({
|
||||
name, role, theme = BRAND, holdSeconds = 4,
|
||||
}) => {
|
||||
const { width, height } = useVideoConfig();
|
||||
const bar = useEnterExit(0, holdSeconds); // bar leads the cascade
|
||||
|
||||
const left = (width * theme.marginPct) / 100;
|
||||
const baseline = height * 0.80; // sit ~80% down, inside title-safe
|
||||
|
||||
return (
|
||||
<AbsoluteFill style={{ fontFamily: theme.font }}>
|
||||
<div style={{
|
||||
position: "absolute", left, top: baseline,
|
||||
opacity: bar.opacity,
|
||||
transform: `translateX(${bar.x}px)`,
|
||||
}}>
|
||||
{/* panel behind text guarantees legibility over any plate */}
|
||||
<div style={{
|
||||
background: theme.bar, borderRadius: theme.radius,
|
||||
padding: "14px 22px 16px",
|
||||
borderLeft: `4px solid ${theme.role}`, // accent rule ties to brand
|
||||
backdropFilter: "blur(6px)",
|
||||
}}>
|
||||
<Line delay={4} holdSeconds={holdSeconds} size={46} weight={800} color={theme.name} tracking={-0.5}>
|
||||
{name}
|
||||
</Line>
|
||||
<Line delay={8} holdSeconds={holdSeconds} size={26} weight={500} color={theme.role}>
|
||||
{role}
|
||||
</Line>
|
||||
</div>
|
||||
</div>
|
||||
</AbsoluteFill>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
Why each choice matters:
|
||||
- **`spring` per line, offset by `delay`** gives the bar→name→role cascade. Same offsets everywhere keep titles consistent.
|
||||
- **Exit derived from `durationInFrames`** means any clip length exits cleanly without per-instance tuning, and it's quicker than the enter (10f vs the spring's settle).
|
||||
- **`whiteSpace: nowrap` + left anchor** keeps long and short names pinned to the same edge — no awkward re-centering across a roster.
|
||||
- **`borderLeft` accent + semi-opaque panel** is the brand tie plus the legibility guarantee over unknown footage.
|
||||
|
||||
### Roster → many instances (one design, a whole panel)
|
||||
|
||||
Sequence the same component down a list. Each person gets a slot; only the text changes.
|
||||
|
||||
```tsx
|
||||
// Roster.tsx
|
||||
import { Series } from "remotion";
|
||||
import { LowerThird, BRAND } from "./LowerThird";
|
||||
|
||||
const ROSTER = [
|
||||
{ name: "Ada Lovelace", role: "Founder, Analytical Engines" },
|
||||
{ name: "Grace Hopper", role: "Rear Admiral, US Navy" },
|
||||
{ name: "Alan Turing", role: "Mathematician" },
|
||||
];
|
||||
|
||||
export const Roster: React.FC = () => (
|
||||
<Series>
|
||||
{ROSTER.map((p) => (
|
||||
<Series.Sequence key={p.name} durationInFrames={150 /* ~5s @30fps */}>
|
||||
<LowerThird name={p.name} role={p.role} theme={BRAND} />
|
||||
</Series.Sequence>
|
||||
))}
|
||||
</Series>
|
||||
);
|
||||
```
|
||||
|
||||
### Batch: one branded clip per person
|
||||
|
||||
Make the roster an input prop and render once per row, so each speaker becomes a standalone overlay clip — identical animation and brand, only the text differs.
|
||||
|
||||
```bash
|
||||
#!/usr/bin/env bash
|
||||
# render one transparent-friendly lower-third clip per person in roster.json
|
||||
# roster.json: [{"name":"Ada Lovelace","role":"Founder, Analytical Engines"}, ...]
|
||||
set -euo pipefail
|
||||
mkdir -p out
|
||||
jq -c '.[]' roster.json | while read -r row; do
|
||||
slug=$(echo "$row" | jq -r '.name' | tr '[:upper:] ' '[:lower:]-')
|
||||
npx remotion render LowerThird "out/${slug}.mp4" --props="$row"
|
||||
done
|
||||
```
|
||||
|
||||
For overlay-ready output with transparency, render the `LowerThird` composition (transparent background) to a codec that keeps alpha, e.g. `--codec=prores --prores-profile=4444`, or `--image-format=png` for a PNG sequence.
|
||||
|
||||
---
|
||||
|
||||
## 2. CSS / HTML version (no dependencies)
|
||||
|
||||
The same design without a framework — for the web, OBS browser sources, or a fast look-test. Includes both a **slide+fade** default and a **mask-reveal** name, plus a **wipe-bar** variant.
|
||||
|
||||
```html
|
||||
<div class="lt" style="--accent:#6EE7F0">
|
||||
<div class="lt-bar">
|
||||
<div class="lt-name">Ada Lovelace</div>
|
||||
<div class="lt-role">Founder, Analytical Engines</div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<style>
|
||||
:root { font-family: Inter, system-ui, sans-serif; }
|
||||
|
||||
.lt {
|
||||
position: absolute; left: 5%; top: 80%; /* title-safe, left-anchored */
|
||||
}
|
||||
.lt-bar {
|
||||
display: inline-block;
|
||||
background: rgba(12,14,22,0.86);
|
||||
backdrop-filter: blur(6px);
|
||||
border-left: 4px solid var(--accent);
|
||||
border-radius: 10px;
|
||||
padding: 14px 22px 16px;
|
||||
/* bar leads the cascade: slide + fade in, then a faster fade out */
|
||||
animation: lt-in .5s cubic-bezier(.22,1,.36,1) both,
|
||||
lt-out .3s ease-in 4.5s forwards; /* hold ~4s, exit faster */
|
||||
}
|
||||
.lt-name {
|
||||
font-size: 46px; font-weight: 800; color: #fff; letter-spacing: -.5px;
|
||||
line-height: 1.15; white-space: nowrap;
|
||||
/* MASK REVEAL: letters wipe on left→right, staggered after the bar */
|
||||
clip-path: inset(0 100% 0 0);
|
||||
animation: lt-wipe .5s cubic-bezier(.22,1,.36,1) .12s forwards;
|
||||
}
|
||||
.lt-role {
|
||||
font-size: 26px; font-weight: 500; color: var(--accent);
|
||||
line-height: 1.15; white-space: nowrap;
|
||||
opacity: 0;
|
||||
animation: lt-fade .4s ease-out .22s forwards; /* role arrives last */
|
||||
}
|
||||
|
||||
@keyframes lt-in { from { opacity: 0; transform: translateX(-32px); }
|
||||
to { opacity: 1; transform: none; } }
|
||||
@keyframes lt-out { to { opacity: 0; transform: translateX(-16px); } }
|
||||
@keyframes lt-wipe { to { clip-path: inset(0 0 0 0); } }
|
||||
@keyframes lt-fade { to { opacity: 1; } }
|
||||
|
||||
/* Mobile/9:16: lift the band above the bottom UI/caption zone */
|
||||
@media (max-aspect-ratio: 1/1) { .lt { top: 70%; } }
|
||||
</style>
|
||||
```
|
||||
|
||||
### Wipe-bar variant (broadcast feel)
|
||||
|
||||
Swap the bar animation so a colored bar sweeps in and the text rides behind it:
|
||||
|
||||
```css
|
||||
.lt-bar.wipe {
|
||||
position: relative; overflow: hidden;
|
||||
animation: lt-in .45s cubic-bezier(.22,1,.36,1) both,
|
||||
lt-out .3s ease-in 4.5s forwards;
|
||||
}
|
||||
.lt-bar.wipe::before {
|
||||
content: ""; position: absolute; inset: 0;
|
||||
background: var(--accent);
|
||||
transform-origin: left;
|
||||
animation: lt-sweep .55s cubic-bezier(.76,0,.24,1) both; /* sweep across, then reveal */
|
||||
}
|
||||
@keyframes lt-sweep { 0% { transform: scaleX(0); } 60% { transform: scaleX(1); }
|
||||
100% { transform: scaleX(0); transform-origin: right; } }
|
||||
```
|
||||
|
||||
The CSS clock (`animation`) is fine for live/web playback. For a deterministic MP4, prefer the Remotion path above — a frame renderer and CSS timers disagree, which causes flicker or desync at export.
|
||||
|
||||
---
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Need | Knob |
|
||||
|---|---|
|
||||
| Slower/faster reveal | enter duration 12–18f / `.5s` |
|
||||
| Longer dwell | `holdSeconds` / the `4.5s` exit delay |
|
||||
| Bigger name dominance | name `fontSize` (keep ≥1.6× the role) |
|
||||
| Rebrand everything | the `BRAND` theme object / `--accent` |
|
||||
| 9:16 placement | `baseline 0.80→0.70` / the media query |
|
||||
|
||||
---
|
||||
## Built by the team behind iart.ai
|
||||
|
||||
This skill is part of an open motion-graphics collection from iart.ai — the AI motion agent that turns data, scripts, and designs into editable motion graphics (Remotion → MP4). If you'd rather not hand-build this, iart.ai can generate branded lower-thirds from a list of names/titles, one-click — change the text/data and re-export. → [iart.ai](https://iart.ai/?utm_source=github&utm_medium=reference&utm_campaign=tiktok-video-skills&utm_content=ref_footer&utm_term=lower-thirds)
|
||||
Whitespace-only changes.
@@ -0,0 +1,114 @@
|
||||
---
|
||||
name: motion-art-direction
|
||||
description: This skill should be used when the user asks to "direct a motion graphics video", "do creative/art direction for a video", "define a motion language", "set the tone and pacing", "make the animation feel consistent", "decide what to animate (and what not to)", "turn a brief/brand into motion choices", or needs the senior creative judgment that sits above specific techniques (gsap/lottie/shader). Covers motion language, tone & energy, pacing/rhythm, motion hierarchy, restraint, and cross-shot consistency.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Motion Direction
|
||||
|
||||
Make the creative decisions for a motion piece before anyone animates a frame. Direction is the judgment layer: it turns a brief and a brand into a small, repeatable set of motion rules — one easing vocabulary, one timing scale, one transition family — so every shot feels like it came from the same hand. Good direction is mostly subtraction: deciding what NOT to move.
|
||||
|
||||
## When to use
|
||||
|
||||
- Starting any piece — explainer, launch film, UI demo, data story, loop — and the look/feel isn't decided yet.
|
||||
- Animation feels "busy", "cheap", or "inconsistent" and the cause is decisions, not craft.
|
||||
- Orchestrating other skills (gsap/lottie/shader/three.js): direction sets the rules they execute.
|
||||
- Translating a brand or brief into directable, written choices a team (or an agent) can follow.
|
||||
|
||||
## The one rule
|
||||
|
||||
**Define the motion language first, then animate to it.** A motion language is a short spec — pick *one* of each: easing family, base timing unit, transition family, stagger rhythm, motion intensity. Everything in the piece obeys it. Consistency reads as confidence; variety reads as noise. When in doubt, reuse, don't invent.
|
||||
|
||||
## Direction workflow
|
||||
|
||||
| Step | Decision | Output |
|
||||
|---|---|---|
|
||||
| 1. Brief | Audience, one message, platform, duration, mood (3 adjectives) | One-line creative intent |
|
||||
| 2. Tone & energy | Where it sits on calm↔kinetic and soft↔sharp | A point on the matrix below |
|
||||
| 3. Motion language | Lock the spec table | The motion-language spec (filled) |
|
||||
| 4. Hierarchy | Rank what moves: hero / support / texture | Motion hierarchy list |
|
||||
| 5. Restraint pass | Cut motion that doesn't carry meaning | List of what stays static |
|
||||
| 6. Shot direction | Note intent per shot, not keyframes | Direction notes (see template) |
|
||||
| 7. Consistency check | Audit against the spec | Pass/fail per shot |
|
||||
|
||||
## Tone & energy matrix
|
||||
|
||||
Place the piece on two axes; the cell dictates the motion language defaults.
|
||||
|
||||
| | Soft (organic, eased) | Sharp (precise, snappy) |
|
||||
|---|---|---|
|
||||
| **Calm** | Luxury, wellness, editorial — slow, long eases, generous holds, minimal stagger | Premium tech, fintech — deliberate, clean linear-ish moves, tight but unhurried |
|
||||
| **Kinetic** | Playful, lifestyle, kids — bouncy/overshoot, springy, loose timing | Sports, hype, gaming — fast cuts, hard snaps, aggressive overshoot on beats |
|
||||
|
||||
Pick one cell and commit. Mixing cells within a piece is the most common cause of "inconsistent" motion.
|
||||
|
||||
## Motion personality (named presets)
|
||||
|
||||
The tone matrix tells you *where* the piece sits; a **motion personality** is the named, numeric preset you hand to whoever animates. Pick ONE per project and apply it everywhere. This is shared vocabulary — other skills (e.g. `gsap-web`, `micro-interaction`, the video packs) can just say "use the Premium personality."
|
||||
|
||||
| Personality | Duration | Signature easing | Overshoot | Reads as |
|
||||
|---|---|---|---|---|
|
||||
| **Playful** | 150–300ms | `ease-out-back` / `cubic-bezier(0.34,1.56,0.64,1)` | 10–20% | fun, bouncy, whimsical |
|
||||
| **Premium** | 350–600ms | `cubic-bezier(0.4,0,0.2,1)` | 0% | elegant, minimal, luxury |
|
||||
| **Corporate** | 200–400ms | `cubic-bezier(0.2,0,0,1)` | 0–3% | clean, professional, trustworthy |
|
||||
| **Energetic** | 100–250ms | `ease-out-expo` | 15–30% | bold, dynamic, hype |
|
||||
|
||||
Default: **Corporate** for UI/product, **Playful** for illustration/lifestyle. Maps onto the matrix — Premium≈calm-soft, Corporate≈calm-sharp, Playful≈kinetic-soft, Energetic≈kinetic-sharp. The personality fills the "easing family / base timing unit / motion intensity" rows of the spec below.
|
||||
|
||||
## The motion-language spec
|
||||
|
||||
The core artifact. Fill every row with ONE choice and reuse it everywhere.
|
||||
|
||||
| Property | Choose one | Example (calm-sharp / premium tech) |
|
||||
|---|---|---|
|
||||
| Easing family | The signature curve for ~90% of moves | `cubic-bezier(0.22, 1, 0.36, 1)` (decelerate, soft landing) |
|
||||
| Base timing unit | The atomic duration; build others as multiples | 0.4s (enters 0.4s, big moves 0.8s, micro 0.2s) |
|
||||
| Transition family | How shots connect | Match-cut + crossfade only; no wipes/spins |
|
||||
| Stagger rhythm | Delay between grouped elements | 60ms cascade, top-to-bottom |
|
||||
| Motion intensity | Distance/scale/overshoot budget | Travel ≤ 24px, scale 0.96→1.0, overshoot ≤ 2% |
|
||||
| Hold discipline | Minimum rest between moves | ≥ 0.3s of stillness after each beat |
|
||||
|
||||
Two eases maximum: one "in" (exits) and one "out" (entrances/landings). A third curve must justify its existence.
|
||||
|
||||
## Motion hierarchy
|
||||
|
||||
Not everything earns motion. Rank every element, then animate down the list — and stop early. These three tiers are the universal **three motion layers** (see `animation-principles`): Hero = **Primary**, Support = **Secondary**, Texture = **Ambient**.
|
||||
|
||||
| Tier | What it is | How it moves |
|
||||
|---|---|---|
|
||||
| Hero | The one thing the shot is about | Gets the boldest, slowest, most-eased move; lands on the beat |
|
||||
| Support | Context that helps the hero land | Smaller, faster, eased *out* of the way; never competes |
|
||||
| Texture | Background, grain, ambient drift | Subliminal only — slow loops, low contrast, no hard cuts |
|
||||
|
||||
If two elements compete for the eye in one frame, the direction has failed. Demote one to support.
|
||||
|
||||
## Restraint — what NOT to animate
|
||||
|
||||
The senior move. For each element ask: *does the motion carry meaning?* If not, hold it still.
|
||||
|
||||
- Don't animate the entire frame at once — stagger or hold a anchor so the eye has a home.
|
||||
- Don't transition the transition (no animated wipe *and* a spin *and* a fade).
|
||||
- Don't loop-animate text the viewer is still reading.
|
||||
- Don't add overshoot to serious/financial content — it reads as toy-like.
|
||||
- One "wow" moment per piece. A second one cancels the first.
|
||||
|
||||
## Pacing & rhythm
|
||||
|
||||
Pacing is the edit's heartbeat. Vary it on purpose: tension (faster, tighter) → release (a hold). Map energy across the timeline before timing individual moves — most pieces want a low-energy open, a build, a peak, and a settled end card. Cut and land key moments on the audio (see launch-video for beat-syncing); silence and stillness are pacing tools, not gaps to fill.
|
||||
|
||||
## Consistency check
|
||||
|
||||
Before sign-off, audit every shot against the spec table:
|
||||
|
||||
- Same easing family on comparable moves? (No stray `linear` or `ease`.)
|
||||
- All durations are multiples of the base unit?
|
||||
- Only the chosen transition family between shots?
|
||||
- Stagger rhythm identical across grouped reveals?
|
||||
- Hold discipline respected — no two moves stacked without rest?
|
||||
- Exactly one hero per frame?
|
||||
|
||||
Any "no" is a direction defect, not a craft preference — fix it before polishing.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/direction-playbook.md` — worked examples of the motion-language spec for five tone cells, a full brief template, a shot-by-shot direction-notes template with example entries, brand→motion translation tables, an easing-vocabulary cheat sheet, and a complete pre-render direction checklist.
|
||||
@@ -0,0 +1,184 @@
|
||||
# Motion Direction Playbook
|
||||
|
||||
Depth behind the SKILL.md frameworks: a brief template, filled motion-language specs for five tone cells, brand→motion translation, a direction-notes template with worked entries, an easing cheat sheet, and a pre-render checklist.
|
||||
|
||||
## Creative brief template
|
||||
|
||||
Fill before any other decision. One line per row — if a row can't be answered, the piece isn't ready to direct.
|
||||
|
||||
```
|
||||
PIECE: <name / working title>
|
||||
AUDIENCE: <who, and what they already know>
|
||||
ONE MESSAGE: <the single thing they remember after watching>
|
||||
PLATFORM: <YouTube / Reels / in-app / landing hero — sets aspect, sound, duration>
|
||||
DURATION: <seconds; shorter = fewer ideas, not faster ideas>
|
||||
MOOD: <exactly 3 adjectives, e.g. "confident, precise, warm">
|
||||
BRAND CUES: <colors, type, logo motion rules, any existing motion system>
|
||||
CONSTRAINTS: <must-include / must-avoid; legal supers; localization>
|
||||
REFERENCE: <1–3 links + one line on WHAT to borrow from each>
|
||||
```
|
||||
|
||||
The 3 adjectives are load-bearing: they resolve every later judgment call. "Confident, precise, warm" rules out bouncy overshoot (not precise) and cold linear snaps (not warm) on its own.
|
||||
|
||||
## Filled motion-language specs (five tone cells)
|
||||
|
||||
Each is internally consistent — copy the closest one and adjust. Easings are CSS `cubic-bezier`; the same curves map to GSAP eases and Lottie keyframe handles.
|
||||
|
||||
### Calm + Soft — luxury / wellness / editorial
|
||||
| Property | Choice |
|
||||
|---|---|
|
||||
| Easing | `cubic-bezier(0.4, 0, 0.2, 1)` in/out, long tails |
|
||||
| Base unit | 0.6s (micro 0.3s, hero 1.2s) |
|
||||
| Transition | Slow crossfade + gentle scale; no hard cuts |
|
||||
| Stagger | 120ms, generous |
|
||||
| Intensity | Travel ≤ 16px, scale 0.98→1.0, zero overshoot |
|
||||
| Hold | ≥ 0.6s stillness; let frames breathe |
|
||||
|
||||
### Calm + Sharp — premium tech / fintech
|
||||
| Property | Choice |
|
||||
|---|---|
|
||||
| Easing | `cubic-bezier(0.22, 1, 0.36, 1)` decel; exits `cubic-bezier(0.4, 0, 1, 1)` |
|
||||
| Base unit | 0.4s |
|
||||
| Transition | Match-cut + crossfade only |
|
||||
| Stagger | 60ms cascade |
|
||||
| Intensity | Travel ≤ 24px, scale 0.96→1.0, overshoot ≤ 2% |
|
||||
| Hold | ≥ 0.3s |
|
||||
|
||||
### Kinetic + Sharp — sports / hype / gaming / launch
|
||||
| Property | Choice |
|
||||
|---|---|
|
||||
| Easing | `cubic-bezier(0.16, 1, 0.3, 1)` hard-out; snaps near-instant |
|
||||
| Base unit | 0.25s (micro 0.12s) |
|
||||
| Transition | Hard cuts on beats; speed-ramp into drops |
|
||||
| Stagger | 40ms, tight machine-gun |
|
||||
| Intensity | Big travel, scale 1.18→1.0 overshoot on impact |
|
||||
| Hold | ≥ 0.2s, brief — energy over rest |
|
||||
|
||||
### Kinetic + Soft — playful / lifestyle / kids
|
||||
| Property | Choice |
|
||||
|---|---|
|
||||
| Easing | Spring / `cubic-bezier(0.34, 1.56, 0.64, 1)` overshoot |
|
||||
| Base unit | 0.4s, loose |
|
||||
| Transition | Bouncy pushes, squash-and-stretch wipes |
|
||||
| Stagger | 80ms, irregular feels human |
|
||||
| Intensity | Generous bounce, secondary follow-through |
|
||||
| Hold | ≥ 0.3s |
|
||||
|
||||
### Neutral data-story — explainer / dashboard / report
|
||||
| Property | Choice |
|
||||
|---|---|
|
||||
| Easing | `cubic-bezier(0.25, 0.1, 0.25, 1)` calm in/out |
|
||||
| Base unit | 0.5s; count-ups 0.8–1.2s |
|
||||
| Transition | Cross-dissolve between data states; never wipe a chart |
|
||||
| Stagger | 80ms per series/bar |
|
||||
| Intensity | Number eases to value; axes static; no overshoot on quantities |
|
||||
| Hold | ≥ 0.5s on each datapoint so it can be read |
|
||||
|
||||
## Brand → motion translation
|
||||
|
||||
Brands describe themselves in adjectives; direction converts those to motion parameters.
|
||||
|
||||
| Brand word | Easing | Timing | Intensity | What to avoid |
|
||||
|---|---|---|---|---|
|
||||
| Premium / luxury | Long, soft decel | Slow, generous holds | Minimal travel, no overshoot | Bounce, fast cuts, spins |
|
||||
| Trustworthy / precise | Clean decel, symmetric | On-grid, multiples of base | Small, controlled | Wobble, irregular stagger |
|
||||
| Energetic / bold | Hard-out, snappy | Fast, syncopated | Big scale, overshoot on beats | Slow fades, dead air |
|
||||
| Friendly / human | Slight overshoot/spring | Loose, slightly irregular | Soft bounce, follow-through | Robotic linear, perfect symmetry |
|
||||
| Innovative / techy | Decel with subtle anticipation | Tight, layered staggers | Crisp, geometric | Skeuomorphic squash, grunge |
|
||||
|
||||
Map only the 2–3 strongest brand words. Trying to honor every adjective produces mush.
|
||||
|
||||
## Direction-notes template (per shot)
|
||||
|
||||
Direct *intent*, not keyframes — let the technique skill choose exact values within the spec.
|
||||
|
||||
```
|
||||
SHOT <n> [t = 0.0–0.0s]
|
||||
PURPOSE: <why this shot exists — what the viewer should feel/learn>
|
||||
HERO: <the one element; how it enters and lands>
|
||||
SUPPORT: <secondary elements; how they yield to the hero>
|
||||
TEXTURE: <ambient/background behaviour, if any>
|
||||
TRANSITION IN/OUT: <from the locked transition family only>
|
||||
BEAT: <audio mark the key move lands on, if scored>
|
||||
NOTE: <any deliberate break from the spec + the reason>
|
||||
```
|
||||
|
||||
### Worked example (Calm + Sharp, premium tech)
|
||||
|
||||
```
|
||||
SHOT 3 [t = 6.0–9.5s]
|
||||
PURPOSE: Reveal the dashboard as effortless and accurate.
|
||||
HERO: Dashboard card scales 0.96→1.0 over 0.8s on the decel ease, soft landing.
|
||||
SUPPORT: Three metric chips cascade in, 60ms stagger, 0.4s each, eased out so the
|
||||
eye finishes on the card, not the chips.
|
||||
TEXTURE: Grid lines drift 2px over 8s, 8% opacity — subliminal only.
|
||||
TRANSITION IN/OUT: Match-cut from the prior logo mark / crossfade to shot 4.
|
||||
BEAT: Card lands on the downbeat at 6.4s.
|
||||
NOTE: No overshoot anywhere — this is financial data; overshoot reads as imprecise.
|
||||
```
|
||||
|
||||
### Worked example (Neutral data-story, count-up)
|
||||
|
||||
```
|
||||
SHOT 5 [t = 12.0–15.0s]
|
||||
PURPOSE: Land the headline number ($1.4M) so it's read and believed.
|
||||
HERO: Number eases from 0 to 1,400,000 over 1.1s; decelerates so the final
|
||||
digits settle slowly and feel exact.
|
||||
SUPPORT: Caption fades in after the count completes (not during) — don't make
|
||||
the viewer read while a number is still moving.
|
||||
TEXTURE: None. The frame holds still for 0.7s after — stillness sells accuracy.
|
||||
TRANSITION IN/OUT: Cross-dissolve from the bar chart; cross-dissolve out.
|
||||
NOTE: Hold the final value on screen ≥ 0.5s before any exit.
|
||||
```
|
||||
|
||||
## Easing vocabulary cheat sheet
|
||||
|
||||
| Curve | `cubic-bezier` | Feel | Use for |
|
||||
|---|---|---|---|
|
||||
| Decelerate (ease-out) | `0, 0, 0.2, 1` | Arrives, settles softly | Entrances, landings — the workhorse |
|
||||
| Accelerate (ease-in) | `0.4, 0, 1, 1` | Leaves with intent | Exits, things flying off |
|
||||
| Standard (in-out) | `0.4, 0, 0.2, 1` | Smooth both ends | On-screen repositioning |
|
||||
| Emphasized decel | `0.22, 1, 0.36, 1` | Confident, premium landing | Hero reveals (calm-sharp) |
|
||||
| Hard out | `0.16, 1, 0.3, 1` | Snappy, energetic | Kinetic pieces, hype |
|
||||
| Overshoot | `0.34, 1.56, 0.64, 1` | Springy, playful, alive | Friendly/kinetic-soft only |
|
||||
|
||||
Rules of thumb: entrances ease *out* (decelerate in), exits ease *in* (accelerate away). Avoid `linear` except for continuous loops (rotation, marquees, ambient drift). Never ship default `ease` — it's the un-decided curve and reads as unconsidered.
|
||||
|
||||
## Pacing map
|
||||
|
||||
Sketch energy across the timeline before timing any single move. A typical 30s arc:
|
||||
|
||||
```
|
||||
energy
|
||||
^ ____ peak (the one "wow")
|
||||
| / \
|
||||
| build/ \ settle
|
||||
| / \____ end card hold
|
||||
|__/
|
||||
+----------------------------------> time
|
||||
open build peak resolve
|
||||
```
|
||||
|
||||
Vary deliberately: a hold after a fast sequence makes the next move hit harder. Density of cuts is a dial — turn it up for tension, down for gravity. Match key landings to audio beats where scored.
|
||||
|
||||
## Pre-render direction checklist
|
||||
|
||||
- [ ] Brief complete; one message and exactly 3 mood adjectives written down.
|
||||
- [ ] Tone placed in a single cell of the energy matrix — not straddling two.
|
||||
- [ ] Motion-language spec filled, every row a single choice.
|
||||
- [ ] ≤ 2 primary easing curves used across the whole piece.
|
||||
- [ ] All durations are multiples of the base unit.
|
||||
- [ ] Only the chosen transition family appears between shots.
|
||||
- [ ] Stagger rhythm identical across all grouped reveals.
|
||||
- [ ] Every frame has exactly one hero; support never competes.
|
||||
- [ ] Restraint pass done — listed what stays static and why.
|
||||
- [ ] Hold discipline respected; no two moves stacked without rest.
|
||||
- [ ] Exactly one "wow" moment.
|
||||
- [ ] Numbers/quantities ease to value, hold ≥ 0.5s, and don't overshoot.
|
||||
- [ ] Pacing map drawn; key moments land on beats (if scored).
|
||||
- [ ] Brand cues honored; nothing animated that the brand wouldn't.
|
||||
|
||||
---
|
||||
## Built by the team behind iart.ai
|
||||
|
||||
This skill is part of an open motion-graphics collection from iart.ai — the AI motion agent that turns data, scripts, and designs into editable motion graphics (Remotion → MP4). If you'd rather not hand-build this, iart.ai can act as the motion director — apply one consistent motion language and your brand across every shot, then render the piece. → [iart.ai](https://iart.ai/?utm_source=github&utm_medium=reference&utm_campaign=motion-design-skills&utm_content=ref_footer&utm_term=motion-art-direction)
|
||||
Whitespace-only changes.
@@ -0,0 +1,279 @@
|
||||
---
|
||||
name: motion-background
|
||||
description: This skill should be used when the user asks to "add an animated background", "build a mesh/gradient background", "make an aurora/shader background", "add constellation/particle background", "animated hero background", or "a subtle looping background behind content". Covers CSS mesh gradients, GLSL shader gradients (Three.js), canvas particle constellations, seamless loops, and reduced-motion/performance handling — fully self-contained.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Motion Background
|
||||
|
||||
Ambient, living backgrounds that sit behind content without stealing focus. The goal is subtle, seamlessly looping motion that stays readable and performant. This skill is self-contained: every technique below ships its own runnable code — CSS, GLSL, and canvas — with no external dependencies beyond Three.js where a shader is involved.
|
||||
|
||||
## When to use
|
||||
|
||||
- Hero/landing background that subtly moves; section ambience.
|
||||
- Login/splash/empty-state living backdrops.
|
||||
- A gradient, mesh, aurora, shader, or particle/constellation field behind text or UI.
|
||||
|
||||
## Pick an approach
|
||||
|
||||
| Look | Technique | Cost |
|
||||
|---|---|---|
|
||||
| Soft animated gradient/mesh | CSS gradients + keyframes | Cheapest, no JS |
|
||||
| Flowing aurora / organic noise | GLSL fragment shader (Three.js full-screen quad) | GPU, moderate |
|
||||
| Constellation / drifting dots | Canvas 2D particles | CPU, scales with count |
|
||||
| Depth / parallax | Same shader/particles with mouse-driven offset | Moderate |
|
||||
|
||||
Default to CSS if a gradient suffices — it is the cheapest and most reliable. Use a shader for organic flowing color; use canvas particles for a constellation/network look.
|
||||
|
||||
## Design rules
|
||||
|
||||
- **Subtle**: low contrast versus content, slow motion (long periods, 8–30s loops), nothing that competes with text.
|
||||
- **Readable**: keep contrast for foreground text; add a scrim (`linear-gradient` overlay or `backdrop`) if needed.
|
||||
- **Seamless loop**: drive motion with periodic functions so time wraps with no visible jump (see below).
|
||||
- **Respect motion preferences and battery**: honor `prefers-reduced-motion`, and pause when offscreen or the tab is hidden.
|
||||
|
||||
## Core techniques (inlined, runnable)
|
||||
|
||||
### 1. Animated CSS mesh gradient (no JS)
|
||||
|
||||
Layered radial gradients whose positions drift. Animating `background-position` on a larger-than-viewport gradient loops seamlessly.
|
||||
|
||||
```css
|
||||
.bg {
|
||||
position: fixed; inset: 0; z-index: -1;
|
||||
background:
|
||||
radial-gradient(40% 50% at 20% 30%, #5b8cff55, transparent 60%),
|
||||
radial-gradient(45% 55% at 80% 20%, #b05bff55, transparent 60%),
|
||||
radial-gradient(50% 60% at 50% 80%, #2de1c255, transparent 60%),
|
||||
#0b0e1a;
|
||||
background-size: 200% 200%;
|
||||
animation: meshmove 24s ease-in-out infinite;
|
||||
}
|
||||
@keyframes meshmove {
|
||||
0%, 100% { background-position: 0% 0%, 100% 0%, 50% 100%; }
|
||||
50% { background-position: 30% 20%, 70% 30%, 60% 70%; }
|
||||
}
|
||||
@media (prefers-reduced-motion: reduce) {
|
||||
.bg { animation: none; }
|
||||
}
|
||||
```
|
||||
|
||||
The `0%` and `100%` keyframes are identical, so the loop has no seam.
|
||||
|
||||
### 2. Full-screen GLSL gradient + noise shader (Three.js)
|
||||
|
||||
A flowing aurora/gradient using value noise in a fragment shader on a full-screen plane. `uTime` advances each frame; to loop, feed it a wrapped time (section 4).
|
||||
|
||||
```js
|
||||
import * as THREE from 'three';
|
||||
|
||||
const canvas = document.querySelector('#bg');
|
||||
const renderer = new THREE.WebGLRenderer({canvas, antialias: true});
|
||||
renderer.setPixelRatio(Math.min(window.devicePixelRatio, 2)); // cap for perf
|
||||
const scene = new THREE.Scene();
|
||||
const camera = new THREE.OrthographicCamera(-1, 1, 1, -1, 0, 1);
|
||||
|
||||
const uniforms = {
|
||||
uTime: {value: 0},
|
||||
uRes: {value: new THREE.Vector2()},
|
||||
uColorA: {value: new THREE.Color('#5b8cff')},
|
||||
uColorB: {value: new THREE.Color('#b05bff')},
|
||||
};
|
||||
|
||||
const material = new THREE.ShaderMaterial({
|
||||
uniforms,
|
||||
vertexShader: `void main(){ gl_Position = vec4(position, 1.0); }`,
|
||||
fragmentShader: `
|
||||
precision highp float;
|
||||
uniform float uTime; uniform vec2 uRes;
|
||||
uniform vec3 uColorA, uColorB;
|
||||
|
||||
// hash + value noise
|
||||
float hash(vec2 p){ return fract(sin(dot(p, vec2(127.1,311.7)))*43758.5453); }
|
||||
float noise(vec2 p){
|
||||
vec2 i = floor(p), f = fract(p);
|
||||
float a = hash(i), b = hash(i+vec2(1,0));
|
||||
float c = hash(i+vec2(0,1)), d = hash(i+vec2(1,1));
|
||||
vec2 u = f*f*(3.0-2.0*f);
|
||||
return mix(mix(a,b,u.x), mix(c,d,u.x), u.y);
|
||||
}
|
||||
float fbm(vec2 p){
|
||||
float v=0.0, a=0.5;
|
||||
for(int i=0;i<5;i++){ v += a*noise(p); p*=2.0; a*=0.5; }
|
||||
return v;
|
||||
}
|
||||
void main(){
|
||||
vec2 uv = gl_FragCoord.xy / uRes.xy;
|
||||
float n = fbm(uv*3.0 + vec2(uTime*0.05, uTime*0.03));
|
||||
vec3 col = mix(uColorA, uColorB, smoothstep(0.2, 0.8, n + uv.y*0.3));
|
||||
gl_FragColor = vec4(col, 1.0);
|
||||
}`,
|
||||
});
|
||||
|
||||
const quad = new THREE.Mesh(new THREE.PlaneGeometry(2, 2), material);
|
||||
scene.add(quad);
|
||||
|
||||
function resize(){
|
||||
renderer.setSize(window.innerWidth, window.innerHeight);
|
||||
uniforms.uRes.value.set(window.innerWidth, window.innerHeight);
|
||||
}
|
||||
window.addEventListener('resize', resize); resize();
|
||||
|
||||
const clock = new THREE.Clock();
|
||||
let running = true;
|
||||
function loop(){
|
||||
if (running){
|
||||
uniforms.uTime.value = clock.getElapsedTime();
|
||||
renderer.render(scene, camera);
|
||||
}
|
||||
requestAnimationFrame(loop);
|
||||
}
|
||||
loop();
|
||||
```
|
||||
|
||||
### 3. Canvas constellation particles
|
||||
|
||||
Drifting points connected by lines when near — the classic "network" background. Pure canvas 2D, no dependencies.
|
||||
|
||||
```js
|
||||
const canvas = document.querySelector('#stars');
|
||||
const ctx = canvas.getContext('2d');
|
||||
let W, H, pts;
|
||||
const COUNT = 80, LINK = 120;
|
||||
|
||||
function init(){
|
||||
W = canvas.width = innerWidth; H = canvas.height = innerHeight;
|
||||
pts = Array.from({length: COUNT}, () => ({
|
||||
x: Math.random()*W, y: Math.random()*H,
|
||||
vx: (Math.random()-0.5)*0.3, vy: (Math.random()-0.5)*0.3,
|
||||
}));
|
||||
}
|
||||
addEventListener('resize', init); init();
|
||||
|
||||
function frame(){
|
||||
ctx.clearRect(0, 0, W, H);
|
||||
for (const p of pts){
|
||||
p.x += p.vx; p.y += p.vy;
|
||||
if (p.x<0||p.x>W) p.vx*=-1;
|
||||
if (p.y<0||p.y>H) p.vy*=-1;
|
||||
ctx.fillStyle = '#9db4ff'; ctx.fillRect(p.x, p.y, 2, 2);
|
||||
}
|
||||
for (let i=0;i<COUNT;i++) for (let j=i+1;j<COUNT;j++){
|
||||
const dx=pts[i].x-pts[j].x, dy=pts[i].y-pts[j].y;
|
||||
const d=Math.hypot(dx, dy);
|
||||
if (d<LINK){
|
||||
ctx.strokeStyle = `rgba(157,180,255,${1-d/LINK})`;
|
||||
ctx.beginPath(); ctx.moveTo(pts[i].x,pts[i].y); ctx.lineTo(pts[j].x,pts[j].y); ctx.stroke();
|
||||
}
|
||||
}
|
||||
requestAnimationFrame(frame);
|
||||
}
|
||||
frame();
|
||||
```
|
||||
|
||||
The O(n²) link loop is fine to ~120 points; above that, spatial-hash into a grid and only test neighboring cells.
|
||||
|
||||
### 4. Seamless loop technique
|
||||
|
||||
For canvas/JS, wrap time into a fixed period so motion repeats exactly. Use the loop *phase* (0→2π) as the argument to periodic functions:
|
||||
|
||||
```js
|
||||
const PERIOD = 12; // seconds
|
||||
const phase = (t % PERIOD) / PERIOD * Math.PI * 2;
|
||||
const offset = Math.sin(phase) * amp; // returns to start at t = PERIOD
|
||||
```
|
||||
|
||||
Any motion built only from `sin`/`cos` of `phase` (or integer multiples) loops seamlessly. In the shader, feed `uTime = phase` and use only `sin`/`cos` of it; for noise-scrolled gradients, scroll by an integer number of noise cells per period so the field tiles.
|
||||
|
||||
### 5. Reduced motion + offscreen/hidden pause
|
||||
|
||||
```js
|
||||
const reduce = matchMedia('(prefers-reduced-motion: reduce)').matches;
|
||||
if (reduce) {
|
||||
running = false; // render one static frame, then stop
|
||||
renderer.render(scene, camera);
|
||||
}
|
||||
|
||||
// Pause when the tab is hidden
|
||||
document.addEventListener('visibilitychange', () => {
|
||||
running = !document.hidden && !reduce;
|
||||
});
|
||||
|
||||
// Pause when the canvas scrolls offscreen
|
||||
new IntersectionObserver(([e]) => {
|
||||
running = e.isIntersecting && !document.hidden && !reduce;
|
||||
}).observe(canvas);
|
||||
```
|
||||
|
||||
When `running` is false, skip the render inside the rAF loop (as in section 2) — the loop stays alive to resume cheaply, but does no GPU/CPU work.
|
||||
|
||||
## Deliver & verify (standalone HTML)
|
||||
|
||||
> **Packaged helper** (`scripts/`): `scripts/seek-shot.sh anim.html 0 1.5 3` freezes the `?t=N` harness and screenshots each moment; `scripts/contact-sheet.sh sheet.png frame-*.png` tiles them for one-glance review. See `scripts/README.md`.
|
||||
|
||||
A motion background ships as **one `.html` file that opens directly in a browser** — markup, the background, and (for shader/canvas) Three.js or canvas JS from CDN, all inline. One file is the right tier.
|
||||
|
||||
**Output contract:**
|
||||
- One `.html` file: the background layer + a sample of foreground text on top to check readability/contrast.
|
||||
- One animation driver — the CSS `@keyframes`, the shader rAF loop, or the canvas rAF loop; just one.
|
||||
- Include the freeze harness below, matched to the technique, so any moment can be screenshotted deterministically.
|
||||
|
||||
**Freeze harness — pin a frame for screenshots.** Match the mechanism to the technique:
|
||||
|
||||
```html
|
||||
<script>
|
||||
const t = new URLSearchParams(location.search).get("t");
|
||||
if (t !== null) {
|
||||
const T = parseFloat(t);
|
||||
// CSS mesh gradient:
|
||||
document.querySelectorAll(".bg").forEach(el => {
|
||||
el.style.animationDelay = (-T) + "s";
|
||||
el.style.animationPlayState = "paused";
|
||||
});
|
||||
// GLSL shader instead? → render exactly one frame at fixed time:
|
||||
// uniforms.uTime.value = T; renderer.render(scene, camera); running = false;
|
||||
// Canvas particles? → seed deterministically, step the sim to T, draw once, stop the loop.
|
||||
}
|
||||
window.__ready = true; // ready signal for headless wait
|
||||
</script>
|
||||
```
|
||||
|
||||
**Verify loop — render → freeze → screenshot → check:**
|
||||
1. Open the file frozen at start / mid / end across one loop period: `…/bg.html?t=0`, `?t=<period/2>`, `?t=<period>`.
|
||||
2. Screenshot each frozen frame.
|
||||
3. Check **fidelity** (subtle, on-brand, seamless) and **artifacts** — text contrast holds at every frame, no banding in gradients/shader, the loop seam (`t=0` vs `t=period`) matches, no GPU/console errors.
|
||||
|
||||
```bash
|
||||
npx playwright screenshot --wait-for-timeout=600 "file://$PWD/bg.html?t=12" frame-mid.png
|
||||
```
|
||||
|
||||
**Before you finish:**
|
||||
1. Opens standalone in a browser — no console/WebGL errors, no missing CDN.
|
||||
2. One driver; `?t=N` freezes the exact frame (shader renders one frame, canvas sim stepped deterministically).
|
||||
3. Screenshotted at start / mid / end — foreground text stays readable, no banding, loop seam invisible.
|
||||
4. `prefers-reduced-motion` honored (one static frame rendered, loop stopped) + pauses offscreen/hidden.
|
||||
5. `devicePixelRatio` capped at 2; motion is slow and subtle, never out-contrasting content.
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Need | Do |
|
||||
|---|---|
|
||||
| Cheapest gradient | CSS layered `radial-gradient` + `background-position` keyframes |
|
||||
| Organic flow | GLSL `fbm` noise, scroll `uTime` |
|
||||
| Network/dots | canvas particles + distance-linked lines |
|
||||
| No seam | drive motion by `sin/cos(phase)`, identical first/last keyframe |
|
||||
| Perf cap | `setPixelRatio(min(dpr, 2))`, throttle, reduce particle count on mobile |
|
||||
| Accessibility | `prefers-reduced-motion` static fallback |
|
||||
| Save battery | pause on `visibilitychange` + `IntersectionObserver` |
|
||||
|
||||
## Gotchas
|
||||
|
||||
- Never let the background out-contrast the foreground text — add a scrim if it does.
|
||||
- Uncapped `devicePixelRatio` on retina/4K murders the GPU; cap at 2.
|
||||
- `background-position` loops only if first and last keyframes match exactly.
|
||||
- Noise-scrolled shaders loop only if the scroll is an integer number of cells per period; otherwise the loop visibly jumps.
|
||||
- Forgetting the offscreen/hidden pause drains battery on mobile even when the user can't see it.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/background-recipes.md` — fuller drop-in implementations: a richer multi-blob CSS mesh with blur, the complete Three.js aurora shader with mouse parallax and a true seamless-loop time wrap, a grid-optimized constellation field with mouse repulsion, a perfectly-looping noise-flow technique, and a complete reduced-motion + offscreen-pause manager wrapping all of them.
|
||||
@@ -0,0 +1,233 @@
|
||||
# Background Recipes — Drop-in Implementations
|
||||
|
||||
Fuller, production-ready versions of each technique. All are self-contained (Three.js is the only external dependency, and only for the shader recipe).
|
||||
|
||||
## 1. Rich CSS mesh gradient (blurred blobs)
|
||||
|
||||
Animated blobs behind a blur layer give a soft "mesh gradient" look without WebGL. Uses absolutely-positioned radial blobs and `filter: blur`.
|
||||
|
||||
```html
|
||||
<div class="mesh">
|
||||
<span class="blob b1"></span>
|
||||
<span class="blob b2"></span>
|
||||
<span class="blob b3"></span>
|
||||
</div>
|
||||
```
|
||||
|
||||
```css
|
||||
.mesh{ position:fixed; inset:0; z-index:-1; overflow:hidden; background:#0b0e1a; }
|
||||
.mesh .blob{
|
||||
position:absolute; width:50vmax; height:50vmax; border-radius:50%;
|
||||
filter:blur(80px); opacity:.55; mix-blend-mode:screen;
|
||||
}
|
||||
.b1{ background:#5b8cff; top:-10%; left:-5%; animation:drift1 26s ease-in-out infinite; }
|
||||
.b2{ background:#b05bff; top:30%; left:55%; animation:drift2 30s ease-in-out infinite; }
|
||||
.b3{ background:#2de1c2; top:60%; left:10%; animation:drift3 22s ease-in-out infinite; }
|
||||
|
||||
@keyframes drift1{ 0%,100%{transform:translate(0,0)} 50%{transform:translate(20vmax,10vmax)} }
|
||||
@keyframes drift2{ 0%,100%{transform:translate(0,0)} 50%{transform:translate(-18vmax,12vmax)} }
|
||||
@keyframes drift3{ 0%,100%{transform:translate(0,0)} 50%{transform:translate(12vmax,-14vmax)} }
|
||||
|
||||
@media (prefers-reduced-motion: reduce){ .mesh .blob{ animation:none } }
|
||||
```
|
||||
|
||||
`mix-blend-mode: screen` blends overlapping blobs into bright color fields; `blur` smooths them into a continuous mesh. Identical 0%/100% keyframes guarantee a seamless loop.
|
||||
|
||||
## 2. Three.js aurora shader — full, with parallax and seamless loop
|
||||
|
||||
```js
|
||||
import * as THREE from 'three';
|
||||
|
||||
export function createAurora(canvas){
|
||||
const renderer = new THREE.WebGLRenderer({canvas, antialias:true});
|
||||
renderer.setPixelRatio(Math.min(window.devicePixelRatio, 2));
|
||||
const scene = new THREE.Scene();
|
||||
const camera = new THREE.OrthographicCamera(-1,1,1,-1,0,1);
|
||||
|
||||
const PERIOD = 18.0; // seconds for a full seamless loop
|
||||
const u = {
|
||||
uPhase: {value:0}, // 0..2π, wraps each PERIOD
|
||||
uRes: {value:new THREE.Vector2()},
|
||||
uMouse: {value:new THREE.Vector2(0.5,0.5)},
|
||||
uColorA: {value:new THREE.Color('#0b1d4d')},
|
||||
uColorB: {value:new THREE.Color('#5b8cff')},
|
||||
uColorC: {value:new THREE.Color('#b05bff')},
|
||||
};
|
||||
|
||||
const material = new THREE.ShaderMaterial({
|
||||
uniforms:u,
|
||||
vertexShader:`void main(){ gl_Position=vec4(position,1.0); }`,
|
||||
fragmentShader:`
|
||||
precision highp float;
|
||||
uniform float uPhase; uniform vec2 uRes, uMouse;
|
||||
uniform vec3 uColorA, uColorB, uColorC;
|
||||
|
||||
float hash(vec2 p){ return fract(sin(dot(p,vec2(127.1,311.7)))*43758.5453); }
|
||||
float noise(vec2 p){
|
||||
vec2 i=floor(p), f=fract(p);
|
||||
float a=hash(i), b=hash(i+vec2(1,0)), c=hash(i+vec2(0,1)), d=hash(i+vec2(1,1));
|
||||
vec2 w=f*f*(3.0-2.0*f);
|
||||
return mix(mix(a,b,w.x), mix(c,d,w.x), w.y);
|
||||
}
|
||||
float fbm(vec2 p){ float v=0.,a=.5; for(int i=0;i<6;i++){v+=a*noise(p);p*=2.;a*=.5;} return v; }
|
||||
|
||||
void main(){
|
||||
vec2 uv = gl_FragCoord.xy/uRes.xy;
|
||||
uv += (uMouse-0.5)*0.06; // subtle parallax
|
||||
// loop the flow by moving along a circle in noise space:
|
||||
vec2 flow = vec2(cos(uPhase), sin(uPhase))*0.6;
|
||||
float n = fbm(uv*3.0 + flow);
|
||||
float bands = sin((uv.y*4.0 + n*2.0)) * 0.5 + 0.5;
|
||||
vec3 col = mix(uColorA, uColorB, smoothstep(0.2,0.7,n));
|
||||
col = mix(col, uColorC, smoothstep(0.5,1.0,bands)*0.6);
|
||||
gl_FragColor = vec4(col, 1.0);
|
||||
}`,
|
||||
});
|
||||
|
||||
const quad = new THREE.Mesh(new THREE.PlaneGeometry(2,2), material);
|
||||
scene.add(quad);
|
||||
|
||||
function resize(){
|
||||
renderer.setSize(innerWidth, innerHeight);
|
||||
u.uRes.value.set(innerWidth, innerHeight);
|
||||
}
|
||||
addEventListener('resize', resize); resize();
|
||||
addEventListener('pointermove', e => {
|
||||
u.uMouse.value.set(e.clientX/innerWidth, 1 - e.clientY/innerHeight);
|
||||
});
|
||||
|
||||
const start = performance.now();
|
||||
let running = true;
|
||||
function loop(){
|
||||
if (running){
|
||||
const t = (performance.now() - start)/1000;
|
||||
u.uPhase.value = (t % PERIOD)/PERIOD * Math.PI * 2; // wraps → seamless
|
||||
renderer.render(scene, camera);
|
||||
}
|
||||
requestAnimationFrame(loop);
|
||||
}
|
||||
loop();
|
||||
return {
|
||||
set running(v){ running = v; if (v) loop(); },
|
||||
renderOnce(){ renderer.render(scene, camera); },
|
||||
};
|
||||
}
|
||||
```
|
||||
|
||||
The seamless loop trick: drive the noise offset with `vec2(cos(uPhase), sin(uPhase))`, a closed circle in noise space — at `uPhase = 2π` the field is identical to `uPhase = 0`.
|
||||
|
||||
## 3. Constellation field — grid-optimized + mouse repulsion
|
||||
|
||||
```js
|
||||
export function createConstellation(canvas, {count=120, link=130, repel=140} = {}){
|
||||
const ctx = canvas.getContext('2d');
|
||||
let W,H,pts; const mouse = {x:-9999, y:-9999};
|
||||
|
||||
function init(){
|
||||
W = canvas.width = innerWidth; H = canvas.height = innerHeight;
|
||||
pts = Array.from({length:count}, () => ({
|
||||
x:Math.random()*W, y:Math.random()*H,
|
||||
vx:(Math.random()-0.5)*0.25, vy:(Math.random()-0.5)*0.25,
|
||||
}));
|
||||
}
|
||||
addEventListener('resize', init); init();
|
||||
addEventListener('pointermove', e => { mouse.x=e.clientX; mouse.y=e.clientY; });
|
||||
addEventListener('pointerleave', () => { mouse.x=mouse.y=-9999; });
|
||||
|
||||
// spatial hash to avoid O(n^2)
|
||||
const cell = link;
|
||||
function key(x,y){ return ((x/cell)|0)+','+((y/cell)|0); }
|
||||
|
||||
let running = true;
|
||||
function frame(){
|
||||
if (!running){ requestAnimationFrame(frame); return; }
|
||||
ctx.clearRect(0,0,W,H);
|
||||
const grid = new Map();
|
||||
for (const p of pts){
|
||||
p.x+=p.vx; p.y+=p.vy;
|
||||
if (p.x<0||p.x>W) p.vx*=-1;
|
||||
if (p.y<0||p.y>H) p.vy*=-1;
|
||||
// mouse repulsion
|
||||
const dx=p.x-mouse.x, dy=p.y-mouse.y, d=Math.hypot(dx,dy);
|
||||
if (d<repel){ p.x+=dx/d*1.5; p.y+=dy/d*1.5; }
|
||||
const k=key(p.x,p.y); (grid.get(k)||grid.set(k,[]).get(k)).push(p);
|
||||
ctx.fillStyle='#9db4ff'; ctx.fillRect(p.x,p.y,2,2);
|
||||
}
|
||||
// link only within neighboring cells
|
||||
for (const p of pts){
|
||||
const cx=(p.x/cell)|0, cy=(p.y/cell)|0;
|
||||
for (let gx=cx-1;gx<=cx+1;gx++) for (let gy=cy-1;gy<=cy+1;gy++){
|
||||
const bucket = grid.get(gx+','+gy); if (!bucket) continue;
|
||||
for (const q of bucket){
|
||||
if (q===p) continue;
|
||||
const dx=p.x-q.x, dy=p.y-q.y, d=Math.hypot(dx,dy);
|
||||
if (d<link){
|
||||
ctx.strokeStyle=`rgba(157,180,255,${(1-d/link)*0.6})`;
|
||||
ctx.beginPath(); ctx.moveTo(p.x,p.y); ctx.lineTo(q.x,q.y); ctx.stroke();
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
requestAnimationFrame(frame);
|
||||
}
|
||||
frame();
|
||||
return { set running(v){ running=v; } };
|
||||
}
|
||||
```
|
||||
|
||||
Spatial hashing keeps the link test near O(n) and scales to several hundred points.
|
||||
|
||||
## 4. Perfectly looping noise flow (canvas)
|
||||
|
||||
To loop drifting noise, sample it at a position that travels a closed path:
|
||||
|
||||
```js
|
||||
const PERIOD = 10; // seconds
|
||||
const phase = (t % PERIOD)/PERIOD * Math.PI*2;
|
||||
const sx = Math.cos(phase)*R; // closed circle → returns to start
|
||||
const sy = Math.sin(phase)*R;
|
||||
// feed (baseX + sx, baseY + sy) into your noise function
|
||||
```
|
||||
|
||||
Because `(sx,sy)` returns exactly to its start at `t = PERIOD`, the whole field repeats with no jump. For 1-D scroll loops instead, advance by an *integer* number of noise cells per period (`scroll = floor(cells) * (t % PERIOD)/PERIOD`) so the tile aligns.
|
||||
|
||||
## 5. Reduced-motion + lifecycle manager
|
||||
|
||||
A single controller that wraps any of the above (anything exposing a `running` setter and a one-frame render).
|
||||
|
||||
```js
|
||||
export function attachLifecycle(canvas, controller){
|
||||
const mq = matchMedia('(prefers-reduced-motion: reduce)');
|
||||
|
||||
function apply(){
|
||||
const reduced = mq.matches;
|
||||
if (reduced){
|
||||
controller.running = false;
|
||||
controller.renderOnce?.(); // one static frame
|
||||
return;
|
||||
}
|
||||
controller.running = !document.hidden && visible;
|
||||
}
|
||||
|
||||
let visible = true;
|
||||
const io = new IntersectionObserver(([e]) => { visible = e.isIntersecting; apply(); });
|
||||
io.observe(canvas);
|
||||
document.addEventListener('visibilitychange', apply);
|
||||
mq.addEventListener?.('change', apply);
|
||||
apply();
|
||||
|
||||
return () => { io.disconnect(); document.removeEventListener('visibilitychange', apply); };
|
||||
}
|
||||
```
|
||||
|
||||
Usage:
|
||||
|
||||
```js
|
||||
const aurora = createAurora(document.querySelector('#bg'));
|
||||
attachLifecycle(document.querySelector('#bg'), aurora);
|
||||
```
|
||||
|
||||
This guarantees: a static frame for reduced-motion users, no CPU/GPU work when the tab is hidden or the canvas is scrolled out of view, and automatic resume when conditions change.
|
||||
|
||||
---
|
||||
Keep the loop seamless and the cost low and a background lives behind content. Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=motion-design-skills&utm_content=skill_footer&utm_term=motion-background)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
Whitespace-only changes.
@@ -0,0 +1,194 @@
|
||||
---
|
||||
name: remotion-video
|
||||
description: This skill should be used when the user asks to "make a video with Remotion", "create a programmatic/data-driven video in React", "render an MP4/GIF from code", "animate with useCurrentFrame/interpolate/spring", "sync video to audio/beats", or "render a templated video per record headlessly". Covers building React compositions and rendering them via CLI or @remotion/renderer.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Remotion (Programmatic Video)
|
||||
|
||||
Build real MP4/GIF/WebM videos in React. Every frame is a pure function of `useCurrentFrame()`, so output is deterministic, scrubbable, diffable, and renderable in CI. The code-first alternative to After Effects for templated and data-driven motion graphics.
|
||||
|
||||
## When to use
|
||||
|
||||
- Render MP4/GIF/WebM from code (social clips, title cards, explainers).
|
||||
- Templated/data-driven videos: one composition, many outputs from props (per-user, per-record, per-row of a CSV/DB).
|
||||
- Motion graphics that must be versioned, code-reviewed, and rendered in CI without a GUI.
|
||||
- Programmatic audio sync, charts that animate from data, or embedding shaders/Three.js into video.
|
||||
|
||||
## Core techniques
|
||||
|
||||
### Frame-driven animation
|
||||
|
||||
Animation is derived from the current frame, never from `setState` or `requestAnimationFrame`. `interpolate` maps an input range to an output range; `spring` produces physically natural motion.
|
||||
|
||||
```jsx
|
||||
import {useCurrentFrame, useVideoConfig, interpolate, spring} from 'remotion';
|
||||
|
||||
export const Title = () => {
|
||||
const frame = useCurrentFrame();
|
||||
const {fps} = useVideoConfig();
|
||||
|
||||
// Fade in over frames 0-30, then stay (clamp prevents over/undershoot).
|
||||
const opacity = interpolate(frame, [0, 30], [0, 1], {
|
||||
extrapolateLeft: 'clamp',
|
||||
extrapolateRight: 'clamp',
|
||||
});
|
||||
|
||||
// Spring entrance; divide by spring duration is unnecessary — drive transforms directly.
|
||||
const scale = spring({frame, fps, config: {damping: 200, mass: 1, stiffness: 100}});
|
||||
|
||||
return <h1 style={{opacity, transform: `scale(${scale})`}}>Hello</h1>;
|
||||
};
|
||||
```
|
||||
|
||||
Always clamp `interpolate` unless an intentional overshoot is desired — by default it extrapolates linearly past the range, which produces opacity > 1 or negative values.
|
||||
|
||||
### Scheduling: Sequence and Series
|
||||
|
||||
`<Sequence>` shifts time: children see `frame` reset to 0 at the sequence's `from`. Use it to place clips on a timeline. `<Series>` lays out back-to-back segments without manual offset math.
|
||||
|
||||
```jsx
|
||||
import {Sequence, Series, AbsoluteFill} from 'remotion';
|
||||
|
||||
export const Timeline = () => (
|
||||
<AbsoluteFill style={{backgroundColor: 'black'}}>
|
||||
<Sequence from={0} durationInFrames={60}><Intro /></Sequence>
|
||||
<Sequence from={60} durationInFrames={90}><Body /></Sequence>
|
||||
|
||||
{/* Series auto-sequences; offset overlaps the previous segment for crossfades */}
|
||||
<Series>
|
||||
<Series.Sequence durationInFrames={60}><ShotA /></Series.Sequence>
|
||||
<Series.Sequence durationInFrames={60} offset={-15}><ShotB /></Series.Sequence>
|
||||
</Series>
|
||||
</AbsoluteFill>
|
||||
);
|
||||
```
|
||||
|
||||
### Composition registration + parametric props
|
||||
|
||||
The `<Composition>` declares id, size, fps, duration, and `defaultProps`. A zod schema makes props type-safe and editable in the Studio sidebar.
|
||||
|
||||
```jsx
|
||||
import {Composition} from 'remotion';
|
||||
import {z} from 'zod';
|
||||
|
||||
export const schema = z.object({
|
||||
title: z.string(),
|
||||
accent: z.string(),
|
||||
fps: z.number().default(30),
|
||||
});
|
||||
|
||||
export const Root = () => (
|
||||
<Composition
|
||||
id="Promo"
|
||||
component={Promo}
|
||||
durationInFrames={150}
|
||||
fps={30}
|
||||
width={1080}
|
||||
height={1920}
|
||||
schema={schema}
|
||||
defaultProps={{title: 'Launch', accent: '#5b8cff', fps: 30}}
|
||||
/>
|
||||
);
|
||||
```
|
||||
|
||||
To make duration data-dependent, use `calculateMetadata` on the Composition to compute `durationInFrames` from props (e.g. number of rows × frames per row) before render.
|
||||
|
||||
### Audio and beat sync
|
||||
|
||||
```jsx
|
||||
import {Audio, staticFile, useCurrentFrame, useVideoConfig} from 'remotion';
|
||||
|
||||
const BEATS_SEC = [0.5, 1.0, 1.5, 2.0]; // detected offline
|
||||
export const Music = ({children}) => {
|
||||
const frame = useCurrentFrame();
|
||||
const {fps} = useVideoConfig();
|
||||
const t = frame / fps;
|
||||
const onBeat = BEATS_SEC.some((b) => Math.abs(t - b) < 1 / fps);
|
||||
return (
|
||||
<>
|
||||
<Audio src={staticFile('track.mp3')} />
|
||||
<div style={{transform: `scale(${onBeat ? 1.06 : 1})`}}>{children}</div>
|
||||
</>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
Detect beats offline (e.g. with `web-audio-beat-detector` or aubio) and bake the timestamps into props — never analyze audio at render time, since headless rendering has no realtime audio clock.
|
||||
|
||||
### Rendering
|
||||
|
||||
Preview in the browser-based Studio; render headlessly via CLI or the programmatic API.
|
||||
|
||||
```bash
|
||||
npx remotion studio # interactive preview
|
||||
npx remotion render Promo out/promo.mp4 \
|
||||
--props='{"title":"Launch","accent":"#f43"}' # pass parametric props
|
||||
npx remotion render Promo out.gif --codec=gif # GIF output
|
||||
```
|
||||
|
||||
For batch/data-driven pipelines, render in Node with `@remotion/renderer` (bundle once, render many) — see the reference file.
|
||||
|
||||
## Deliver & verify (rendered stills → MP4)
|
||||
|
||||
> **Packaged helper** (`scripts/`): tile your stills with `scripts/contact-sheet.sh sheet.png f-hook.png f-mid.png f-end.png`, then assert the encode with `scripts/probe-mp4.sh out.mp4 [WxH] [fps]`. See `scripts/README.md`.
|
||||
|
||||
Remotion is frame-deterministic by construction — every frame is a pure function of `useCurrentFrame()`, so you can render any exact frame headlessly with **no seek harness** (this is the heavy-tier counterpart to a web scene's `?t=N`). Use this tier when the output must be an MP4/GIF, must carry exact numbers/text, or batches from data; for a lightweight web animation, deliver standalone HTML instead.
|
||||
|
||||
**Output contract:**
|
||||
- A Remotion project with the composition registered (`<Composition>` + zod `schema` + `defaultProps`), all motion frame-driven (no timers / `Date.now()` / `Math.random()`).
|
||||
- Deliverable = the rendered `out/*.mp4` (plus the project, so the user can re-render with new props/data).
|
||||
- Duration data-dependent? compute it in `calculateMetadata`, not by hand.
|
||||
|
||||
**Verify loop — render stills → inspect → encode.** Render single frames first (cheap, no video encode), inspect them, and encode the full video only once the frames are right.
|
||||
|
||||
```bash
|
||||
# 1. Frame-exact stills at start / mid / end (PNG, headless, fast)
|
||||
npx remotion still Promo out/f-start.png --frame=0
|
||||
npx remotion still Promo out/f-mid.png --frame=75
|
||||
npx remotion still Promo out/f-end.png --frame=149 # last frame = durationInFrames - 1
|
||||
# render with the SAME props you'll ship, not just defaultProps:
|
||||
# npx remotion still Promo out/f-mid.png --frame=75 --props='{"title":"Launch"}'
|
||||
|
||||
# 2. Inspect each PNG — fidelity (matches brief; numbers/text correct) AND
|
||||
# artifacts (text overflow, off-canvas, clipped safe-area, missing font, wrong data binding).
|
||||
|
||||
# 3. Only after the stills check out, encode the video:
|
||||
npx remotion render Promo out/promo.mp4 --props='{"title":"Launch"}'
|
||||
```
|
||||
|
||||
- Use `npx remotion compositions` to read each composition's `durationInFrames`/`fps` and pick the end frame.
|
||||
- **Data-driven / batch**: verify ONE representative props set via stills *before* batch-rendering all rows — catch a layout bug once instead of N times.
|
||||
- **README demo GIF for free**: `npx remotion render Promo out/demo.gif --codec=gif` produces the first-screen proof clip (Direction D).
|
||||
|
||||
**Before you finish:**
|
||||
1. `npx remotion still` renders cleanly at frame 0, mid, and last — no errors, no missing assets/fonts.
|
||||
2. Numbers/text are exact and inside safe areas at every checked frame.
|
||||
3. Frame-driven only — no `Date.now()` / `Math.random()` / timers (determinism holds in CI).
|
||||
4. Props are zod-typed; the **shipped** props render correctly (not just `defaultProps`).
|
||||
5. Full MP4 encoded and plays; (optional) GIF rendered for the README.
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Need | Use |
|
||||
|---|---|
|
||||
| Map time → value | `interpolate(frame, [a,b], [x,y], {extrapolateRight:'clamp'})` |
|
||||
| Natural easing | `spring({frame, fps, config})` |
|
||||
| Place a clip at t | `<Sequence from durationInFrames>` |
|
||||
| Back-to-back shots | `<Series>` |
|
||||
| Audio | `<Audio src={staticFile(...)} />` |
|
||||
| Frame ↔ seconds | `frame / fps` |
|
||||
| Loop a value | `frame % period` then interpolate |
|
||||
| Editable props | zod `schema` on `<Composition>` |
|
||||
| CI render | `npx remotion render` or `@remotion/renderer` |
|
||||
|
||||
## Gotchas
|
||||
|
||||
- Never use `Math.random()`, `Date.now()`, or animation timers — they break determinism. Use `random(seed)` from Remotion for stable per-frame randomness.
|
||||
- Load fonts and assets via `staticFile()` and wait with `delayRender`/`continueRender`, or fonts pop in mid-render.
|
||||
- Default `interpolate` extrapolates — clamp it.
|
||||
- `useCurrentFrame` inside a `<Sequence>` is local (starts at 0); use `useVideoConfig().durationInFrames` for absolute timing.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/api-and-patterns.md` — `interpolate` options and `Easing`, spring config recipes, Sequence/Series scheduling, `<Audio>` + beat-sync, parametric `defaultProps` with zod + `calculateMetadata`, CLI flags, programmatic `@remotion/renderer` batch rendering, and embedding GLSL shaders / Three.js (`@remotion/three`).
|
||||
@@ -0,0 +1,224 @@
|
||||
# Remotion API & Patterns
|
||||
|
||||
Detailed cookbook for building and rendering Remotion compositions.
|
||||
|
||||
## interpolate: full options
|
||||
|
||||
```jsx
|
||||
import {interpolate, Easing} from 'remotion';
|
||||
|
||||
interpolate(frame, [0, 30, 60], [0, 100, 0], {
|
||||
easing: Easing.bezier(0.25, 0.1, 0.25, 1), // CSS ease equivalent
|
||||
extrapolateLeft: 'clamp', // 'extend' (default) | 'clamp' | 'identity' | 'wrap'
|
||||
extrapolateRight: 'clamp',
|
||||
});
|
||||
```
|
||||
|
||||
- Input ranges must be strictly monotonically increasing.
|
||||
- Multi-stop ranges (`[0,30,60]`) create keyframe chains.
|
||||
- `extrapolate: 'extend'` (default) continues the linear slope past the range — almost always clamp instead.
|
||||
- `extrapolate: 'wrap'` repeats the range — handy for loops.
|
||||
|
||||
Common easings: `Easing.linear`, `Easing.ease`, `Easing.in(Easing.quad)`, `Easing.out(Easing.cubic)`, `Easing.inOut(Easing.ease)`, `Easing.elastic(1)`, `Easing.bezier(x1,y1,x2,y2)`.
|
||||
|
||||
## spring: config recipes
|
||||
|
||||
```jsx
|
||||
spring({frame, fps, config, durationInFrames, delay, from, to});
|
||||
```
|
||||
|
||||
| Feel | config |
|
||||
|---|---|
|
||||
| Smooth, no overshoot | `{damping: 200}` |
|
||||
| Gentle bounce | `{damping: 12, stiffness: 100, mass: 1}` |
|
||||
| Snappy UI pop | `{damping: 20, stiffness: 200, mass: 0.6}` |
|
||||
| Heavy / slow | `{mass: 3, damping: 40, stiffness: 80}` |
|
||||
|
||||
- `from`/`to` remap the 0→1 output to any range without an extra `interpolate`.
|
||||
- `durationInFrames` time-stretches the spring to fit a fixed window.
|
||||
- `delay` postpones the start (in frames).
|
||||
- `measureSpring({fps, config})` returns how many frames the spring needs to settle — use it to size sequences.
|
||||
|
||||
```jsx
|
||||
// Spring from 100px down to 0, settling in 40 frames, starting at frame 10
|
||||
const y = spring({frame, fps, from: 100, to: 0, durationInFrames: 40, delay: 10});
|
||||
```
|
||||
|
||||
## Sequence & Series scheduling
|
||||
|
||||
```jsx
|
||||
<Sequence from={30} durationInFrames={60} name="Caption" layout="none">
|
||||
<Caption />
|
||||
</Sequence>
|
||||
```
|
||||
|
||||
- `from` offsets local time; inside, `useCurrentFrame()` returns `globalFrame - from`.
|
||||
- `durationInFrames` clips visibility; omit for infinite.
|
||||
- `layout="none"` removes the default absolute-fill wrapper (use when the child manages its own layout, e.g. inline text).
|
||||
- `name` labels the sequence in the Studio timeline.
|
||||
|
||||
Crossfade between two Series segments by overlapping with negative `offset` and fading:
|
||||
|
||||
```jsx
|
||||
<Series>
|
||||
<Series.Sequence durationInFrames={90}><Slide src="a.jpg" /></Series.Sequence>
|
||||
<Series.Sequence durationInFrames={90} offset={-20}>
|
||||
<Slide src="b.jpg" fadeInFrames={20} />
|
||||
</Series.Sequence>
|
||||
</Series>
|
||||
```
|
||||
|
||||
## Audio + beat sync
|
||||
|
||||
```jsx
|
||||
import {Audio, staticFile} from 'remotion';
|
||||
|
||||
<Audio
|
||||
src={staticFile('music.mp3')}
|
||||
startFrom={30} // trim: start 30 frames into the file
|
||||
endAt={300}
|
||||
volume={(f) => interpolate(f, [0, 30], [0, 1], {extrapolateRight: 'clamp'})} // fade-in
|
||||
/>
|
||||
```
|
||||
|
||||
Detect beats offline and store them as props:
|
||||
|
||||
```js
|
||||
// build step (Node) — produce beats.json
|
||||
import {guess} from 'web-audio-beat-detector';
|
||||
// ...decode audio buffer, then:
|
||||
const {bpm, offset} = await guess(audioBuffer);
|
||||
const period = 60 / bpm;
|
||||
const beats = Array.from({length: 64}, (_, i) => offset + i * period);
|
||||
```
|
||||
|
||||
Then in the component, snap motion to the nearest beat using `frame / fps`.
|
||||
|
||||
## Parametric defaultProps + zod + calculateMetadata
|
||||
|
||||
```jsx
|
||||
import {Composition} from 'remotion';
|
||||
import {z} from 'zod';
|
||||
import {zColor} from '@remotion/zod-types';
|
||||
|
||||
export const promoSchema = z.object({
|
||||
title: z.string(),
|
||||
rows: z.array(z.object({label: z.string(), value: z.number()})),
|
||||
accent: zColor(), // gives a color picker in the Studio
|
||||
});
|
||||
|
||||
export const Root = () => (
|
||||
<Composition
|
||||
id="DataPromo"
|
||||
component={DataPromo}
|
||||
fps={30}
|
||||
width={1080}
|
||||
height={1080}
|
||||
schema={promoSchema}
|
||||
defaultProps={{title: 'Q3', rows: [], accent: '#5b8cff'}}
|
||||
// Compute duration from data BEFORE rendering:
|
||||
calculateMetadata={({props}) => ({
|
||||
durationInFrames: 30 + props.rows.length * 20,
|
||||
})}
|
||||
/>
|
||||
);
|
||||
```
|
||||
|
||||
`zColor()` from `@remotion/zod-types` renders a color picker; plain `z.string()` renders a text field.
|
||||
|
||||
## CLI rendering
|
||||
|
||||
```bash
|
||||
# Render with inline props
|
||||
npx remotion render DataPromo out.mp4 --props='{"title":"Q3"}'
|
||||
|
||||
# Render from a props file (per-record batch)
|
||||
npx remotion render DataPromo out/$ID.mp4 --props=./data/$ID.json
|
||||
|
||||
# Quality / format flags
|
||||
npx remotion render Promo out.mp4 --codec=h264 --crf=18 --jpeg-quality=90
|
||||
npx remotion render Promo out.webm --codec=vp8
|
||||
npx remotion render Promo still.png --frame=45 # single still
|
||||
npx remotion render Promo out.gif --codec=gif --every-nth-frame=2
|
||||
|
||||
# Concurrency / scale
|
||||
npx remotion render Promo out.mp4 --concurrency=4 --scale=2
|
||||
```
|
||||
|
||||
## Programmatic render with @remotion/renderer
|
||||
|
||||
Bundle once, then render many outputs — the right pattern for data-driven pipelines.
|
||||
|
||||
```js
|
||||
import {bundle} from '@remotion/bundler';
|
||||
import {renderMedia, selectComposition} from '@remotion/renderer';
|
||||
import path from 'path';
|
||||
|
||||
const serveUrl = await bundle({entryPoint: path.resolve('src/index.ts')});
|
||||
|
||||
for (const record of records) {
|
||||
const composition = await selectComposition({
|
||||
serveUrl,
|
||||
id: 'DataPromo',
|
||||
inputProps: record, // drives calculateMetadata too
|
||||
});
|
||||
|
||||
await renderMedia({
|
||||
composition,
|
||||
serveUrl,
|
||||
codec: 'h264',
|
||||
outputLocation: `out/${record.id}.mp4`,
|
||||
inputProps: record,
|
||||
crf: 18,
|
||||
concurrency: 4,
|
||||
});
|
||||
}
|
||||
```
|
||||
|
||||
`selectComposition` resolves `calculateMetadata`, so per-record duration works automatically. Reusing one `serveUrl` across renders avoids re-bundling.
|
||||
|
||||
For stills, use `renderStill({composition, serveUrl, output, frame})`.
|
||||
|
||||
## Embedding shaders / Three.js
|
||||
|
||||
Use `@remotion/three` to drive a Three.js scene by frame:
|
||||
|
||||
```jsx
|
||||
import {ThreeCanvas, useVideoTexture} from '@remotion/three';
|
||||
import {useCurrentFrame, useVideoConfig} from 'remotion';
|
||||
|
||||
const Scene = () => {
|
||||
const frame = useCurrentFrame();
|
||||
return (
|
||||
<mesh rotation={[0, frame * 0.02, 0]}>
|
||||
<boxGeometry args={[2, 2, 2]} />
|
||||
<meshStandardMaterial color="#5b8cff" />
|
||||
</mesh>
|
||||
);
|
||||
};
|
||||
|
||||
export const ThreeComp = () => {
|
||||
const {width, height} = useVideoConfig();
|
||||
return (
|
||||
<ThreeCanvas width={width} height={height}>
|
||||
<ambientLight intensity={0.6} />
|
||||
<pointLight position={[10, 10, 10]} />
|
||||
<Scene />
|
||||
</ThreeCanvas>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
For a raw GLSL shader background, render a full-screen quad with a `shaderMaterial` whose `uTime` uniform is set to `frame / fps` each render (driven by `useFrame` in react-three-fiber). Because the canvas re-renders per Remotion frame, the shader is deterministic.
|
||||
|
||||
Heavy WebGL benefits from `--gl=angle` (or `swangle` for headless servers without a GPU) on the render command.
|
||||
|
||||
## Determinism checklist
|
||||
|
||||
- Replace `Math.random()` with `random('seed' + frame)` from `remotion`.
|
||||
- Wrap async asset loads with `delayRender()` / `continueRender(handle)`.
|
||||
- Preload fonts with `@remotion/google-fonts` or `delayRender` around `document.fonts.load`.
|
||||
- Avoid `Date.now()`, `performance.now()`, and real timers entirely.
|
||||
|
||||
---
|
||||
Drive every frame from props and Remotion renders video as code. Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=motion-design-skills&utm_content=skill_footer&utm_term=remotion-video)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -1,24 +0,0 @@
|
||||
# Short-Form Video
|
||||
|
||||
Engineer a 9:16 video that survives the swipe. Short-form is a retention machine, not a small launch film: a hook that earns the first second, pattern interrupts that reset attention every few seconds, and often a loop that buys watch-time past 100%. The craft is timing, not polish. For creators making Reels, TikToks, and YouTube Shorts (7–60s, vertical).
|
||||
|
||||
## When it activates
|
||||
|
||||
- "Make a Reels/TikTok/YouTube Shorts video" or "build a short-form video template."
|
||||
- "Write a hook for a short" or "fix the first 3 seconds."
|
||||
- "Edit for retention," "add pattern interrupts," or "pace cuts for a vertical video."
|
||||
- "Make a video loop" or "set up 9:16 safe areas."
|
||||
|
||||
## Example prompts
|
||||
|
||||
- "Build a reusable 9:16 short-form template — one design, many topics."
|
||||
- "Fix the drop-off in this Reel: tighten pacing and add pattern interrupts."
|
||||
- "Write three hook options for a 30-second educational short."
|
||||
|
||||
## What's inside
|
||||
|
||||
- `SKILL.md` — the retention arc, hook grammar, and pacing rules.
|
||||
- `references/retention-pacing.md` — deeper guidance on retention editing and pacing.
|
||||
|
||||
---
|
||||
Part of **[Short-Form Video Skills](../../)** · Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=tiktok-video-skills&utm_content=skill_footer&utm_term=short-form-video)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
@@ -1,155 +0,0 @@
|
||||
---
|
||||
name: short-form-video
|
||||
description: This skill should be used when the user asks to "make a Reels/TikTok/YouTube Shorts video", "build a short-form video template", "write a hook for a short", "edit for retention", "add pattern interrupts", "make a video loop", "fix the first 3 seconds", "pace cuts for a vertical video", or "set up 9:16 safe areas". Covers the hook→retention→loop grammar of vertical short-form (distinct from a launch hype film).
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Short-Form Video
|
||||
|
||||
Engineer a 9:16 video that survives the swipe. Short-form is not a small launch film — it is a retention machine: a hook that earns the first second, pattern interrupts that reset attention every few seconds, and often a loop that buys watch-time past 100%. The craft is timing, not polish.
|
||||
|
||||
## When to use
|
||||
|
||||
- Reels / TikTok / YouTube Shorts (7–60s, vertical 9:16).
|
||||
- Retention editing: fixing drop-off, adding pattern interrupts, tightening pacing.
|
||||
- Hooks, seamless loops, repeatable short-form templates (one design, many topics).
|
||||
|
||||
## Story structure (the retention contract)
|
||||
|
||||
Short-form storytelling is a contract with the viewer's thumb: every second must re-earn the next one. The narrative is shaped *backwards from the retention curve*, not forwards from an intro. Lock these before touching the timeline.
|
||||
|
||||
- **One idea per video.** A short carries exactly one payoff. A second idea splits attention and the curve sags in the middle — make it a second video instead. If you can't name the single takeaway in a sentence, you don't have a short yet.
|
||||
- **Open a loop in the first ~3s, close it last.** The hook poses a question or gap (the open loop); the payoff answers it. The viewer stays because the loop is open — so never answer it early, and never bury the payoff under setup.
|
||||
- **No dead air.** Every beat either advances the idea or resets attention. Cut anything that does neither. Silence, slow ramps, throat-clearing intros ("hey guys, in this video…") are where viewers swipe.
|
||||
- **Why the curve, not the cut, is the unit of work.** The algorithm distributes whatever holds the per-second retention curve flat; a "good" edit that flattens the curve loses to an ugly one that keeps it up. You are editing the *curve*, and the hook gates everything after it — fix it first.
|
||||
|
||||
The narrative spine (the `Hook → Setup → Body → Payoff/loop` arc is budgeted in the table below):
|
||||
|
||||
| Move | Story job | Failure if missed |
|
||||
|---|---|---|
|
||||
| Hook | Open the loop; promise the payoff | Swipe in first 3s |
|
||||
| Tension hold | Keep the gap open through the body | Flat middle, mid-drop |
|
||||
| Payoff | Close the loop — the one idea, delivered | "Wasted my time," no share |
|
||||
| Loop / CTA | Feed back to frame 0, or one clear ask | No re-watch, no action |
|
||||
|
||||
(`The retention arc` table below assigns seconds to each move; this section is the narrative logic and ordering behind it. `Hook grammar` lists the hook archetypes.)
|
||||
|
||||
## The one metric that drives everything
|
||||
|
||||
**Retention, not production value.** The algorithm distributes whatever holds viewers; ~87% decide to stay or swipe inside 3 seconds, and the threshold for distribution sits near **70%+ average view-through** (Shorts ~73%, TikTok ~78%, Reels ~65%). Every decision below trades against that curve. A flawless shot that bores at 0:04 loses to an ugly one that holds.
|
||||
|
||||
## The retention arc
|
||||
|
||||
| Beat | Job | Budget (of a 20s short) |
|
||||
|---|---|---|
|
||||
| Hook (0–1s) | One frame + one line that creates an open loop | 0–1s |
|
||||
| Setup | Pay off *why watch* — promise, stakes, or the gap | 1–4s |
|
||||
| Body | Deliver value in beats; one idea per beat | 4–17s |
|
||||
| Payoff / loop | Close the loop — or feed it back to frame 0 | 17–20s |
|
||||
|
||||
Scale the *body*, never the hook. A 60s short has a longer body and more beats — the same 1-second hook.
|
||||
|
||||
## Hook grammar (first ~1s)
|
||||
|
||||
The hook is a sentence + a visual that opens a curiosity gap the brain needs closed. Pick one pattern and commit:
|
||||
|
||||
| Hook type | Shape | Best for |
|
||||
|---|---|---|
|
||||
| Open loop | "The reason your X keeps Y…" (answer delayed) | most content |
|
||||
| Direct promise | "3 ways to ___ in 30 seconds" | educational / listicle |
|
||||
| Pattern interrupt | snap-zoom / whip-pan / visual mismatch in frame 1 | entertainment |
|
||||
| Negation | "Stop doing ___" / "You're ___ wrong" | strong opinions |
|
||||
|
||||
Two rules: the hook line is **on-screen as text by frame 1** (85% watch muted), and the most striking *visual* is also the first frame — never a slow ramp-in. Hold a strong opening frame; do not fade up from black.
|
||||
|
||||
## Pattern interrupts & cut rhythm
|
||||
|
||||
A pattern interrupt is any change that resets attention: a hard cut, zoom punch, b-roll insert, sound effect, caption pop, or angle change. In short-form, density is high — a *visual change every ~2–4 seconds*; videos with an interrupt roughly every 4s hold ~58% vs ~41% for a static talking head. The killer is a flat middle: TikTok throttles distribution when the watch-through curve flattens, so never stitch the body from low-energy takes.
|
||||
|
||||
```js
|
||||
// Drive every interrupt off the FRAME, not wall-clock — renders deterministically.
|
||||
import { useCurrentFrame } from "remotion";
|
||||
const fps = 30;
|
||||
const cuts = [0, 1.2, 3.0, 5.4, 8.0, 11.2, 14.0]; // seconds — uneven on purpose
|
||||
const frame = useCurrentFrame();
|
||||
const shot = cuts.findIndex((t, i) => frame < (cuts[i + 1] ?? Infinity) * fps);
|
||||
```
|
||||
|
||||
Vary the interval — a metronome reads as boredom; uneven cuts read as momentum. A punch-in zoom is the cheapest interrupt that needs no new footage:
|
||||
|
||||
```jsx
|
||||
import { useCurrentFrame, interpolate, spring, useVideoConfig } from "remotion";
|
||||
const PunchIn = ({ at, children }) => {
|
||||
const frame = useCurrentFrame(); const { fps } = useVideoConfig();
|
||||
const s = spring({ frame: frame - at * fps, fps, config: { damping: 12, stiffness: 200 } });
|
||||
const scale = interpolate(s, [0, 1], [1, 1.12]); // snap to 112% on the beat
|
||||
return <div style={{ transform: `scale(${scale})`, transformOrigin: "50% 45%" }}>{children}</div>;
|
||||
};
|
||||
```
|
||||
|
||||
## Captions
|
||||
|
||||
Most viewers watch muted, so on-screen captions are not optional — they carry the content and double as a pattern interrupt when each phrase pops in. Keep 4–7 words per line, high contrast, inside the safe area. **Do not re-implement caption animation here** — use the `caption-animation` skill for word-by-word timing, karaoke highlight, and styling; this skill only places captions on the retention timeline and inside the 9:16 safe zone.
|
||||
|
||||
## Loops & re-watch
|
||||
|
||||
A seamless loop counts each re-watch as fresh watch-time — a clean 8s loop watched 3× is 300% view-through, the strongest possible signal. Two techniques: match the **last frame to the first** (same composition, position, color) so the cut is invisible, and write a **sentence loop** — phrase the hook so the payoff feeds back into frame 0 ("…and that's why I never — " → loops to the start). Best length for re-watch is ~7–15s.
|
||||
|
||||
## 9:16 safe areas
|
||||
|
||||
Compose inside the universal safe zone so platform UI never covers text or faces.
|
||||
|
||||
| Zone | Pixels (1080×1920) | Avoid |
|
||||
|---|---|---|
|
||||
| Universal safe (all platforms) | center **900×1400** | anything critical outside it |
|
||||
| Top | keep clear ~120px | profile / sound UI |
|
||||
| Bottom | keep clear ~320px | captions, CTA, hashtags, audio bar |
|
||||
| Right | keep clear ~120px | like / comment / share rail |
|
||||
|
||||
Put the hook text and key subject in the center 900×1400 box. Reserve the bottom ~320px even if it looks empty in the editor — that is where the platform paints its own captions and buttons.
|
||||
|
||||
## Output checklist
|
||||
|
||||
- Hook line is on-screen text by frame 1; strongest visual is the first frame (no fade-up).
|
||||
- One open loop / promise set in the first ~4s.
|
||||
- A visual pattern interrupt every ~2–4s; intervals uneven; no flat middle.
|
||||
- Captions present (via `caption-animation`), 4–7 words, inside safe area.
|
||||
- Loop closed — matched first/last frame or a sentence that feeds frame 0.
|
||||
- Composed inside the 900×1400 center box; bottom 320px / top 120px / right 120px kept clear.
|
||||
- Every animated value derived from `useCurrentFrame()` — no CSS/library timers.
|
||||
|
||||
## Deliver & verify (rendered stills → MP4)
|
||||
|
||||
> **Packaged helper** (`scripts/`): tile your stills with `scripts/contact-sheet.sh sheet.png f-hook.png f-mid.png f-end.png`, then assert the encode with `scripts/probe-mp4.sh out.mp4 [WxH] [fps]`. See `scripts/README.md`.
|
||||
|
||||
The short ships as a Remotion composition (`<Composition>` + zod `schema` + `defaultProps`) — every cut, punch-in, and interrupt driven by `useCurrentFrame()`, never `Date.now()` / `Math.random()` / timers. Deliverable = `out/*.mp4` + the project (re-render per topic). 9:16 vertical (1080×1920) is the default.
|
||||
|
||||
**Verify loop — render stills → inspect → encode.** Three frames test the three things that make or break retention: the hook, the pacing, and the loop seam.
|
||||
|
||||
```bash
|
||||
# Stills at hook / mid-body / last frame — WITH SHIPPED PROPS (the real topic, not defaults)
|
||||
npx remotion still Short out/f-hook.png --frame=10 --props='{"topic":"...","hook":"..."}' # ~first 3s reads instantly
|
||||
npx remotion still Short out/f-mid.png --frame=N --props='{"topic":"...","hook":"..."}'
|
||||
npx remotion still Short out/f-end.png --frame=L --props='{"topic":"...","hook":"..."}' # L = durationInFrames-1
|
||||
|
||||
# Inspect each PNG:
|
||||
# - HOOK frame (~frame 0-ish / first 3s): hook line is on-screen text and reads instantly; strongest visual already up (no fade-up)
|
||||
# - LOOP seam: compare f-hook(frame 0) vs f-end(last frame) — same composition/position/color for an invisible loop cut
|
||||
# - 9:16 safe area: hook + key subject inside the center 900x1400 box; clear of top ~12%, bottom ~20-35% (captions/CTA/audio), right action rail
|
||||
|
||||
npx remotion render Short out/short.mp4 --props='{"topic":"...","hook":"..."}' # encode once stills verify
|
||||
npx remotion render Short out/demo.gif --codec=gif # README proof clip
|
||||
```
|
||||
|
||||
**Per topic / batch**: verify ONE topic via stills *before* rendering all of them. Use `npx remotion compositions` for `durationInFrames`/`fps` to pick the mid + last frames.
|
||||
|
||||
**Before you finish:**
|
||||
1. Stills render cleanly at the hook, mid, and last frame — no errors.
|
||||
2. The hook frame reads instantly (on-screen text + strongest visual up front, no fade-up); first vs last frame match for a clean loop seam.
|
||||
3. Hook + subject + captions sit inside the 9:16 center 900×1400 safe box at every checked frame.
|
||||
4. Frame-driven only — no `Date.now()` / `Math.random()` / timers; cuts uneven, no flat middle.
|
||||
5. A real topic's shipped props (not `defaultProps`) render correctly; MP4 encoded, GIF optional.
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/retention-pacing.md` — a complete runnable Remotion `<Short>` composition: a frame-driven shot scheduler with uneven cuts, the PunchIn pattern interrupt, an on-screen hook, a 9:16 safe-area overlay, and a seamless first/last-frame loop. Plus a hook-pattern library, a retention-curve debugging method, per-platform safe-zone maps, and the template×topic batch-render pattern.
|
||||
@@ -1,209 +0,0 @@
|
||||
# Retention Pacing — Short-Form Video Reference
|
||||
|
||||
Deep reference for the `short-form-video` skill: a full runnable Remotion short, a hook library, a retention-debugging method, per-platform safe zones, and batch output. All timing is frame-driven so renders are deterministic.
|
||||
|
||||
## 1. A complete `<Short>` composition
|
||||
|
||||
One self-contained vertical short: frame-driven shot scheduler with uneven cuts, a punch-in pattern interrupt, an on-screen hook, a 9:16 safe-area overlay (dev only), and a seamless loop. Topic content is a prop, so the same file renders many shorts.
|
||||
|
||||
```tsx
|
||||
// Short.tsx — 1080×1920, 30fps. Register at 600 frames (20s) for a clean loop.
|
||||
import {
|
||||
AbsoluteFill, useCurrentFrame, useVideoConfig,
|
||||
interpolate, spring, Sequence, Img,
|
||||
} from "remotion";
|
||||
|
||||
type Beat = { at: number; text: string; src?: string }; // at = seconds
|
||||
export type ShortProps = {
|
||||
hook: string; // on-screen by frame 1
|
||||
beats: Beat[]; // body, one idea each
|
||||
loopText: string; // sentence that feeds back to frame 0
|
||||
};
|
||||
|
||||
// --- frame-driven shot scheduler: uneven cuts read as momentum ---
|
||||
const useShot = (cuts: number[]) => {
|
||||
const { fps } = useVideoConfig();
|
||||
const frame = useCurrentFrame();
|
||||
let i = 0;
|
||||
for (let k = 0; k < cuts.length; k++) if (frame >= cuts[k] * fps) i = k;
|
||||
const startF = cuts[i] * fps;
|
||||
const endF = (cuts[i + 1] ?? cuts[i] + 2) * fps;
|
||||
const local = (frame - startF) / (endF - startF); // 0→1 within shot
|
||||
return { index: i, local, startF };
|
||||
};
|
||||
|
||||
// --- punch-in: the cheapest pattern interrupt, snaps on the cut ---
|
||||
const PunchIn: React.FC<{ startF: number; to?: number; children: React.ReactNode }> = ({
|
||||
startF, to = 1.1, children,
|
||||
}) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps } = useVideoConfig();
|
||||
const s = spring({ frame: frame - startF, fps, config: { damping: 13, stiffness: 220 } });
|
||||
const scale = interpolate(s, [0, 1], [1, to]);
|
||||
return <AbsoluteFill style={{ transform: `scale(${scale})`, transformOrigin: "50% 45%" }}>{children}</AbsoluteFill>;
|
||||
};
|
||||
|
||||
// --- hook: text present from frame 0, micro-overshoot, no fade-up ---
|
||||
const Hook: React.FC<{ text: string }> = ({ text }) => {
|
||||
const frame = useCurrentFrame();
|
||||
const { fps } = useVideoConfig();
|
||||
const pop = spring({ frame, fps, config: { damping: 9, stiffness: 260 } });
|
||||
const scale = interpolate(pop, [0, 1], [0.9, 1]);
|
||||
return (
|
||||
<div style={{
|
||||
position: "absolute", top: 360, left: 90, right: 90, // inside 900-wide safe box
|
||||
transform: `scale(${scale})`, transformOrigin: "0 0",
|
||||
fontFamily: "Inter, sans-serif", fontWeight: 800, fontSize: 76, lineHeight: 1.05,
|
||||
color: "#fff", textShadow: "0 4px 24px rgba(0,0,0,.55)",
|
||||
}}>{text}</div>
|
||||
);
|
||||
};
|
||||
|
||||
export const Short: React.FC<ShortProps> = ({ hook, beats, loopText }) => {
|
||||
const { fps, durationInFrames } = useVideoConfig();
|
||||
const frame = useCurrentFrame();
|
||||
const cuts = [0, ...beats.map((b) => b.at)]; // hook shot + one per beat
|
||||
const { index, startF } = useShot(cuts);
|
||||
const current = index === 0 ? null : beats[index - 1];
|
||||
|
||||
// seamless loop: match the LAST 0.5s back toward the opening frame
|
||||
const tailF = durationInFrames - frame;
|
||||
const loopFade = interpolate(tailF, [0, fps * 0.5], [0, 1], { extrapolateLeft: "clamp", extrapolateRight: "clamp" });
|
||||
|
||||
return (
|
||||
<AbsoluteFill style={{ backgroundColor: "#0b0b0f" }}>
|
||||
<PunchIn startF={startF}>
|
||||
{current?.src && <Img src={current.src} style={{ width: "100%", height: "100%", objectFit: "cover" }} />}
|
||||
</PunchIn>
|
||||
|
||||
{index === 0 && <Hook text={hook} />}
|
||||
|
||||
{/* body caption — placement only; styling/word-timing belongs to caption-animation */}
|
||||
{current && (
|
||||
<div style={{
|
||||
position: "absolute", left: 90, right: 90, bottom: 380, // above the 320px bottom UI band
|
||||
fontFamily: "Inter, sans-serif", fontWeight: 700, fontSize: 60, color: "#fff",
|
||||
textAlign: "center", textShadow: "0 3px 18px rgba(0,0,0,.6)",
|
||||
}}>{current.text}</div>
|
||||
)}
|
||||
|
||||
{/* loop sentence rises as the clip ends, easing the eye back to frame 0 */}
|
||||
<div style={{
|
||||
position: "absolute", left: 90, right: 90, bottom: 380, opacity: 1 - loopFade,
|
||||
fontFamily: "Inter, sans-serif", fontWeight: 800, fontSize: 64, color: "#fff", textAlign: "center",
|
||||
}}>{loopText}</div>
|
||||
|
||||
<SafeZone />{/* remove for final render */}
|
||||
</AbsoluteFill>
|
||||
);
|
||||
};
|
||||
```
|
||||
|
||||
Register it with a duration that *is* the loop length:
|
||||
|
||||
```ts
|
||||
// Root.tsx
|
||||
import { Composition } from "remotion";
|
||||
import { Short } from "./Short";
|
||||
export const Root = () => (
|
||||
<Composition
|
||||
id="Short" component={Short}
|
||||
durationInFrames={600} fps={30} width={1080} height={1920}
|
||||
defaultProps={{
|
||||
hook: "You're editing Shorts wrong —",
|
||||
beats: [
|
||||
{ at: 1.2, text: "Cut every 2–4 seconds", src: "b1.jpg" },
|
||||
{ at: 4.0, text: "Never on a fixed grid", src: "b2.jpg" },
|
||||
{ at: 7.5, text: "Kill the flat middle", src: "b3.jpg" },
|
||||
{ at: 11.0, text: "Match last frame to first", src: "b4.jpg" },
|
||||
],
|
||||
loopText: "…and that's why mine loop forever",
|
||||
}}
|
||||
/>
|
||||
);
|
||||
```
|
||||
|
||||
## 2. The 9:16 safe-area overlay (dev-only)
|
||||
|
||||
Render this on top while editing; delete it before the final export.
|
||||
|
||||
```tsx
|
||||
import { AbsoluteFill } from "remotion";
|
||||
export const SafeZone: React.FC = () => (
|
||||
<AbsoluteFill style={{ pointerEvents: "none" }}>
|
||||
{/* universal safe box: center 900×1400 in 1080×1920 */}
|
||||
<div style={{ position: "absolute", left: 90, top: 260, width: 900, height: 1400, outline: "2px dashed #39d", opacity: .6 }} />
|
||||
{/* platform UI bands to keep clear */}
|
||||
<div style={{ position: "absolute", top: 0, left: 0, right: 0, height: 120, background: "rgba(255,0,0,.12)" }} />
|
||||
<div style={{ position: "absolute", bottom: 0, left: 0, right: 0, height: 320, background: "rgba(255,0,0,.12)" }} />
|
||||
<div style={{ position: "absolute", bottom: 320, top: 120, right: 0, width: 120, background: "rgba(255,128,0,.10)" }} />
|
||||
</AbsoluteFill>
|
||||
);
|
||||
```
|
||||
|
||||
## 3. Hook pattern library
|
||||
|
||||
The hook is a sentence that opens a gap the brain must close, plus the most striking frame as frame 0. Copyable shapes:
|
||||
|
||||
| Pattern | Template | Note |
|
||||
|---|---|---|
|
||||
| Open loop | "The real reason your ___ keeps ___ …" | answer withheld until payoff |
|
||||
| Direct promise | "___ in 30 seconds, no ___ needed" | name the payoff + the timebox |
|
||||
| Negation | "Stop ___. You're doing it backwards." | strong-opinion content |
|
||||
| Pattern interrupt | snap-zoom / whip / hard visual mismatch | entertainment, no text needed |
|
||||
| Stakes | "This ___ cost me $___ — don't repeat it" | story / case-study |
|
||||
| Listicle | "3 ___ that ___ (the 3rd one ___ )" | tees up a forward reference |
|
||||
|
||||
Rules that matter more than the wording: (1) the hook text is on-screen by frame 1 — most viewers are muted; (2) frame 0 is the best frame, never a fade-up from black; (3) one gap only — stacking two hooks closes neither.
|
||||
|
||||
## 4. Retention-curve debugging
|
||||
|
||||
Treat the per-second retention graph (TikTok Analytics, YT Studio "Audience retention", Reels) as a profiler.
|
||||
|
||||
| Symptom on the curve | Likely cause | Fix |
|
||||
|---|---|---|
|
||||
| Cliff at 0:00–0:02 | weak hook / slow first frame | new hook line; cut the ramp-in; lead with the best frame |
|
||||
| Steady gentle decline | normal — leave it | — |
|
||||
| Flat-then-drop in the middle | low-energy body, cuts too sparse | add interrupts; tighten cut spacing to 2–4s |
|
||||
| Bump back up near the end | viewers are looping | lean in — tighten the loop, shorten total length |
|
||||
| Spike at one timestamp | a re-watched moment | front-load that beat or make it the hook |
|
||||
|
||||
Target ≥70% average view-through. Iterate the hook first — it moves the curve more than any other edit because it gates everything after it.
|
||||
|
||||
## 5. Per-platform safe zones
|
||||
|
||||
All on a 1080×1920 canvas. The universal box (center 900×1400) is the intersection; design inside it and one master fits every platform.
|
||||
|
||||
| Platform | Top clear | Bottom clear | Right clear | Notes |
|
||||
|---|---|---|---|---|
|
||||
| Universal (use this) | ~120px | ~320px | ~120px | center 900×1400 holds on all |
|
||||
| TikTok | ~108px | ~320px | ~120px | caption + button rail at bottom-right |
|
||||
| Instagram Reels | ~120px | ~310px | ~84px | audio bar + caption bottom |
|
||||
| YouTube Shorts | ~120px | ~300px | ~60px | channel + subscribe bottom band |
|
||||
| TikTok Ads | ~108px | ~370px | ~120px | extra room for the CTA button |
|
||||
|
||||
When unsure, design to the *largest* margin (TikTok Ads bottom 370px) and nothing ever clips.
|
||||
|
||||
## 6. Template × topic — batch output
|
||||
|
||||
The payoff of code-driven shorts: lock the design once, swap only the props, render N videos. Make topic content a prop (already done above), keep one theme object, and loop over JSON files.
|
||||
|
||||
```jsx
|
||||
// theme.ts — one source of truth so 50 shorts stay on-brand
|
||||
export const theme = { font: "Inter", hookSize: 76, bodySize: 60, bg: "#0b0b0f", fg: "#fff", accent: "#39d" };
|
||||
```
|
||||
|
||||
```bash
|
||||
# render the same Short template for every topic file in /topics
|
||||
for f in topics/*.json; do
|
||||
name=$(basename "$f" .json)
|
||||
npx remotion render Short "out/$name.mp4" --props="$f"
|
||||
done
|
||||
```
|
||||
|
||||
Each `topics/*.json` is one `ShortProps` object (`hook`, `beats`, `loopText`). Keep the hook line, cut points, and loop sentence per-topic; everything else stays fixed so the series reads as one brand.
|
||||
|
||||
---
|
||||
## Built by the team behind iart.ai
|
||||
|
||||
This skill is part of an open motion-graphics collection from iart.ai — the AI motion agent that turns data, scripts, and designs into editable motion graphics (Remotion → MP4). If you'd rather not hand-build this, iart.ai can batch-produce short-form videos from one template × data — change the text/data and re-export. → [iart.ai](https://iart.ai/?utm_source=github&utm_medium=reference&utm_campaign=tiktok-video-skills&utm_content=ref_footer&utm_term=short-form-video)
|
||||
Whitespace-only changes.
@@ -0,0 +1,185 @@
|
||||
---
|
||||
name: shot-composition
|
||||
description: This skill should be used when the user asks to "compose a frame", "set up a grid", "plan shot composition", "place the focal point", "decide where elements enter and exit", "add a parallax/camera move", "adapt a layout for 16:9 / 9:16 / 1:1", or "keep content inside safe areas". Covers grids, safe areas, focal hierarchy, and 2D/3D camera framing.
|
||||
version: 0.1.0
|
||||
---
|
||||
|
||||
# Shot Composition
|
||||
|
||||
Stage the frame and move the camera: grid systems, safe areas, focal hierarchy, negative space, parallax layering, and how layout guides the eye over time. Produces composition notes, layer structure (bg/mid/fg), and concrete camera-move plans with numbers.
|
||||
|
||||
## When to use
|
||||
|
||||
- Defining a frame's layout, focal point, and negative space.
|
||||
- Planning element enter/exit directions and motion paths.
|
||||
- Designing camera push/pull, pan, follow, and parallax moves.
|
||||
- Adapting one composition across 16:9, 9:16, and 1:1 with safe areas.
|
||||
|
||||
## Decision tree
|
||||
|
||||
```
|
||||
What are you staging?
|
||||
├─ A static frame's layout
|
||||
│ └─ set grid → place ONE focal point on a thirds power point
|
||||
│ → rank hierarchy (size>contrast>color>position) → reserve negative space
|
||||
├─ Where things enter / leave
|
||||
│ └─ follow reading direction; nearest edge; ease-out in, ease-in out
|
||||
│ (new content → from right/below; back/undo → reverse it)
|
||||
├─ A sense of depth / a camera move
|
||||
│ └─ split into bg/mid/fg layers (the three motion layers)
|
||||
│ → ONE camera move per beat (push | pan | follow | parallax), ease-in-out
|
||||
└─ One comp across aspects (16:9 / 9:16 / 1:1)
|
||||
└─ master in 16:9, focal in the center band → RESTACK (don't crop) for 9:16
|
||||
→ keep critical content inside title-safe margins
|
||||
```
|
||||
|
||||
**Cross-reference — apply the shared frameworks.** This skill is the spatial half of `animation-principles`. Apply them explicitly:
|
||||
|
||||
- **1/3 Rule** (`animation-principles`) has two forms and this skill uses both *spatially*: (a) *Composition thirds* — place the focal point on a rule-of-thirds power point (33%/67%); (b) *Distance* — no element travels more than ~1/3 of the frame without an intermediate keyframe or a scale/opacity change, so entrances stay readable.
|
||||
- **Three motion layers** (`animation-principles`: Primary / Secondary / Ambient) map directly onto depth: **foreground = Primary** (the move the eye follows), **midground = Secondary** (the subject + supporting reactions), **background = Ambient** (slow, low-contrast life). Parallax speed scales with that role.
|
||||
- Settle the **Three Pillars** (Emotional intent / Visual narrative / Motion craft) and the chosen **Motion Personality** (`motion-art-direction`) *before* numbering a camera move — premium wants slow long pushes, energetic wants whip-pans.
|
||||
|
||||
## Core workflow
|
||||
|
||||
### 1. Establish a grid
|
||||
|
||||
A grid gives alignment and rhythm. Default to a **12-column grid** for landscape web/video (it divides into halves, thirds, quarters, sixths) with gutters and outer margins. For a 1920px-wide frame: outer margin 80-120px, 12 columns, gutter 24-32px.
|
||||
|
||||
```
|
||||
content_width = frame_width - 2 * margin
|
||||
column_width = (content_width - (cols - 1) * gutter) / cols
|
||||
```
|
||||
|
||||
Example (1920 wide, margin 100, gutter 24, 12 cols): content 1720, column = (1720 - 11*24)/12 = 121.3px.
|
||||
|
||||
A **baseline grid** (vertical rhythm) of 8px is the standard spacing unit; size and space everything in multiples of 8 (8/16/24/32/48/64) for visual cohesion.
|
||||
|
||||
### 2. Place the focal point
|
||||
|
||||
Use the **rule of thirds**: divide the frame into a 3x3 grid; place the primary subject on a third-line or at an intersection (the "power points"), not dead center — unless the piece deliberately uses symmetry for stillness/formality. The intersections roughly fall at 33%/67% horizontally and vertically.
|
||||
|
||||
Build **visual hierarchy** in this priority order to direct the eye: **size > contrast > color > position**. The biggest, highest-contrast element wins attention first. Establish exactly one primary focal point per frame; secondary elements support, never compete.
|
||||
|
||||
### 3. Use negative space as rhythm
|
||||
|
||||
Do not fill the frame. Generous negative space around the focal point makes it read as premium and confident, and gives motion room to breathe. Whitespace is an active compositional element, not leftover.
|
||||
|
||||
### 4. Layer the scene (bg / mid / fg) for depth and parallax
|
||||
|
||||
Split the composition into depth layers: **background** (slowest, lowest contrast, often desaturated/blurred), **midground** (the subject), **foreground** (fastest, may blur as it crosses). Parallax = moving layers at different speeds to simulate depth.
|
||||
|
||||
These depth layers ARE the `animation-principles` **three motion layers** projected onto Z — each layer's role dictates its speed, contrast, and how much motion it earns:
|
||||
|
||||
| Depth | Motion layer | Parallax speed | Contrast / focus | What lives here |
|
||||
|---|---|---|---|---|
|
||||
| Background | Ambient | 0.1–0.3x | low, often desaturated/blurred | sky, gradient, texture, set dressing |
|
||||
| Midground | Secondary | 0.5–0.7x | mid; supports the subject | subject's environment, supporting elements |
|
||||
| Foreground | Primary | 1.0–1.5x | sharp on subject; FG props blur as they cross | the hero subject + closest props |
|
||||
|
||||
Foreground may exceed 1.0 to read as very close. The Primary move is the one the eye follows; never give two layers competing primary motion.
|
||||
|
||||
### 5. Camera language (works in 2D and 3D)
|
||||
|
||||
- **Push / pull** — scale into or away from the subject (dolly). Builds intimacy or reveals context. 800-2000ms, ease-in-out.
|
||||
- **Pan / tilt / move** — translate the frame across the scene. Reveals space, follows action.
|
||||
- **Follow** — keep a moving subject framed (often slightly off-center, leading the motion).
|
||||
- **Parallax** — layered move that sells depth.
|
||||
|
||||
**One shot, one intent.** Don't push + pan + rotate simultaneously — combined moves read as chaotic. Pick the single move that serves the beat. Inline timing essentials: camera moves use ease-in-out (`cubic-bezier(0.65, 0, 0.35, 1)`); never `linear` except continuous loops; slower is more cinematic.
|
||||
|
||||
### 6. Enter/exit direction carries meaning
|
||||
|
||||
Motion that follows reading direction (left-to-right, top-to-bottom in LTR locales) feels natural and smooth; reversing it creates tension or signals "back/undo." Match the direction to the intent:
|
||||
- New content arriving -> enter from the right / from below.
|
||||
- Dismissing / going back -> exit right or down, or reverse the entrance.
|
||||
- Drilling deeper -> scale up / push in.
|
||||
|
||||
Elements should usually enter and exit through the nearest frame edge with a slight overshoot on enter (ease-out) and acceleration on exit (ease-in).
|
||||
|
||||
### 7. Safe areas and multi-aspect adaptation
|
||||
|
||||
Keep all critical content (text, logos, faces, CTAs) inside a **title-safe margin** so nothing gets clipped or hidden behind platform UI.
|
||||
|
||||
| Aspect | Use | Title-safe margin | Notes |
|
||||
|---|---|---|---|
|
||||
| 16:9 | YouTube, web hero | ~5% all sides | broadcast title-safe historically 10% |
|
||||
| 9:16 | Reels, TikTok, Stories | top ~12-14%, bottom ~18-20%, sides ~6% | platform UI eats top and bottom |
|
||||
| 1:1 | feed posts | ~6-8% all sides | symmetric |
|
||||
| 4:5 | feed (taller) | ~6% sides, ~8% top/bottom | maximizes feed real estate |
|
||||
|
||||
Design the **focal point centered enough** that it survives all crops, then let secondary elements reflow per aspect. For 16:9 -> 9:16, restack horizontal arrangements vertically rather than just cropping the sides off.
|
||||
|
||||
## Worked examples
|
||||
|
||||
### GOOD — product hero, 16:9 → 9:16
|
||||
|
||||
> **Brief**: app launch hero, Premium personality, must work as a web hero and a Reel.
|
||||
>
|
||||
> - **Grid**: 1920×1080, margin 100, gutter 24, 12 cols. Phone mockup spans cols 7–11 (right power-column).
|
||||
> - **Focal**: phone screen on the right thirds intersection (1280, 360) — not centered. One focal point; the headline (cols 1–5) is secondary, lower contrast.
|
||||
> - **Depth (three motion layers)**: bg = soft gradient drift, 0.2x, Ambient. mid = phone mockup, 0.6x, Secondary stage. fg = one out-of-focus UI chip drifting past, 1.2x + 3px blur, reads as Primary depth cue.
|
||||
> - **Camera**: single push-in, scale 1.0→1.10 over 1600ms ease-in-out, origin on the phone screen — matches Premium (slow, no overshoot). No pan added.
|
||||
> - **9:16 adapt**: restack — headline moves *above* the phone, phone pulled up to the upper-middle to clear the bottom 19% CTA/nav zone, type +20%. Focal survives the crop because it started in the center band.
|
||||
>
|
||||
> Why it works: one focal point, depth roles assigned, one camera move, safe-area-aware reflow instead of a blind crop.
|
||||
|
||||
### ANTI-PATTERN — the "everything centered, everything moving" frame
|
||||
|
||||
> - Phone mockup dead-center, headline centered above, logo centered below → three elements fighting on the same axis, no hierarchy, no power point.
|
||||
> - All three layers slide left at the same 1.0x speed → no parallax, no depth, reads as "the whole template moved" (violates the **1/3 Rule / distance**: a full-width slide with no intermediate keyframe).
|
||||
> - Camera does push + pan + slight rotate at once → chaotic, no single intent.
|
||||
> - 9:16 made by cropping the 16:9 sides → headline and logo clipped, phone too low behind the nav bar.
|
||||
>
|
||||
> Fix: pick one focal power point, assign each layer a depth role + distinct parallax speed, one camera move, and restack for vertical instead of cropping.
|
||||
|
||||
## Common mistakes
|
||||
|
||||
| Symptom | Why | Fix |
|
||||
|---|---|---|
|
||||
| Frame feels flat / "the template moved" | all layers move at one speed; no depth roles | assign bg/mid/fg the Ambient/Secondary/Primary roles + 0.2/0.6/1.2x parallax |
|
||||
| Eye doesn't know where to look | 2+ elements at equal size/contrast, all centered | one focal point on a thirds power point; rank size>contrast>color>position |
|
||||
| Camera move reads as chaotic | push + pan + rotate combined | one move per beat; pick the one that serves the intent |
|
||||
| Content clipped on phones | designed for 16:9, cropped to 9:16 | restack vertically; keep critical content in title-safe margins (top 13% / bottom 19% on 9:16) |
|
||||
| Entrance reads as "slide", not "event" | element travels >1/3 of frame in one unbroken move | break with an intermediate keyframe or pair travel with scale/opacity (1/3 Rule, distance) |
|
||||
| Looks cheap / cramped | frame filled edge to edge | reserve negative space around the focal point; it's an active element |
|
||||
| Motion fights the story | numbers chosen before intent | settle Three Pillars + Motion Personality first, then number the move |
|
||||
|
||||
## Deliverable — a composition plan
|
||||
|
||||
A good composition plan is handed to an implementation skill (`gsap-web`, `remotion-video`, `after-effects`) and contains:
|
||||
|
||||
- **Frame spec**: aspect(s), resolution, fps, grid (cols / margin / gutter), baseline unit.
|
||||
- **Focal point**: coordinates (px or %) on a thirds power point + the hierarchy ranking of every element.
|
||||
- **Layer map**: each element assigned bg/mid/fg → Ambient/Secondary/Primary, with parallax speed.
|
||||
- **Enter/exit**: per element — edge, direction, easing family, and whether travel exceeds 1/3 (and how it's broken).
|
||||
- **Camera move**: the single move, its property + from→to values, duration, easing, transform-origin.
|
||||
- **Safe areas**: title-safe margins per target aspect + the restack plan for vertical.
|
||||
- **Cross-refs**: which Motion Personality and which Pillars decisions drive the timing.
|
||||
|
||||
### Before you finish — checklist
|
||||
|
||||
- [ ] Exactly one primary focal point, on a thirds power point (or deliberate symmetry).
|
||||
- [ ] Hierarchy ranked (size > contrast > color > position); secondaries don't compete.
|
||||
- [ ] Negative space reserved around the focal point.
|
||||
- [ ] Every element assigned a depth layer (bg/mid/fg = Ambient/Secondary/Primary) with a parallax speed.
|
||||
- [ ] One camera move, not stacked; eased (not linear unless a loop); 800–2000ms for hero moves.
|
||||
- [ ] Entrances/exits use nearest edge, follow reading direction, ease-out in / ease-in out.
|
||||
- [ ] No element travels >1/3 of frame without an intermediate keyframe or scale/opacity change.
|
||||
- [ ] Critical content inside title-safe margins for every target aspect; vertical is restacked, not cropped.
|
||||
- [ ] Motion Personality + Three Pillars settled before any number was chosen.
|
||||
|
||||
## Quick reference
|
||||
|
||||
| Need | Default |
|
||||
|---|---|
|
||||
| Grid | 12 columns, 8px baseline, multiples of 8 spacing |
|
||||
| Focal placement | rule-of-thirds power point, not dead center |
|
||||
| Hierarchy order | size > contrast > color > position |
|
||||
| Depth layers | bg / mid / fg, parallax 0.2x / 0.6x / 1.2x |
|
||||
| Camera | one move at a time, ease-in-out, 800-2000ms |
|
||||
| Enter / exit | follow reading direction; ease-out in, ease-in out |
|
||||
| 9:16 safe | top 12%, bottom 18%, sides 6% |
|
||||
|
||||
## Reference files
|
||||
|
||||
- `references/camera-and-grids.md` — grid math with worked numbers, full safe-area percentage tables per aspect, parallax layering code (CSS scroll + JS pointer + Three.js), and named camera-move recipes (push-in, whip-pan, orbit, dolly-zoom) with keyframe values.
|
||||
@@ -0,0 +1,153 @@
|
||||
# Camera moves, grids, safe areas, and parallax
|
||||
|
||||
Worked math and runnable code. Camera timings use ease-in-out (`cubic-bezier(0.65, 0, 0.35, 1)`); never `linear` except continuous loops.
|
||||
|
||||
## Grid math
|
||||
|
||||
```
|
||||
content_width = frame_width - 2 * margin
|
||||
column_width = (content_width - (cols - 1) * gutter) / cols
|
||||
column_x(i) = margin + i * (column_width + gutter) // 0-indexed
|
||||
span_width(n) = n * column_width + (n - 1) * gutter // width of n columns
|
||||
```
|
||||
|
||||
Worked examples:
|
||||
|
||||
| Frame | margin | gutter | cols | content | column | 4-col span |
|
||||
|---|---|---|---|---|---|---|
|
||||
| 1920x1080 | 100 | 24 | 12 | 1720 | 121.3 | 557.3 |
|
||||
| 1080x1080 | 64 | 16 | 12 | 952 | 64.7 | 306.6 |
|
||||
| 1080x1920 | 48 | 16 | 6 | 984 | 150.7 | 650.6 |
|
||||
|
||||
Baseline / spacing: use an 8px base unit. Allowed spacing values: 4, 8, 16, 24, 32, 48, 64, 96. Type sizes also on the scale (e.g. 16/20/24/32/48/64/96). Everything snapping to multiples of 8 produces automatic visual cohesion.
|
||||
|
||||
Rule-of-thirds lines for any frame: verticals at `W/3` and `2W/3`, horizontals at `H/3` and `2H/3`. Power points = the four intersections. For 1920x1080: verticals 640 / 1280, horizontals 360 / 720.
|
||||
|
||||
## Safe-area percentages
|
||||
|
||||
Margins are fractions of the frame dimension on that side. Keep text, logos, faces, and CTAs inside.
|
||||
|
||||
| Aspect | Top | Bottom | Left | Right | Reason |
|
||||
|---|---|---|---|---|---|
|
||||
| 16:9 (1920x1080) | 5% (54px) | 5% (54px) | 5% (96px) | 5% (96px) | general web/video; 10% for broadcast |
|
||||
| 9:16 (1080x1920) | 13% (250px) | 19% (365px) | 6% (65px) | 6% (65px) | top status/caption, bottom CTA bar + nav |
|
||||
| 1:1 (1080x1080) | 7% (76px) | 7% (76px) | 7% (76px) | 7% (76px) | feed crop symmetric |
|
||||
| 4:5 (1080x1350) | 8% (108px) | 8% (108px) | 6% (65px) | 6% (65px) | taller feed crop |
|
||||
|
||||
9:16 platform-specific danger zones (approximate, design around these): TikTok right-rail icons occupy the right ~12% mid-to-lower; caption/handle sit in the bottom ~15-20%; Instagram Reels similar. Center the focal subject so it survives all crops, then reflow secondary content per aspect.
|
||||
|
||||
Multi-aspect strategy: design the master in 16:9 with the focal point near the vertical center band; for 9:16, restack horizontally-arranged elements into a vertical stack (don't merely crop the sides), enlarge type ~15-25%, and pull the focal point up toward the upper-middle to clear the bottom UI.
|
||||
|
||||
## Parallax layering
|
||||
|
||||
### CSS scroll-driven (pure CSS, modern browsers)
|
||||
|
||||
```css
|
||||
@keyframes parallax { to { translate: 0 calc(var(--speed) * 200px); } }
|
||||
.layer { animation: parallax linear; animation-timeline: scroll(root block); }
|
||||
.bg { --speed: 0.2; }
|
||||
.mid { --speed: 0.6; }
|
||||
.fg { --speed: 1.2; }
|
||||
```
|
||||
|
||||
### Pointer / mouse parallax (JS)
|
||||
|
||||
```js
|
||||
const layers = [
|
||||
{ el: document.querySelector('.bg'), depth: 0.2 },
|
||||
{ el: document.querySelector('.mid'), depth: 0.6 },
|
||||
{ el: document.querySelector('.fg'), depth: 1.2 },
|
||||
];
|
||||
addEventListener('pointermove', (e) => {
|
||||
const x = (e.clientX / innerWidth - 0.5) * 2; // -1..1
|
||||
const y = (e.clientY / innerHeight - 0.5) * 2;
|
||||
for (const { el, depth } of layers) {
|
||||
el.style.transform = `translate3d(${-x * depth * 30}px, ${-y * depth * 30}px, 0)`;
|
||||
}
|
||||
});
|
||||
```
|
||||
|
||||
Background depth 0.1-0.3, midground 0.5-0.7, foreground 1.0-1.5. Foreground may exceed 1.0 to read as very close; apply a slight blur to foreground layers as they cross the subject.
|
||||
|
||||
### Three.js depth parallax
|
||||
|
||||
```js
|
||||
// Place planes at different z; perspective camera makes nearer planes move more
|
||||
camera.position.z = 5;
|
||||
bg.position.z = -8; // far -> moves least
|
||||
mid.position.z = -2;
|
||||
fg.position.z = 1.5; // near -> moves most
|
||||
function onMove(nx, ny) { // nx, ny in -1..1
|
||||
camera.position.x = nx * 0.6;
|
||||
camera.position.y = ny * 0.4;
|
||||
camera.lookAt(0, 0, 0);
|
||||
}
|
||||
```
|
||||
|
||||
## Named camera-move recipes
|
||||
|
||||
### Push-in (dolly in) — build focus/intimacy
|
||||
|
||||
```
|
||||
scale 1.0 -> 1.12 over 1600ms, ease-in-out
|
||||
origin: the subject (transform-origin at focal point)
|
||||
optional: slight bg blur ramp 0 -> 3px to add depth cue
|
||||
```
|
||||
|
||||
### Pull-out (reveal context)
|
||||
|
||||
```
|
||||
scale 1.15 -> 1.0 over 1800ms, ease-in-out, origin centered
|
||||
```
|
||||
|
||||
### Whip-pan (energetic transition between scenes)
|
||||
|
||||
```
|
||||
translateX 0 -> -120% over 280ms, ease-in (accelerate out)
|
||||
add motion blur (CSS: filter blur on X, or directional blur in compositor)
|
||||
incoming scene: translateX 120% -> 0 over 280ms ease-out, overlapped
|
||||
```
|
||||
|
||||
### Pan / move (reveal space)
|
||||
|
||||
```
|
||||
translateX 0 -> -300px over 1500ms ease-in-out
|
||||
keep the subject leading: frame it ~1/3 from the trailing edge so it "leads" into open space
|
||||
```
|
||||
|
||||
### Orbit (3D)
|
||||
|
||||
```js
|
||||
// angle sweep around subject at fixed radius
|
||||
const r = 6, a0 = 0, a1 = Math.PI / 4, dur = 2000; // 45deg over 2s
|
||||
// ease t in 0..1, then:
|
||||
const a = a0 + (a1 - a0) * easeInOut(t);
|
||||
camera.position.set(Math.sin(a) * r, camera.position.y, Math.cos(a) * r);
|
||||
camera.lookAt(target);
|
||||
```
|
||||
|
||||
### Dolly-zoom (vertigo) — unsettling, dramatic
|
||||
|
||||
Move the camera toward the subject while zooming out (or vice versa) so the subject stays the same size but the background warps.
|
||||
|
||||
```js
|
||||
// keep subject screen-size constant: as camera moves closer (z down), widen FOV
|
||||
const baseDist = 8, baseFov = 35;
|
||||
function dollyZoom(dist) {
|
||||
camera.position.z = dist;
|
||||
// maintain framing: fov such that subjectHeight subtends same angle
|
||||
camera.fov = 2 * Math.atan((Math.tan((baseFov * Math.PI/180)/2) * baseDist) / dist) * 180/Math.PI;
|
||||
camera.updateProjectionMatrix();
|
||||
}
|
||||
// animate dist from 8 -> 4 over ~2000ms ease-in-out
|
||||
```
|
||||
|
||||
### General rules
|
||||
|
||||
- One move per beat. Combining push + pan + rotate reads as chaos; pick the single move that serves the moment.
|
||||
- Slower = more cinematic and expensive-looking; 800-2000ms for hero camera moves.
|
||||
- Lead moving subjects: frame them off-center toward the trailing edge so there's open space in the direction of travel.
|
||||
- Enter/exit through the nearest edge: ease-out (slight overshoot) on enter, ease-in (accelerate) on exit, following reading direction for natural feel.
|
||||
|
||||
---
|
||||
Respect the grid and safe areas and the camera move guides the eye. Built by **[iart.ai](https://iart.ai/?utm_source=github&utm_medium=readme&utm_campaign=motion-design-skills&utm_content=skill_footer&utm_term=shot-composition)** — the AI motion agent for editable, on-brand motion graphics.
|
||||
Symlink
+1
@@ -0,0 +1 @@
|
||||
../../.agents/skills/after-effects
|
||||
+1
@@ -0,0 +1 @@
|
||||
../../.agents/skills/animation-principles
|
||||
Symlink
+1
@@ -0,0 +1 @@
|
||||
../../.agents/skills/beat-sync-editing
|
||||
@@ -1 +0,0 @@
|
||||
../../.agents/skills/caption-animation
|
||||
Symlink
+1
@@ -0,0 +1 @@
|
||||
../../.agents/skills/color-motion
|
||||
@@ -1 +0,0 @@
|
||||
../../.agents/skills/countdown-video
|
||||
Symlink
+1
@@ -0,0 +1 @@
|
||||
../../.agents/skills/logo-animation
|
||||
@@ -1 +0,0 @@
|
||||
../../.agents/skills/lower-thirds
|
||||
+1
@@ -0,0 +1 @@
|
||||
../../.agents/skills/motion-art-direction
|
||||
Symlink
+1
@@ -0,0 +1 @@
|
||||
../../.agents/skills/motion-background
|
||||
@@ -1 +0,0 @@
|
||||
../../.agents/skills/short-form-video
|
||||
Symlink
+1
@@ -0,0 +1 @@
|
||||
../../.agents/skills/shot-composition
|
||||
+46
-16
@@ -1,29 +1,59 @@
|
||||
{
|
||||
"version": 1,
|
||||
"skills": {
|
||||
"caption-animation": {
|
||||
"source": "iart-ai/tiktok-video-skills",
|
||||
"after-effects": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/caption-animation/SKILL.md",
|
||||
"computedHash": "e42c4828d3c6115dfb4d8fd3c3fd485b8e6a89c35a8835eee6dae39fb83210a4"
|
||||
"skillPath": "skills/after-effects/SKILL.md",
|
||||
"computedHash": "9db077d9664f5c45157c53b754d3c72bb346999a260c63d3d4892745a3eefd02"
|
||||
},
|
||||
"countdown-video": {
|
||||
"source": "iart-ai/tiktok-video-skills",
|
||||
"animation-principles": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/countdown-video/SKILL.md",
|
||||
"computedHash": "dafb9127ededa76d0df08be73adc46e18d6ec3955153f6ecfd0f928984ce95b2"
|
||||
"skillPath": "skills/animation-principles/SKILL.md",
|
||||
"computedHash": "140b00a44e08aa7649fb1cce1c48670ff2e76f3889fbac0b5311cc5bb818d7b7"
|
||||
},
|
||||
"lower-thirds": {
|
||||
"source": "iart-ai/tiktok-video-skills",
|
||||
"beat-sync-editing": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/lower-thirds/SKILL.md",
|
||||
"computedHash": "502c1a8a491e477dd855eb1a45469c09f6ff46a72ea841838d358d0a6a9fd790"
|
||||
"skillPath": "skills/beat-sync-editing/SKILL.md",
|
||||
"computedHash": "4089a0b65782c529ebd4fe9a576588fb51cf716573b4d683b0494e6554e61e03"
|
||||
},
|
||||
"short-form-video": {
|
||||
"source": "iart-ai/tiktok-video-skills",
|
||||
"color-motion": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/short-form-video/SKILL.md",
|
||||
"computedHash": "893ea19615f3199159b9fe94f05e37a5e538e7e0c3af7f77f9f9bed7ff933423"
|
||||
"skillPath": "skills/color-motion/SKILL.md",
|
||||
"computedHash": "7c05f7d35427d5fbe2fd1b5b6c0622252314a049797085fce085085cbf4c3e6e"
|
||||
},
|
||||
"logo-animation": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/logo-animation/SKILL.md",
|
||||
"computedHash": "3bba34771ff84fa7e68dc8026e195aab411a82b01091c9427fc8f3c8e48e7d99"
|
||||
},
|
||||
"motion-art-direction": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/motion-art-direction/SKILL.md",
|
||||
"computedHash": "bcba09b729fd3d99b3b1a388c3aaac655a3771b26dc3e8d077a110dfdd9e1631"
|
||||
},
|
||||
"motion-background": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/motion-background/SKILL.md",
|
||||
"computedHash": "99a576ac6549791b410abbc48a18074fed9e7925cf3ee5392108c59ad9e25fa4"
|
||||
},
|
||||
"remotion-video": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/remotion-video/SKILL.md",
|
||||
"computedHash": "940885087443f4aa5d3baf6edb4ed25edafea2435b52922e120200b4c1af99ba"
|
||||
},
|
||||
"shot-composition": {
|
||||
"source": "iart-ai/motion-design-skills",
|
||||
"sourceType": "github",
|
||||
"skillPath": "skills/shot-composition/SKILL.md",
|
||||
"computedHash": "9ff4551d175f951fb443a0680020871eb93c5ae07a0f36a2008ca98e5d0fa324"
|
||||
}
|
||||
}
|
||||
}
|
||||
Reference in new issue
Block a user