wizz-method 2.22.0 → 2.23.0

package/src/skills-lib/wizz-video-styles/references/film-styles/caption-news.md

@@ -0,0 +1,137 @@
+<!-- Vendored from lemo-opuscar (https://github.com/lemomo-ai/lemo-opuscar/tree/413a6bd6/styles/caption-news), MIT, (c) 2026 LemoLab; see ../../LICENSE-lemo-opuscar.txt. Paths under demo/ and core/ name that repo's engine, which is not shipped here: build the look in HyperFrames (../film-styles.md). -->
+
+# Caption News — Style Prompt
+
+> Social-video news that reads with the sound off. Real footage, photos and the source document are the ground; huge captions are the lead actor, slammed in a chunk at a time with one keyword on a yellow block. Hard cuts on a music grid, numbers that roll, a source tag that slides in whenever a figure appears, and the original document highlighted on screen before its number is believed. 16:9, built for a feed.
+> References (grammar only): the captions-first social news of the mid-2010s (NowThis, AJ+), horizontal explainers that set big type over footage (Bloomberg Quicktake). Never copy their logos, fonts, colour ratios, idents or shots, and never name them in the film.
+
+## 1. Essence, and what it is not
+
+1. **Sound-off first.** Every beat must be understood from the picture and the captions alone; the voice adds pace, not information.
+2. **The caption is the shot.** One Chinese line, set huge, slammed in two or three chunks, one keyword in a yellow block; one English line under it. The footage is chosen to sit under the line.
+3. **Evidence on screen.** Each number arrives with its source: a source tag slides in; key figures are first seen highlighted in the original document; per-shot footage credits sit in the corner like a news "Video:" tag.
+4. **A hook on frame 4.** The film's largest number lands in the first 0.15 s and rolls to its value.
+5. **One turn.** The bad-news hook is answered by a fact that changes its meaning; the rest of the film explains that turn, one new fact per shot.
+
+Not **Data Storytelling** (charts are short inserts between footage, never the whole film, never a dashboard). Not a **TV news package** (no anchor, no lower-third ticker, no studio, no scrolling crawl). Not a keynote: no headers, no centred title-and-bullets, no page turns.
+
+## 2. Frame and layout (1920×1080)
+
+| Zone | Position | Content |
+|---|---|---|
+| Footage | full bleed | video, photo, document or data card; never a vertical frame pillar-boxed in the middle |
+| Caption card | left aligned at x 120; Chinese baseline y ≈ 948, English baseline y ≈ 1014 | one line per language, never wrapped |
+| Channel bug | top left, x 96 y 52, 46 px high | yellow block with a fictional channel name + topic in Chinese and English small caps |
+| Source tag | under the bug, y 108, 36 px | white "来源" + dark label (`USDA NASS · 2026.08.03`, `… · 第 13 页`); only on data and document shots, slides in 0.2 s |
+| Footage credit | top right, 19 px on a 50 % dark chip | `视频 作者 · CC BY 3.0`; every CC BY shot carries it |
+| Lower third | x 120, y 742, 96 px high | yellow bar + label + one English line; for a place, date or speaker, never for the main sentence |
+
+Under the caption: a bottom gradient (0 → 84 % from y 560) plus a soft dark radial from the bottom-left corner; a top gradient (55 % → 0 over 200 px) under the bug and the credit. Data cards use a lighter gradient that starts at y 820. Graphics and charts sit right of centre and bleed to the frame edge; their title and source sit top left, never centred.
+
+Vertical 9:16 is a different layout, not a crop: put the caption under a 1080×1560 footage band (horizontal sources then scale ≤ 1.5×), stack the chunks on two lines there (the one-line rule is for 16:9 subtitles), and move the bug and source tag into the top safe area.
+
+## 3. Editing grammar
+
+- **Cuts on a 120 BPM half-beat grid** (every cut at a multiple of 0.5 s). Footage shots 1.5–3 s; data cards and documents 3–4.5 s; a shot that holds a comparison stays until every item has been fully on screen ≥ 2.5 s.
+- **Captions cross cuts.** A caption can run over two shots (picture changes, words don't); captions change at most every ~2.5 s even when shots change every 1.5 s. Reading time, not shot length, sets the caption clock.
+- **The order of a fact**: footage (feeling) → the number on a card → the document line that says it → the consequence. Never a number without its source tag in the same shot.
+- **Transitions are few and fixed**: hard cut (default); a 1–2-frame white flash on the turn and on the freeze; a card sliding in from the right with speed streaks (0.18 s); a split screen opening from its divider; grid panels rising staggered (0.1 s apart); a punch-in (0.3 s ease-out to ~1.25×) on the culprit; a **mask transition** where a photo grows out of a map shape. No dissolves, no page turns, no 3D flips.
+- **Pace**: a strong action within 0.5 s of every shot; no frame flat for more than 1 s (rolling counters, line drawing, highlighter sweeps, slow pushes keep it alive); one held breath (2–3 s) after the turn; two real silences.
+
+## 4. Caption card (the subtitle *is* the graphic)
+
+- **Chinese line**: Noto Sans SC 900, 88 px (verdict lines 110 px), ivory `#F7F5EF`, 20 px dark soft shadow. Split into 2–3 chunks with `|` in the timeline; each chunk slams 0.12 s after the previous: scale 1.3 → 1 in 0.13 s (back ease, overshoot 1.6), opacity full in 0.04 s.
+- **Keyword**: `{…}` inside a chunk; a honey-yellow block `#FFC21A` wipes left → right in 0.12 s behind it, text turns ink `#0E0E0C`. One keyword per caption, at most.
+- **English line**: Inter Tight 500, 33 px, ivory, fades and rises 10 px in 0.18 s, 0.1 s after the first chunk. It is the spoken line (numbers as digits), not a translation of the Chinese.
+- **One line per language, always.** If either language would wrap, split the sentence into two captions. A caption is on screen ≥ 1.8 s and ideally ≥ the spoken line + 0.6 s; the Chinese is a paraphrase short enough to read in that time (≤ ~14 characters).
+- The `.srt` is generated from the same timeline (`timeline.json` → `tools/cues.py` → `core/render/srt.py`): one Chinese line, one English line per cue.
+- A text-only beat (caption with no voice) is native to the style: use it for the breath after the turn.
+
+## 5. Graphics
+
+- **Ledger card** (the data-card ground): `#151512` sheet, faint ruled lines every 54 px (4.5 % white), a thin red double margin at x 78/86 — the "account book" that ties every chart together. It slides in from the right over 0.18 s; inside, a slow push (×1.025–1.03 over the shot).
+- **Colour meaning**: honey `#FFC21A` = the point / gains / the highlight; red `#FF4A3D` = losses, the spike, the bad number; ivory = everything else; grey `#8D8B83` = labels and units; `#5E5C55` = the bars that don't matter.
+- **Numbers**: Inter Tight 800 with fixed-width digit cells (`num()` in `demo/main.js`), so a rolling counter never jitters. Counters roll with ease-out and turn their colour when they arrive. Units in Chinese (箱, 万箱) set smaller after the number.
+- **Charts**: titled top left in Chinese (Noto Sans SC 900 64 px, ≤ 6 characters so it can be read in the shot) with one grey English line; lines draw segment by segment (eio); the key point gets a bigger dot, an expanding ring (0.7 s) and a block label; bars grow from the left one after another (first one slower and longer if it is the point).
+- **Map**: state shapes from Natural Earth, projected (Albers), sequential fill in five steps of one hue, states without data in dark grey with a "未单列" legend swatch; the subject state outlined in yellow with a leader and a label; the map pushes in on it.
+- **Documents**: the PDF page rendered at 220 dpi as white paper on a dark desk (±1–2° tilt, soft shadow, vignette), pushed in slowly; highlighter rectangles placed from `pdftotext -bbox` coordinates (1 pt = 3.0556 px), `multiply` blend, swept left → right in 0.3–0.5 s; the number read from it is re-set as a big yellow block beside the page. The printed text is picture; what must be read is re-set in type.
+- **Split screen**: two halves with a 4 px yellow divider that drops first; the "bad" half desaturated, darkened and tinted red, the "good" half in full grade; a label block (Chinese + English word) and a counter in each; a shared stamp centred on the divider.
+- **Grid**: three panels (632 px, 12 px gaps) rising staggered; each one lights up (yellow border, label turns yellow) on the word that names it; numbered 1–2–3.
+- **Freeze frame**: the shot stops on a frame, a 0.14 s white flash, punch 1.0 → 1.12, saturation drops to 45 %, yellow corner brackets close in on the subject, a timecode chip with a pause mark appears top right.
+- **Title sting**: a yellow ruled page (the ledger again, in yellow) wipes across the left 45 % over live footage; the title slams, the English title and a dateline chip follow; the page wipes off left and the title becomes the channel bug.
+
+## 6. Grade and image quality
+
+- One grade for every piece of footage: `contrast(1.07) saturate(0.92) brightness(0.98)` + a 6 % warm soft-light layer; documents 4 %. Light grain in the master (`mux.sh … 30 1`).
+- **Scale limits**: modern footage ≤ 1.5× (prepare frames at their final size with ffmpeg: crop, lanczos, `unsharp=5:5:0.45`); photos ≤ 1:1 of their source; archive material ≤ 2.5× after deblocking. Pushes stay inside those limits — a 720p source already at 1.5× gets no push.
+- Remove station logos by cropping (never cover them with a caption), blur legible private details (phone numbers, names on hives, number plates) with `delogo`, skip shots with burned-in text.
+- Captions, numbers and charts are always vector, never scaled footage.
+
+## 7. Sound palette
+
+| Layer | What | Level |
+|---|---|---|
+| Voice | one brisk narrator (Kokoro, speed 1.1–1.17), numbers spelled out in the TTS text | compressed, ≈ 8 dB over the music |
+| Music | 120 BPM pulse in sampled instruments: kit (kick on the beat, hats on off-beats), eighth-note electric bass, a sixteenth-note marimba ostinato, vibraphone off-beat chords, glockenspiel accents, staccato brass stabs on the big numbers, timpani and suspended cymbal for turns | ducked ~45 % under the voice |
+| Structure | the hook hit on frame 4; a hard stop into silence before the turn (only natural sound and the question); the turn lands on a major-key brass hit; half-time and low strings for the threat; a snare build into the spike; a second short silence before the verdict; a major chord with bells on the verdict | |
+| Foley | every graphic action has a sound: chunk slam (click + thump), counter rattle (a tick every 45 ms), whoosh into each card, highlighter squeak, camera shutter on the freeze, stamp thump on a big number, pops for labels, ticks as map states fill, a riser into the map push | |
+| Natural sound | the subject's own location sound (here a swarm), J-cut 0.3 s before the freeze, bed under the ending | not ducked below audibility |
+
+Master −14 LUFS, true peak −1.2 dB (`core/render/mux.sh`).
+
+## 8. Native moves
+
+1. **Frame-4 hook** — kicker line, then a giant number in a yellow block rolling to its value in 0.45 s, then the unit and the subject word (`hook()`).
+2. **Chunk slam with one yellow keyword** (`capCard()`).
+3. **Proportion by darkness** — a yellow line sweeps across the footage to the percentage; what it passed goes grey and dark; the figure appears on the dark side (`shotHalf()`). Works for any "X % of …".
+4. **The document says it** — headline or sentence highlighted from PDF bbox coordinates, then the number popped as a yellow block beside the paper (`paper()`, `shotDoc6()`, `shotQuote()`).
+5. **Freeze and question** — the image stops on the viewer's question, music cut to silence (`shotFreeze()`).
+6. **Ledger split** — two counters racing side by side, the difference stamped after both settle (`shotSplit()`).
+7. **Lit grid** — three ways / three causes, each panel lighting on its word (`shotGrid()`).
+8. **Punch and circle** — 0.3 s punch-in on the culprit, a circle drawn round it, a leader to its name and its Latin name (`shotSem()`).
+9. **Map to ground** — the map pushes into a region, then a photo of that place grows out of the region's outline (`shotMap()`).
+
+## 9. Material
+
+1. **Ask the user first** whether they have their own footage, photos, documents or data (and the rights to them). Their material comes first; the sources below are the fallback.
+2. **Where to look** (judge every file on its own page, not the search result):
+   - **Wikimedia Commons** (video and photos): read the licence on each file page. Fine: CC0, public domain, CC BY (credit author, title, licence on screen and in `CREDITS`). Not used: CC BY-SA, "no known restrictions", unclear licences, YouTube re-uploads, colourised or "restored" versions.
+   - **U.S. federal agencies** (USDA, USGS, NASA, NOAA, NARA): works of federal employees are public domain; check that the page does not credit a contractor or a third party. NASA: https://images.nasa.gov. NARA: https://catalog.archives.gov.
+   - **Internet Archive**: only items whose own page states public domain or a CC0/CC BY licence; U.S. works published before 1931 are public domain.
+   - **Library of Congress**: read each item's rights statement; "no known restrictions" is not enough.
+   - **Official documents and data** (reports, statistics, court filings) from the agency's own site: the source of every fact; keep `FACTS.md` (claim, URL, quoted wording, derived sums with their arithmetic).
+   - **Map geometry**: Natural Earth (public domain).
+3. **Before downloading, tell the user** what you will fetch, from where and under which licence, and keep within a size budget. The finder copied into this style, `demo/tools/footage.py` (`search commons|archive|nasa`, `fetch`, `used`, `md`), logs licence, author, page and the seconds used of each file into `SOURCES.jsonl` and writes `SOURCES.md`; `demo/tools/commons-tools/peek.py` previews a Commons video from thumbnails before downloading it, `seg.py` downloads only a segment of a long one, `commons_thumb.py` takes a standard-width thumbnail when the original is rate-limited. Wikimedia rate-limits bursts (HTTP 429): one request at a time with pauses.
+4. **People on screen**: no close-up or slow motion of an identifiable private person; public figures and people in the news event are fine. Never use generated images as footage.
+
+## 10. Engine
+
+Copy `demo/` into the project and replace the content:
+- `main.js` — the page (`window.render(t)` draws synchronously with placeholders first, so `TEXTS(t)` is valid for `readcheck.mjs`, then loads the frames it needs and draws again). Shot functions per format; `SHOTS` table for cuts.
+- `timeline.json` — captions (`zh` with `|` chunks and `{keyword}`, `en`), voice start times, silences and hits; read by the page, `tools/cues.py` (.srt + timing check) and `audio/mix.py`.
+- `tools/prep.sh` — every video shot as a 30 fps JPEG sequence at its display size (crop, lanczos, unsharp, delogo); report pages; photos; the map projection (`tools/map_prep.py`).
+- `audio/mix.py` — sampled score from the timeline, foley from `events.json`, voice, natural sound, ducking.
+- `tools/fonts.sh` (OFL subsets for every character in `main.js`), `tools/final_check.sh` (loudness, black frames, speech-to-text of the final mix).
+
+## 11. Pitfalls
+
+- **Text that can't be read in its shot.** Chart titles ≤ 6 Chinese characters; ornamental stamps (datelines) must recur readably elsewhere; a label that changes text (a counter) is reported to `TEXTS` with its final value. Run `readcheck.mjs`.
+- **Two-line Chinese captions in 16:9** read as a slide. Split the sentence instead.
+- **The voice running to the cut**: lines that end on the cut feel rushed; speed the TTS (≤ 1.17 — faster, Kokoro starts to mangle names: "varroa" became "Veroa") or shorten the line.
+- **Homophones in ASR**: "requeen the weak" was heard as "week"; "varroa mite" as "might". Rewrite the line ("give old ones a new queen", "varroa mites").
+- **Words stronger than the source**: the report says "stressor", not "killer"; quarterly losses summed are not "annual net loss". Write what the source says.
+- **Footage from elsewhere**: illustrative footage (here Dutch hives) never carries a place label it doesn't have; dates on photos (an almond orchard in 2014) are stated in the credit.
+- **Blank first frames on transitions**: start split/grid/sting animations 0.1–0.15 s before the cut so frame 1 already has picture.
+- **Station logos**: crop them out; never park a caption over a logo.
+
+## 12. Variation space
+
+The agent decides per film: topic, the turn, which formats are spent where, footage, the channel's fictional name and colour (keep one accent colour), the title.
+
+- **Three structures:** the inverted pyramid with a mid-film turn (bad number → "really?" → the document says otherwise → why); a list of three that turns on the third; a countdown of numbers that ends on the one that matters.
+- **Three openings:** the biggest number on frame 4; the document headline highlighted in the first second; a freeze on the event with a question.
+- **Three endings:** a two-part verdict caption on footage with music returning after a silence; back to the hook's number, now re-read; the last data card's key point becoming the closing caption.
+
+---
+
+How our demo was made (story, shots, score, end card, build): [DEMO.md](https://github.com/lemomo-ai/lemo-opuscar/blob/413a6bd6/styles/caption-news/DEMO.md). Read it after your treatment exists.

package/src/skills-lib/wizz-video-styles/references/film-styles/cel-anime-80s.md

@@ -0,0 +1,119 @@
+<!-- Vendored from lemo-opuscar (https://github.com/lemomo-ai/lemo-opuscar/tree/413a6bd6/styles/cel-anime-80s), MIT, (c) 2026 LemoLab; see ../../LICENSE-lemo-opuscar.txt. Paths under demo/ and core/ name that repo's engine, which is not shipped here: build the look in HyperFrames (../film-styles.md). -->
+
+# 80s Cel Anime — Style Prompt
+
+> Hand-painted cel animation from the golden age of Japanese OVAs, re-created in code: flat two-tone cels over airbrushed backgrounds, backlit neon, limited-animation timing, played back as a late-80s videotape on a CRT television.
+> References (grammar only): 1980s Japanese OVA and TV cel animation in general (painted-background vs flat-cel contrast, backlit light, held drawings, speed grammar, city-pop openings). Borrow the era's craft, never a specific work's characters, mecha, logos, place names or signature shots.
+
+## 1. Essence, and what it is not
+
+Four layers, and the contrast between them is the look:
+- **Background painting**: soft airbrushed gradients, glowing windows, wet reflections; painted once at load (colour + emissive pair), then only slid.
+- **Cel** (characters, vehicles, props): flat fills, **hard-edged** shadow and highlight shapes, **coloured** outlines, never pure black; stepped at 12 or 8 fps.
+- **Light**: backlit cel (透過光): signs, lamps, title letters glowing *through* the frame; flares, halation.
+- **Playback**: videotape on a CRT: chroma bleed, scanlines, grille, phosphor glow, darkened corners, faint noise.
+
+If a frame looks like modern digital anime, one layer is missing. Not moe anime (round eyes, giant irises, hatched blush), not vector clip-art (uniform black lines, gradients in shapes), not a film-print pastiche (no gate weave or scratches: this is tape).
+
+## 2. Materials & rendering
+
+**Cel rendering.** Fill, then a **hard** shadow crescent, then a **hard** highlight crescent, then the coloured outline. No gradients inside a cel shape (exception: one "harmony" painted still at a peak). Automatic shading: clip to the shape and fill `shape − shape shifted toward the light` with the shadow colour (even-odd); the same shifted away for the highlight and for a saturated neon **rim light** (2–4 px). Add hand-placed cast shadows. Outlines ~2–3 px at 1080p, scaled down (0.45–0.8×) in extreme close-ups. **Colour models per lighting** (色指定): define each character in daylight, derive other sets by multiplying lit / shadow / line separately, plus a backlit silhouette set with a warm rim.
+
+**People the 80s way.** Construct the head (sphere + jaw): long egg-shaped face, soft-point chin, far eye foreshortened with a cheekbone turn in 3/4. **Almond** eyes near the vertical middle of the head, a thick upper lash line flicking outward, a tall oval layered iris, two highlights, long thin brows. Nose = a shadow plane + one stroke; one faint blush layer, never hatched. Hair in overlapping locks with sharp uneven tips over a wider dark under-layer, a continuous "angel ring" highlight with a zigzag lower edge. Lines: skin thin warm dark brown, hair dark purple; only hard props near-black. Emotion lives in brows, lid height, iris size, mouth corners and shoulders. Build and approve a **model sheet** first; draw every shot from the same head function.
+
+**Backgrounds.** Painted once at load: sky gradients, airbrushed clouds with lit undersides, buildings or landscape with painted light sources, and always a matching **emissive plate** that slides with the colour plate. Wet ground = a flipped, blurred, streaked copy of the plate darkened with ripple bands. Generic sign words only, never brands. Moving plates get a light horizontal smear (~16 px) against strobing.
+
+**Light and lens.** Bloom strongly from the emissive canvas, weakly from the colour canvas above ~0.9; tint the widest bloom red-orange (halation). Flares: star-burst, anamorphic streak, hexagonal ghosts through frame centre. Stamp every cel's silhouette in black onto the emissive canvas so glow never bleeds through it.
+
+**Videotape + CRT, restrained and deterministic:** chroma bleed (sharp luma, horizontally blurred and slightly shifted chroma), fine scanlines anchored to output pixels whose gap shrinks on brights, a faint RGB grille, phosphor glow, darkened corners without curved borders, faint signal noise. Subtitles are composited inside the pass after grading. Two optional accents: a **power-on** (dot → line → picture) and **tracking noise** (a few frames of tearing, rolling band, vertical hop). Check moiré on a 720p downscale. Frame: full 16:9 as a theatrical feature (4:3 pillarbox is valid for an OVA feel).
+
+## 3. Colour logic
+
+- Painted plates carry gradients and atmosphere; cels are flat, few-colour, saturated. Each time of day gets its own colour model, never a filter over the frame.
+- The protagonist has **one accent colour** that pops in every lighting set. Line colours are dark versions of their fill.
+- Skin gets its own colour in every set; multiplying daylight skin by a tint turns faces pink under neon.
+- Examples: night city (indigo, violet, neon magenta, cyan, window gold); seaside noon (cobalt, white cumulus, sand, coral); classroom dusk (amber, chalk green, brown wood, purple shadows).
+
+## 4. Type & subtitles
+
+- Subtitles live on the tape, composited inside the CRT pass: a semi-condensed sans (e.g. Barlow Semi Condensed 600, ~50–56 px), cream-yellow with a thick dark outline and soft shadow, ~96 px above the bottom; a second speaker may take a second colour and a small boxed tag. Hold ≥ max(1.8 s, speech + 0.6 s); a line yields to the next.
+- Titles: heavy italic Latin in a **chrome gradient** (sky → white horizon line → violet → pink → gold), dark + neon outlines, optional katakana line (e.g. Dela Gothic One), glowing through the emissive layer, flickering on at 12 fps with a light sweep and a star glint.
+
+## 5. Motion quality
+
+- **Characters on twos (12 fps)**; performance beats (blinks, eye openings, lip flaps, gestures) on threes (8 fps). Camera, light sweeps, flares and rain on ones (24 fps); a stepped camera judders.
+- **Cycles**: hair and cloth as ribbons driven by a travelling sine, 4-drawing cycle (6 when calm); bounce = a 12 fps jitter of a few px.
+- **Move light, not drawings**: coloured bands sweeping across a held cel (`source-atop`).
+- **Impact frames**: 2 frames of an inverted or monochrome silhouette over radial lines, then the result.
+- **Hold the peak**: a painted still with a slow pan, only cycles moving.
+- Fast blinks (one 8 fps closed drawing); eye openings in three drawings; lip flaps of 3–4 shapes at 8 fps ending on a held expression.
+
+## 6. Camera grammar
+
+A vocabulary, not a route. Opening and ending come from the topic.
+
+| Move | What it expresses | Can serve |
+|---|---|---|
+| Multiplane crane over a tall plate | a world revealed layer by layer | arriving; descending into a place; rising above it |
+| Side tracking, 3–4 parallax layers | steady momentum | a journey; a chase; progress |
+| Rear view into a pseudo-3D canyon | speed, pressure ahead | a race; a tunnel; a countdown |
+| Held bust with light bands | speed or tension without animation | a drive; waiting; a decision |
+| Insert with bokeh and a sliding highlight | this object matters | a letter; a key; a ticket |
+| Extreme close-up on eyes | resolve, recognition | a choice; a promise |
+| Slow pan across a painted still | awe, a held breath | a peak; a memory; a landscape |
+| Radial speed lines + one-beat freeze | the instant before impact | a leap; a collision; a goal |
+| Truck-in through a window or screen | entering another world | a flashback; a broadcast; a dream |
+| Locked two-shot with lip flaps | conversation | banter; a confession; an argument |
+| An earlier composition repainted | what changed | after a storm; years later |
+
+Cut on bars. Framing: cels readable against the painted plate; keep the lower ~120 px clear for subtitles. Transitions: cuts, white flashes, impact frames, tracking noise at a tape-related beat; no digital dissolves or wipes.
+
+## 7. Sound palette
+
+- **Instruments** (city pop, synth funk, anime-opening pop): saxophone or synth-brass lead, FM electric piano with a bell transient, slap bass, chorus-y 16th guitar chanks, gated-reverb snare and big toms, drum machine, arpeggiated analog synth, string pad, FM bell sparkle. Extended jazz-pop harmony (maj7, 9ths).
+- **Techniques, as options**: band-pass the score to cassette/AM when a diegetic source plays, then open it; a brass hit on a title; a drum fill left open for an effect; one beat of true silence before a peak; a semitone lift for a last chorus.
+- **Foley** (synthesized): engines from an RPM curve, rain, tire hiss, wind, bells, radio squelch, impacts, sparks, cassette clicks, crowd murmur, door buzzers, footsteps on wet pavement; tape accents get their own thunk, whine, hiss.
+- **Voice**: an English-dub feel, short lines; a second voice through a phone/radio band-pass (~380–2800 Hz). Verify every line with whisper.
+- **Mix**: limit then compress TTS, voices ~10 dB over ducked music (duck ~9 dB), beds ducked ~3 dB, choruses louder than intros, two-pass loudnorm to −14 LUFS.
+
+## 8. Native moves
+
+A menu: use the ones your story needs.
+
+- **Small flat hero in a huge painted world.** *Fits content like:* a first day in a big company; a child in a museum hall; a lone researcher at an observatory.
+- **Stillness as drama** (held close-up or harmony still). *Fits content like:* an exam result; a goodbye at a gate; an unveiling.
+- **Backlit light.** *Fits content like:* a lantern festival; a screen full of code; a lighthouse; stage lights.
+- **Speed grammar** (speed lines, looping plates, impact frames). *Fits content like:* a sprint final; a data packet crossing the world; a kitchen rush.
+- **Light-band sweep on a held cel.** *Fits content like:* a night train; a scanner reading a page; a heartbeat quickening.
+- **A song as an object** (tape, radio, jukebox changes the score). *Fits content like:* a band's origin; a family memory; a call-in show.
+- **Tape artefacts as punctuation.** *Fits content like:* found home video; a rewind to the past; an interrupted broadcast.
+- **Colour-model change** (one place, another hour). *Fits content like:* a day in a shop; a farm's seasons; before/after a renovation.
+
+## 9. Pitfalls of the medium
+
+- White cels bloom pink → bloom from the emissive canvas; weak, high-threshold bloom on colour.
+- Neon bleeds through characters → silhouettes stamped into the emissive canvas.
+- Black holes in highlights → clamp before S-curves.
+- Thick rim light swallows thin shapes → 2–4 px offsets.
+- Hair like tentacles, a face like a paper mask, moe eyes, uniform black lines → the drawing rules in §2.
+- Pink faces under coloured light → skin gets its own colour per set; weaker light sweeps on faces.
+- Front views are the hardest drawing → hide them behind a visor, a windscreen or backlight when you can.
+- Pseudo-3D points behind the camera flip into the sky → clamp z to the near plane.
+- Canvas silently falls back to a serif → `document.fonts.load` every weight before painting.
+- CRT detail inflates files → noise in the shader, only `noise=c0s=2:allf=t` in ffmpeg, CRF ≈ 23; loudnorm in two passes with `linear=true`; don't let a limiter flatten the dynamics.
+
+## 10. Engine
+
+In `demo/`: `cel.js` (path, cel, ribbon, flutter), `pal.js` (colour models), `head80.js` (80s head, 3 views × expressions), `hands.js`, `bg.js` (painted plates, crane plates, wet reflections), `fx.js` (rain, speed lines, flares, trails), `post.js` (WebGL tape + CRT pass: bloom, grade, subtitles, power-on, tracking noise), `hud.js`, `music/score.py` (numpy FM/slap/brass synthesis), `tools/`. `?test=model|close|side` draws model sheets, `?raw=1` skips the CRT pass. A minimal new use: a harbour sunset plate with its emissive pair, one cel boat shaded with `cel.js`, through `post.js`.
+
+## 11. Variation space
+
+You decide the structure, the characters, the world, the opening, the ending, the camera path, the pacing and the colour models. All far from our demo:
+
+- Structures: **slice of life in four seasons** (one place, four colour models, a small change each time); **two-hander dialogue** (a locked two-shot comedy that breaks into one fantasy sequence); **the mock opening credits** (a character roster introduced one by one, each with a pose and a name card).
+- Openings: **in the middle of a conversation** (lip flaps already moving, no establishing shot); **a TV being switched on** to an in-world broadcast; **a close-up of a drawing hand** in a classroom, then cut to the world it draws.
+- Endings: **a freeze-frame** on a laugh with a "to be continued" card; **the tape runs out** into blue screen; **a slow pan up to a starry sky** from a quiet rooftop, no dialogue.
+
+---
+
+How our demo was made (story, shots, score, end card, build): [DEMO.md](https://github.com/lemomo-ai/lemo-opuscar/blob/413a6bd6/styles/cel-anime-80s/DEMO.md). Read it after your treatment exists.

package/src/skills-lib/wizz-video-styles/references/film-styles/city-symphony.md

@@ -0,0 +1,142 @@
+<!-- Vendored from lemo-opuscar (https://github.com/lemomo-ai/lemo-opuscar/tree/413a6bd6/styles/city-symphony), MIT, (c) 2026 LemoLab; see ../../LICENSE-lemo-opuscar.txt. Paths under demo/ and core/ name that repo's engine, which is not shipped here: build the look in HyperFrames (../film-styles.md). -->
+
+# Time-lapse Symphony — Style Prompt
+
+> A city film cut like a piece of minimalist music: time-lapse turns traffic, crowds and clouds into rivers of light, the cuts get denser with the score's subdivisions, one thing is slowed almost to a stop, and the film ends on a hard cut to black on a downbeat. No narrator. The music carries it.
+> Grammar learned from the city-symphony tradition and the "Qatsi" films (Reggio / Fricke, scores in the Philip Glass manner). Nothing is taken from those films: no footage, no melody, no title lettering, no names in the film.
+
+## 1. Essence, and what it is not
+
+- **An edit style.** The picture is real footage of a place (time-lapse, hyperlapse, aerials, slow motion, archive film, photographs). The style lives in *how time is bent* and *where the cuts fall*, not in a drawn look.
+- **The score is the script.** It is written first, with a clear arc: minimal repetition → acceleration → climax → a sudden stop → acceleration again → hard stop → a quiet coda. Every cut sits on a beat or a subdivision of it.
+- **Machines fast, people slow.** Cars, trains, clouds and light are sped up until they become lines; the one slowed moment is something fragile (a soap bubble, steam, a flag, a flock), and people appear as crowds, tracks or long-exposure ghosts.
+- **Few words, each one earned.** On-screen text is a title, a handful of numbers that state the size of the place (each from an official source), an archive date, and one line from a real public-domain text at the end.
+- **It is not** a travel reel (no drone-porn montage on a pop track), a vlog, a news package, an explainer with a narrator, or a slideshow of postcards. No lower thirds with names, no headers, no centred title cards with page-turn transitions, no logo stings.
+
+## 2. Materials & rendering
+
+- Footage only (2D). Sources mixed freely: tripod time-lapse, long single-camera time-lapse (≥ 12 h, sunrise and sunset in the same frame), hyperlapse, drone aerials (top-down is gold), high-frame-rate slow motion, archive film (pre-1931), photographs (a pair from the same camera position is gold), a scanned page of a historic text.
+- Everything is unified into one **35 mm release-print** look: 1.85:1 matte (21 px bars on 1920×1080), warm highlights, lifted blacks, highlight bloom, clumped mid-tone grain, ±0.5 px gate weave, soft vignette.
+- Never generated images; never restored, colourised or speed-corrected archive copies; never YouTube mirrors of archive films.
+- Upscale limits: archive ≤ 2.5× (denoise → lanczos → light sharpen → grain); modern ≤ 1.5×.
+
+## 3. Colour logic
+
+| Use | Value |
+|---|---|
+| Type (warm white) | `#EDE8DF` |
+| Secondary line (English, labels) | `#B0AAA0` |
+| Sodium orange — the only accent (rules under numbers, source lines, crowd tracks, the day ring's hand) | `#D9772F` |
+| Matte and black | `#000000` |
+
+Grade (`tools/render.py → make_luts / grade`): per-shot partial grey-world white balance first (only pixels with luminance 0.12–0.88, strength 0–0.6; **0 for night shots**, which keep their sodium and blue), then one LUT for all shots: S-curve (sigmoid slope 6.2, pivot 0.47, blended 55/45 with linear) → black 7/255, white 250/255 → R ×1.025, B ×0.965 (warm highlights, a little cold in the shadows) → saturation 0.86. Bloom: pixels above 170/255, blurred 6 px at ⅛ size, added back × (1, .82, .62); 0.09 by day, 0.15–0.2 at night. Grain: half-resolution Gaussian noise up-sampled and blurred 0.6 px, weighted `0.35 + 1.3·L·(1−L)`, strength 9/255. Archive film stays black-and-white and goes through the same LUT, so it reads as part of the same print.
+
+## 4. Type & subtitles
+
+- Latin: **Michroma** (OFL), capitals, tracking 0.06–0.32 em. Chinese: **Noto Sans SC** variable, weight 300–350, tracking 0.08–0.14 em. Credits: Noto Sans SC 17 px.
+- **Left-aligned**, left margin 168–196 px, never centred, never on a box. Legibility comes from a very soft elliptical darkening behind the block (α ≤ 0.6), not a panel.
+- **Number card**: the number in Michroma 84 px (44 px when three numbers stack); a 2 px sodium rule grows under it; one Chinese line (28 px) and one English line (14–16 px) say what it counts and the source year. A census-type number locks digit by digit on sixteenths (one glockenspiel note per digit); a flow number counts up from 0 in 0.6 s (ease-out). "About" numbers carry `≈` and "约".
+- **Archive date**: just the year, Michroma 22 px, bottom-left of the archive panel.
+- **Closing quote**: on the page shot the whole sentence is lit; on the card, a short fragment over the last moving shot — Chinese 60 px, English original in capitals 20 px, source line 14 px sodium, fading in within 0.25 s and held just past the reading time.
+- One line per language, never wrapped; every text block stays fully in frame at least CJK ÷ 4.5 + other ÷ 15 + 1.5 s (`readcheck.mjs`). Text may stay up across cuts underneath it, but only while the shots under it belong to what it counts.
+- `.srt`: the film has no voice, so the subtitle file carries the on-screen text (one Chinese and one English line per cue) for muted or accessible playback.
+
+## 5. Motion quality
+
+- Speed is the main instrument. Shot speeds run from 0.3× to 75×; every speed above 1× uses **frame blending** (a 360° shutter): each output frame averages the source frames it covers, capped at N frames. Headlights become continuous lines, crowds become fog, clouds become water.
+- **Speed ramps** are smooth (smoothstep between key speeds) and always mean something: a corridor accelerates out of real time (1× → 6×), a time-lapse ramps into a frenzy (1× → 7×), a fast shot brakes into slow motion (2.2× → 0.3× in 0.55 s) on the bar where the music stops.
+- Camera moves are added sparingly in post: slow push-ins (1.00 → 1.05–1.10 over the shot), a falling push on the opening still (1.0 → 1.3 in 0.35 s, ease-out, then accelerating to ×2 by the cut). No shakes, no whip pans, no digital zooms past the upscale limit.
+
+## 6. Camera grammar — editing
+
+**Cut-point table (the score decides; numbers at 120 BPM and 30 fps: 1 beat = 15 frames, triplet = 5, sixteenth = 3.75)**
+
+| Score texture | Cut length | Frames | Use |
+|---|---|---|---|
+| Opening hit + eighth-note arpeggio | 2 bars → 1 bar | 120 → 60 | the opening still, the first single-camera time-lapse |
+| Eighths | 2 beats → 1 beat | 30 → 15 | dawn, match cuts |
+| Triplets against eighths (3:2) | 1 beat → ½ beat | 15 → 7/8 alternating | morning, the montage of streets |
+| Sixteenths, full ensemble | grid cells light every 1–2 frames; cells re-cut every sixteenth | 1–4 | the grid |
+| Arpeggios stop, one long chord | 3–4 s, one or two shots | 90–120 | the brake: the slowed thing, the archive split |
+| Sixteenths + triplets + horns | ½ beat → 4 frames → 2 frames | 7 → 4 → 2 | dusk to night, the reverse flashback |
+| Dry cut of the whole ensemble, reverb left ringing | hard cut to black on the bar line | — | the hard stop |
+
+Rules:
+
+- **Cuts land on beats.** All cut times come from one function, `bar(n, beat)` in `tools/timeline.py`; shot lengths are never typed in seconds. The music is rendered from the same file.
+- **One long single-camera time-lapse is the clock.** It returns three or four times (night → dawn, noon clouds, a one-second nightfall, a day/night flicker), always from the same position, so the viewer knows the time without a caption.
+- **Density follows the score.** The editor never cuts faster than the current subdivision, and the slowest picture sits under the place where the arpeggios stop.
+- **Shape match cuts** on the beat: clock → clock, arch → arch, grid of streets → grid of cells. A small-to-large punch-in on the same shape is allowed.
+- **The grid** is the split screen of this style: 1×2 → 2×2 → 3×3 → 4×4, one stage per half bar or bar, the new cells lighting up in a scattered order over the previous stage within 0.3 s (never back to black), gutters 6 px black; in the last beat the cells re-cut on every sixteenth.
+- **Counterpoint split**: two halves, archive left, today right, the same kind of place or shape; the gutter grows from the centre in 8 frames.
+- **Day/night flicker**: the same camera, a noon frame and a night frame alternating every 2 frames; a pair of same-position photographs alternating every 2 frames.
+- **Reverse flashback**: in the last bar before the hard stop, earlier shots return in reverse order, 4 frames each, then 2 frames each.
+- Transitions: hard cuts only, plus the hard cut to black. No dissolves except fades from black for the page and the end card.
+
+## 7. Sound palette
+
+- **Music only.** No narrator, no source sound (the footage soundtracks are other people's music), no sound effects. A narrator would turn watching into listening and would fight the score for the structure; that is why the style has none. If a voice is ever required, one short line at the very start or end, never over the cuts.
+- Orchestration (sampled, CC0 — `core/audio/sampler.py`): pipe organ (full and soft stops, pedal), winds staccato (flute, clarinet), strings (spiccato cellos and violins, sustained cellos, violins, double bass), horn pair for the last two bars, concert bass drum on section downbeats, suspended-cymbal crescendos into the hits, glockenspiel only for numbers locking.
+- Writing: arpeggios over a one-chord-per-bar progression; additive layering (eighths → triplets over eighths → sixteenths); the *break* is all arpeggios stopping at once into one held chord (maj7 or 9th); the return is a full-organ hit; the hard stop is a dry cut of every instrument in 15 ms with the reverb tail left ringing; the coda is one soft chord that does not resolve, with the opening figure returning in quarter notes under the end card.
+- Reverb: `S.room(size .7, mix .28)` for the body, `size .78, mix .36` for the coda. Master to −14 LUFS / −1.2 dBTP with `core/render/mux.sh` (grain 0: the grain is already in the picture).
+
+## 8. Native moves (and how they are built)
+
+| Move | Code |
+|---|---|
+| **Light river** | `tools/prep.py`: integrate the speed curve to source time; each output frame averages the source frames in its window (cap `blend`). Speed 40× / blend 30 at night, 37.5× / 24 for clouds, 75× / 30 for a one-second nightfall. |
+| **Speed ramp** | `speed=[(t, v), …]` keyframes, smoothstep between them, same integrator. |
+| **Brake into slow motion** | ramp 2.2× → 0.3× with `interp=3` (ffmpeg `minterpolate` mci, bidir) for soft subjects; `mix=True` (crossfade of the two nearest frames) for thin lines such as cables, where motion compensation tears. |
+| **Crowd → tracks** (people as data) | `Trails` in `tools/render.py`: background = per-pixel median of the whole fixed-camera shot; blobs that differ by > 22/255 (scipy `ndimage.label`, area 8–900 px at half size) are people; each centroid joins the nearest one from the previous frame (< 70 px) with a 2 px line; lines fade × 0.9 per frame; the plate is darkened and desaturated as the tracks appear (0.15 → 0.55 s). Fixed cameras only. |
+| **Traffic → light** | a static or slowly drifting top-down aerial at 5× with full frame blending. |
+| **Grid** | `grid_plate()`: pool clips of 30 frames held in memory at 960×540; stages from `GRID_STAGES`; light-up order scattered; last beat re-cuts cells every sixteenth. |
+| **Counterpoint split** | `split_plate()`: archive 4:3 centre 960 px | today centre 960 px, 6 px gutter grown from the centre. |
+| **Day/night flicker** | two 15-frame clips from the same camera (noon, night), alternating every 2 frames. |
+| **Photo that moves** | a same-position photo pair alternating every 2 frames. |
+| **Number lock** | `lock_digits()`: one digit locks per sixteenth, the others roll; `counter()`: ease-out count-up; music adds one glockenspiel note per lock. |
+| **Day ring** | a thin 24-hour ring bottom-right with a sodium hand and a 3-hour arc: it runs with the film's day, **freezes and dims during the brake**, and **spins backwards during the reverse flashback**. It is a graphic, not a clock reading. |
+| **Reading light on a page** | scanned page, 16:9 window pushed 1.06 → 1.56 page widths, side light (bright top-left), paper tint (1, .93, .80), a Gaussian band (σ 95–140 px) over the quoted line growing to +38 %. |
+| **Hard stop** | picture cuts to black on the bar line; the music's dry signal is cut in 15 ms on the same frame; only reverb remains. |
+
+## 9. Pitfalls of the medium
+
+- **Watermarks and slates hidden in CC footage**: corner logos (scale from the centre until they leave frame, within 1.5×), opening thank-you cards, end-card music credits, title text over the first shots. Check a 1-second contact sheet (`tools/secsheet.py`) of every source before choosing in-points.
+- **Hidden cuts inside sources.** Hyperlapse reels hide 2–3-frame jump cuts and whip transitions; frame blending across a cut makes ghosts. Scan with `ffmpeg … select='gt(scene,0.25)',showinfo` and keep every clip's source window inside one shot.
+- **Too much blending** turns noon clouds into grey mush: about 40 % shutter (24 of 60 frames) looks better than 100 %.
+- **Grey-world balance kills night**: it washes sodium light to white. Strength 0 for night shots.
+- **Tracking needs a fixed camera.** A drifting drone flags every building as motion. Use a median background only on tripod shots; for moving aerials use blending instead.
+- **Motion-compensated slow motion tears thin lines** (cables, rails, text): use the crossfade mode there.
+- **Text over bright sky** disappears: put number cards over the dark half of the frame or add the soft darkening; never a box.
+- **A slow face is not a free subject.** Do not make an identifiable ordinary person the slowed or close-up hero. Slow a thing; show people as crowds, tracks or ghosts. Public figures and people in news events are a different case.
+- **Numbers drift.** Official pages update; quote the page and the year, keep the quote in `FACTS.md`, and recheck before the last render. Prefer one clean figure over a range.
+- **CC BY credits are long.** Each one needs author, full title and licence on the end card. Treat them as fine print (16 px, 29 px line pitch, the licence and changes on one header line) and keep the end card short (4–5 s); a 10 s card of text on black loses the audience. The film's last moving shot runs on under the card, darkened.
+- **Don't end on dead black.** After the hard stop, the coda is a page (≈1 s), then the quote laid over one last moving city shot (≈4 s, just long enough to pass `readcheck`), then the end card over the same shot.
+- Commons rate-limits (HTTP 429): search and download serially with back-off (`tools/s.sh`, `tools/f.sh`); for long 4K files download only the needed excerpt (`tools/seg.py`) or the 1080p transcode.
+
+## 10. Footage: where it comes from
+
+1. **Ask the user first** whether they have their own footage, photographs, documents or data for the place: their own time-lapses, a phone hyperlapse, family photos, a city's open data. Their material comes first.
+2. If not, search these sources (all searchable with `demo/tools/footage.py`, which records title, author, licence and page in `SOURCES.jsonl` and writes `SOURCES.md`):
+   - **Wikimedia Commons** — modern time-lapse, aerial and slow-motion footage under CC BY / CC0, photographs (for example the Carol M. Highsmith Archive, dedicated to the public domain), scans of old maps and books. Check the licence on each file page. Many videos are YouTube or Vimeo imports: use one only when its page carries a passed review (YouTubeReview / VimeoReview / LicenseReview / Flickrreview with a reviewer and a date) or the author's own source page still shows the licence. A pending review is not enough: a channel can later switch to the standard licence and claim the footage through Content ID.
+   - **Library of Congress** (loc.gov, also mirrored on Commons) — archive films published in the US before 1931 are public domain; a rights statement of "no known restrictions" alone is **not** enough: you need the publication date or another public-domain basis.
+   - **NASA** (images.nasa.gov, ISS crew photographs) — public domain; the night photos of cities from orbit make a strong opening.
+   - **Internet Archive** — scanned books and documents marked NOT_IN_COPYRIGHT with a pre-1931 publication date; take the line you quote from the scan, not from a modern edition.
+   - **National archives** (US National Archives, other governments' open archives) — check each country's terms.
+   - Licences allowed: public domain, CC0, CC BY. Not allowed: CC BY-SA, "no known restrictions" without a basis, YouTube mirrors, unknown sources, restored or colourised archive copies.
+3. **Before downloading, tell the user** what you will download, from where, under which licence, and how it will be credited.
+
+Data for number cards comes from official pages (census, transport department, the place's own statistics); write the claim, URL and quoted wording into `FACTS.md`.
+
+Recipe for another place: one single-camera time-lapse of ≥ 12 hours (the clock); 10–15 ground time-lapses and hyperlapses (streets, stations, bridges); 2–4 aerials, one of them top-down at night; one slow-motion shot of a fragile thing; one fixed-camera crowd shot for tracks; one archive film of the same kind of place; one night photograph from orbit or above; a same-position photo pair; three official numbers; one public-domain sentence written about the place.
+
+## 11. Variation space
+
+- Place: any city, port, airport, market, campus, factory, a river over a year (the clock becomes a season).
+- Length: 30–90 s; one brake at least; one hard stop always.
+- Score: organ/winds/strings is the default; a choir, a piano ensemble or analogue synth arpeggios also work if the writing stays additive and the break stays a sudden held chord.
+- Tempo 96–132 BPM; at 30 fps prefer tempos whose beat is a whole number of frames (120: 15, 100: 18, 112.5: 16).
+- Grid sizes and the counterpoint pair change with the material (archive vs today, day vs night, two cities).
+- Number cards: 2–4 at most; the day ring may be dropped for non-daily structures.
+- Fixed: footage only, the frame-blended light, cuts on the beat, the brake, few words, one quote, the hard stop, the release-print grade.
+
+See [`DEMO.md`](https://github.com/lemomo-ai/lemo-opuscar/blob/413a6bd6/styles/city-symphony/DEMO.md) for the example film.

package/src/skills-lib/wizz-video-styles/references/film-styles/countdown.md

@@ -0,0 +1,132 @@
+<!-- Vendored from lemo-opuscar (https://github.com/lemomo-ai/lemo-opuscar/tree/413a6bd6/styles/countdown), MIT, (c) 2026 LemoLab; see ../../LICENSE-lemo-opuscar.txt. Paths under demo/ and core/ name that repo's engine, which is not shipped here: build the look in HyperFrames (../film-styles.md). -->
+
+# Live Countdown Split-screen — Style Prompt
+
+> An edit style, not a look. Real footage of one event with a known zero (a launch, a verdict, a final whistle, polls closing), shown in real time: screen time equals event time, a clock ticks every second, and the screen splits into more and more cells as the moment nears, every cell showing the same second. At zero the cells merge into one picture and the clock flips.
+> References (grammar only): the real-time clock and splitting screens of the TV series *24*; the multi-camera archive panels and mission-time codes of the documentary *Apollo 11* (2019). Never copy their type, colours, sounds, graphics or shots, and never name them in the film.
+
+## 1. Essence, and what it is not
+
+1. **Real time.** One continuous stretch of the event, never cut, never re-ordered, never sped up. Slow-motion sources are played back at real speed. The clock and the picture can be checked against each other at any frame.
+2. **The clock is the lead.** It ticks every whole second (sound and a falling digit), it says who is in control, and it changes state at zero.
+3. **Cells show the same second.** Every video cell is synced to the clock, or says plainly that it is not (§8, "sync badge").
+4. **The layout is the edit.** Cuts are layout changes, and every layout change lands on a real spoken call or event in the native sound, not on the music.
+5. **More cells, closer to zero; one picture at zero.** The number of cells is the tension curve.
+
+Not a montage (time jumps), not a security wall (the layout never moves), not a flashback structure (real time cannot go back), not a slide deck (cells that don't follow events, text filling the cells, centred titles with page turns).
+
+## 2. Materials & rendering
+
+- **Three kinds of material at once**: video cameras of the same moment (2–8), documents (a fact sheet, a filing, a schedule: shown as paper, never retyped as a slide), and photographs (the people of the day). Our own type explains them; type printed in the material is only picture.
+- **Data drawn by us**: checklists, bars, a flight or process profile, a second countdown to the next milestone. Every number comes from an official source (§10).
+- **Cells**: hard rectangles on near-black `#06080A`, 24 px outer margin, 8–9 px gutters, 1 px border `rgba(222,232,238,.16)`, a light radial vignette (to 36 % black) on video cells. No rounded corners, no drop shadows, no 3D.
+- **Camera chip** top-left of every cell: dark plate 84 %, a square dot (orange = synced video, grey = unsynced or still), code (`CAM 4`, `DOC`, `LIST`), Chinese name, small English name.
+- **Burned-in text in the sources** (broadcast bugs, camera IDs, timers, tracker readouts) is cropped by a per-shot usable region, or blurred out (ffmpeg `delogo`) after it has been used for syncing. Never shown as if it were ours.
+- **One grade for every source** (ffmpeg): `eq=contrast=1.07:saturation=0.9:gamma=0.97:brightness=-0.01,colorbalance=rs=-0.02:bs=0.03:rh=0.03:bh=-0.02`, then film grain 1 at the master (`mux.sh … 1`). Cool shadows, slightly warm highlights: control rooms and launch pads look like one film.
+- **Resolution**: modern footage at most 1.5× its source pixels; for a punch-in, extract that stretch at a higher resolution from the 4K source instead of enlarging a 1080p frame. Documents at ≥ 1 source pixel per screen pixel.
+
+## 3. Colour logic
+
+| Role | Value |
+|---|---|
+| Ground | `#06080A`; panels `#0B0E11` |
+| Main text | `#F2F5F7`; secondary `#7C878D` |
+| Accent: the T− sign, synced dots, ticks done, event flags, the active control segment | `#FF5B1F` |
+| Speaking light, "heard" tags | `#FF3B30` |
+| Plates over full-screen footage | `rgba(6,8,10,.80)` |
+| After zero, the clock block | ground `#FF5B1F`, text `#0A0B0C` (inverted) |
+
+Orange means "now / synced / done". Grey means "pending / not synced". Nothing else is coloured; colour belongs to the footage.
+
+## 4. Type & subtitles
+
+- Clock digits: a condensed sans with fixed-width digit slots (ours: Barlow Condensed SemiBold, drawn digit by digit in equal slots so the clock never jitters); size up to 132 px, the T−/T+ sign at 0.85×, tenths at 0.41× in grey, a ten-segment sub-second bar under it.
+- Chinese: a sturdy sans (ours: Noto Sans SC 500/700, subset). English text: a clean grotesque (Barlow). Labels, codes and English small print: a monospace (IBM Plex Mono). All OFL, in the project's `fonts/`.
+- **The live transcript is the subtitle.** A fixed box (bottom right, or a cell in the grid): a speaking light and the speaker type (commentary / loop / narrator — the narrator, our own voice, is marked in orange so it is never mistaken for the event's audio), the English line appearing word by word at the spoken times (0.08 s fade per word, a block cursor), the Chinese line whole when its cue starts. **One line per language, never wrapped**: a long sentence becomes two cues; each cue ≥ max(1.8 s, speech + 0.6 s). Transcribe only what is said; mark doubtful words in `FACTS.md`.
+- Reading time for everything else: CJK ÷ 4.5 + other ÷ 15 + 1.5 s from the moment it is complete (`readcheck.mjs`). A comparison (two bars, two numbers) stays fully on screen ≥ 2.5 s.
+
+## 5. Motion quality
+
+- 30 fps. Clock: on each whole second the new digits drop 14 px in 0.12 s (ease-out); the sub-second bar empties one segment per 0.1 s (fills after zero).
+- Re-layout: 0.34 s `easeInOutCubic`. Cells that stay interpolate their rectangles; new cells open from their left edge; leaving cells close to the left; a new cell flashes a 3 px orange border for ~0.2 s when it lands. A grid filling up pops one cell per spoken count in 0.30 s.
+- Swaps (one source replaces another in the same rectangle) wipe in from the left. Hard cuts only between full-screen sources.
+- Photographs push in ≤ 6 % over their whole stay; names pop one by one 0.16 s apart. Documents move their window down a row in 0.45 s; the highlight sweeps left to right in 0.35 s.
+- Never: slow push-ins on video to fake motion, page-turn transitions, centred pop-up words, speed ramps.
+
+## 6. Edit grammar
+
+- **Screen time = event time.** One mapping, `M = t − T0`, for the whole film. Each camera has its own source offset (`s = M + c`) measured once and baked into the frame extraction (`demo/tools/prep.sh`): frames are named by film frame number, so the page never computes sync.
+- **Cut points are calls.** Each layout change starts on the first word of a real call in the native audio ("ten", "go for…", "ignition", a name) or on an event visible in the footage. The music follows; it never sets a cut.
+- **Cell count rises toward zero**: 1 → 2 → 3–4 → many (6–9, the clock itself becomes the centre cell, the transcript a cell) → **one** at zero → a few again after zero (the event's consequences) → one under the title. The shortest layouts are nearest to zero.
+- **Inside a cell, the source cuts itself** (a broadcast switching cameras): keep it; we only cut the layout.
+- **Silence before zero**: the music drops out a quarter-second before zero; only the native sound and the clock remain.
+- Full-screen moments carry the clock and transcript as plates at the bottom; a gradient (0 → 75 % black) under them.
+
+## 7. Sound palette
+
+Layers, top to bottom:
+
+1. **Native sound** — the event's own audio (commentary, radio loops, crowd), high-pass 90 Hz, compressed, always on top; never ducked.
+2. **Narrator** (sparingly, only in gaps of the native sound, never over it): clean and close, a different gender or accent from the event's voices; Kokoro, every line through `asr_check.py`.
+3. **The clock**: a tick every whole second (two alternating pitches around 1.5–1.9 kHz, panned ±15 %), louder in the last ten seconds, a soft single pitch after zero.
+4. **Interface foley** from `window.EV`: re-layout whoosh + metallic snap; cell pop snap; checklist tock; event-flag whoosh; full-screen cut thump.
+5. **Score**, 60 BPM so one beat = one second = one tick: a low synth drone whose low-pass opens toward zero, a bass drum per beat, off-beat hats, a tom per spoken count, timpani roll into zero, silence, then a low brass major chord with bass drum, crash and tubular bell at zero; after zero a new texture (pizzicato bass ostinato, lighter tick). Low brass, timpani, toms, tubular bells, glockenspiel; no piano-and-strings pad.
+- Mix: music about 8–10 dB under the native sound and ducked under speech; the ignition (or the event's own climax sound) is the loudest moment; master −14 LUFS via a pre-limited mix so `mux.sh` meets the true-peak target.
+- **Changing the music** (e.g. a Suno track): never move the clock or the cuts; choose or stretch a track so a downbeat lands exactly on zero and a section change lands on T−10, and keep the quarter-second silence before zero.
+
+## 8. Native moves (and how to build them)
+
+- **Many into one.** At zero, every cell moves to its tile of the full frame (a 3×3 grid becomes nine 640×360 tiles) while each tile cross-fades to the same full-screen shot, chips and borders vanish in the first half, a warm white flash (`rgba(255,248,236,.7)`, decay τ ≈ 0.11 s) peaks, and the clock and transcript slide from their grid cells back to the bottom. 0.42 s. *Fits:* a verdict read out; the goal at the final whistle.
+- **Clock inversion.** At zero the clock block flips from dark with an orange T− to orange with dark T+, and its header changes (countdown → elapsed). *Fits:* polls closing, a store opening.
+- **Control strip.** Segments in the clock header (who has control: ground → vehicle → mission control; prosecution → jury → judge); the active one fills orange on the call that hands it over, with a short flash. The story's turns live here.
+- **The grid fills on the count.** Empty grid slots show as dashed "standby" frames; each spoken number opens one camera. *Fits:* "three, two, one" before a premiere.
+- **Sync badge.** Top-right of each video cell: `● 实时同步 LIVE SYNC` (synced), `○ 同一时段 · 未逐帧同步 SAME WINDOW` (hatched, grey: footage from the same minutes without a usable clock), `慢动作还原 1× REAL-TIME` (slow-motion re-timed). Honesty is part of the look.
+- **Planned vs heard checklist.** Official milestones with their planned times tick as the clock passes them; when the native audio actually says one, a red "heard −32.1" tag with a small waveform appears beside it. *Fits:* a court schedule against what was said; a train timetable against the platform announcement.
+- **Second countdown.** After zero, a smaller clock counts to the next official milestone (`−1:47 to booster separation, expected T+2:09`) along a schematic profile drawn by time, not distance. *Fits:* the next race split; the next ballot count.
+- **Event flag.** An orange bar with black type slides in at the bottom of the relevant cell on its call (`RS-25 ×4 启动 ENGINE START · T−6.36`), held ≥ the reading time.
+- **Punch-in on a call.** A hard digital punch-in on the relevant detail when the commentary names it, from a higher-resolution extraction of the same source.
+- **Document by rows.** A paper close-up shows two or three rows; a highlight bar and an annotation band below the paper translate the row; the window moves only between rows (its edges in the white space).
+
+## 9. Pitfalls of the medium
+
+- **Unsyncable footage** (an edited "rover" reel with no clock or event to align on). Never fake it: label the cell "same window · not frame-synced", keep it out of the countdown's key seconds or use it only where nothing it shows contradicts the time, and say so in `FACTS.md`. Look first for a source with a burned-in timer, a visible ignition / flash / whistle, or the same audio.
+- **Measuring zero.** Never trust metadata. Combine two independent marks: a visible event (ignition flash, a broadcast countdown bug's flip) and the spoken word timing ("booster" at zero). Re-measure on the file you actually cut: `ffmpeg -ss … -c copy` starts on the previous keyframe.
+- **Audio and picture of one file can disagree** (a switched feed whose audio lagged its video by ~1 s). Sync each camera on its picture.
+- **Slow-motion sources**: measure the factor from two events with a known gap (e.g. engine start to booster ignition = 6.36 s) and re-time with `setpts=PTS/k`.
+- **Burned-in text**: broadcast bugs, camera numbers, timers, tracker data. Define per-shot usable regions (they change when the source cuts) or `delogo` them; check every source cut with a frame-difference scan.
+- **A static wide shot for ten seconds** reads as flat. Give every 1–2 seconds a new element (a cell, a tick on the list, a flag, a punch-in), not a fake camera move.
+- **Spoilers in the subtitle**: a whole-sentence Chinese line reveals the end of a count early. Split counts into separate cues.
+- **The clock jitters** if digits are proportional: draw them in fixed slots.
+- **Hot peaks** from broadcast audio and the ignition hit miss the true-peak target: pre-limit the mix at about −4 dBFS after normalising to −14 LUFS.
+- **Logos.** No agency insignia as a graphic; cut slates and bugs; the film says it is unofficial.
+
+## 10. Materials, sources, facts, engine
+
+**Materials — ask first.** Before searching, ask the user whether they have their own material of the event: footage from several angles, recordings of the audio, photos, documents (schedules, filings, results), data. What they give comes first and is credited as they say.
+
+**If not, where to look** (each item's own page decides; record it):
+- **NASA** (images.nasa.gov, svs.gsfc.nasa.gov): U.S. federal government works, public domain in the U.S.; fine to use. Don't use NASA insignia as graphics or imply endorsement; check each item for third-party material.
+- **Other U.S. federal agencies** (NOAA, National Archives, DoD DVIDS): public domain, same check for third-party material.
+- **Internet Archive**: only items whose own record shows public domain (pre-1931 U.S. publications, U.S. government works) or CC0 / CC BY; skip `youtube-*` / "mirrortube" uploads and anything of unclear origin.
+- **Wikimedia Commons**: read each file page; CC0, public domain and CC BY are fine (credit CC BY in full); CC BY-SA and "no known restrictions" are not fine without the user's explicit agreement.
+- **Library of Congress**: check each item's rights statement; "no known restrictions" is not CC0.
+- Not usable: re-uploads without a source, restored or colourised versions, news-agency footage, anything whose licence is not stated.
+
+**Before downloading, tell the user** what you will fetch, from where and under which licence, in one message. Then fetch only the seconds you need: `python3 demo/tools/grab.py <nasa_id> <orig|large> <start> <duration> <name> [--copy]` reads an HTTP range of a NASA original; `python3 demo/tools/footage.py search|fetch|used|md` searches NASA, Commons and the Internet Archive and records licence, author and page in `SOURCES.jsonl` / `SOURCES.md` automatically. Keep `raw/` under about 1.5 GB.
+
+**Facts**: a `FACTS.md` (claim, URL, quoted wording, or the measurement) covering every number, time, name and call on screen and in the narration; official sources only; planned times and actual (heard) times are kept apart. Check the script against it before the last render.
+
+**Engine** (copy into the project): `demo/index.html` (Canvas 2D: `LAY` layout keyframes with `grid` / `merge` / `cut` / `swap`, `camRegion()` usable regions, `crop()` cover framing, `drawClock()`, `drawTx()`, `drawCheck()`, `drawDoc()`, `drawProfile()`, `drawThrust()`, `window.EV` for the mix, `window.TEXTS` for `readcheck`, `window.SUBFIT` for the one-line subtitle check), `demo/tools/prep.sh` (the sync table → frames), `demo/tools/subs.mjs` (cues → `audio/cues.json`, one-line and duration check), `demo/audio/mix.py` (native sound, ticks, foley, score, ducking, pre-master), `demo/tools/final_asr.py` (Whisper on the finished film), `demo/tools/fetch_fonts.py`.
+
+## 11. Variation space
+
+The agent decides per film: the event, where real time starts and ends, the number of cameras, which documents and data, the control segments, the narrator's lines (or none), the closing layout.
+
+- **Clocks**: T−/T+; a match clock; a wall clock (HH:MM:SS) for a news day; a countdown to a deadline that then counts up as "overtime".
+- **Three structures**: rising cells to a merge (this demo); one cell that splits outward after zero (results arriving from many places); two parallel countdowns side by side that meet at zero (two teams, two cities).
+- **Three openings**: a single wide shot and the first call; the clock alone on black, then the first camera opens; a document (the schedule) with the first item ticking.
+- **Three endings**: the inverted clock over one full-screen shot; the grid refilling with consequences, then one title; the second countdown reaching its own zero.
+
+---
+
+How our demo was made (story, layouts, score, end card, build): [DEMO.md](https://github.com/lemomo-ai/lemo-opuscar/blob/413a6bd6/styles/countdown/DEMO.md). Read it after your treatment exists.