Nobel Chemistry, in English and Spanish

Videos (1920x1080, 30 fps, narrated, captions burned in):

• English: final.github.mp4 · 149.6 s · 9.7 MB · -14.0 LUFS / -1.6 dBTP
• Spanish: final-es.github.mp4 · 154.2 s · 9.6 MB · -14.0 LUFS / -1.6 dBTP

The masters, final.mp4 (88 MB) and final-es.mp4 (93 MB), are the assets 26-nobel-chemistry-en-es--final.mp4 and 26-nobel-chemistry-en-es--final-es.mp4 of the media release.

The requests#

"Full Chemistry 2026 explainer (~2:10, 16:9) for the nobel-2026-lab: Frank's recipe, Kagan's bend, Soai's copier, the 2003 staircase, the coin flip, our mirror-race toy, caveats, medicines. NOT socratic: no questions to the viewer. American voice (af_bella; not the British female). Paperback lab-notebook look, CC0 bed, showtime credited (corner mark, end card, closing line). Must look great; post soon."

"Spanish version of the full Chemistry video for LinkedIn: translate narration and every on-screen string, Spanish voice, captions; same numbers and claims"

Mode: quick, quality review. Made on 7 October 2026 with showtime 0.4.0 on a 6-core Intel Mac, from the public lab repository nobel-2026-lab (MIT), whose mirror-race toy model gives the simulation shots.

What it shows#

• A 2:30 explainer from a storyboard table. Twelve shots (project/storyboard.md, 145 s planned) became the project with showtime new dom <dir> --from-storyboard; the film runs on a persistent roadmap, Frank's three-box recipe from 1953, which each laureate's shot ticks (Kagan 1986, box two; Soai 1995, box three), then the 2003 staircase, a coin flip, the lab's mirror-race toy model, what it does not show, and why medicines care.
• A light look signature. Paperback: warm paper, a burnt-orange accent, Fraunces with IBM Plex Sans, calm motion; one hand is always solid orange, its mirror hatched teal.
• Honest labels on screen. Illustrations say so ("shape only, not Kagan's data", "tints illustrative", "our toy model · trend only", "our drawing"); the quotes are attributed to the Nobel Committee.
• Localisation of the whole film. The Spanish version translates the narration (project-es/narration.es.md) and every on-screen string, with a Spanish voice (Kokoro ef_dora) and Spanish captions, the same numbers and claims, and every scene retimed to the Spanish voice (154.2 s against 149.6 s).

Commands, in order#

showtime is skills/showtime/bin/showtime; reconstructed from the jobs' ledgers and the projects' files.

showtime job init chem-full --platform x --request "Full Chemistry 2026 explainer (~2:10, 16:9) ..."
showtime new dom <job>/project --from-storyboard storyboard-full.md --look paperback   # 12 scenes, narration.md
#   the shots built by hand: scenes/s-shot-*.js and .css, assets/ (the toy model's runs and histograms)
showtime voice script <job>/project/narration.md -o <job>/project/voice                # Kokoro af_bella, x1.08
showtime retime <job>/project --from-voice <job>/project/voice/timeline.json --total 145
showtime check <job>/project
showtime render <job>/project --job <job>                      # final.mp4; spans spliced: final-2 ... final-4
showtime qa <job>
showtime review-pack <job> --against <the earlier final>       # round 1: pairwise, two critics, both orders
showtime review-pack <job>                                     # round 2
showtime review-respond <job> --fixed r1o1-B1 "..."            # every finding fixed or waived
showtime deliver exports <job> --targets github,x,linkedin

showtime job init chem-full-es --platform linkedin --request "Spanish version of the full Chemistry video ..."
#   project-es: a copy of the project with every string translated, narration.es.md
showtime voice script <job>/project/narration.es.md -o <job>/project/voice              # Kokoro ef_dora, x1.0-1.08
showtime retime <job>/project --from-voice <job>/project/voice/timeline.json
showtime render <job>/project --job <job>
showtime qa <job>
showtime review-pack <job>
showtime deliver exports <job> --targets github,linkedin

Timings#

On the 6-core Intel Mac:

QA summary#

• English, at delivery (final-4): PASS (0 fail, 0 warn), -14.0 LUFS, -1.6 dBTP. The job's platform was X, whose standard accounts take 140 s; the film is 2:30, so qa's too_long was waived as the owner's call (he approves before posting). Today (0.4.1, the same file) qa still FAILs too_long against X; loudness, true peak, frame 0, black and frozen stretches pass.
• Spanish (final-3, LinkedIn): PASS (0 fail, 0 warn), -14.0 LUFS, -1.6 dBTP, at delivery and again today.
• The hearing pass (English round 2, audio.txt): every voice line 16.8-18.0 dB over the bed, 129-182 words a minute; the one pause over 2 s (115.1-117.7 s) has the coin-flip result on screen.

What the critic found#

English, round 1, pairwise (final.mp4 against final-2.mp4, two critics, one per order, neither told which was newer: review/en-round-1/): both preferred final-2, but would not post it: 3 blockers (a hidden quote, and the length twice) and 8 should-fix. Fixed in the next render (RESPONSE):

English, round 2 (FINDINGS): ship, would post: yes; every round-1 fix confirmed; polish only.

Spanish, round 1 (review/es-round-1/): ship after fixes, no blockers, no English left on screen, every number matching the English plan; 4 should-fix, all from the localisation, all fixed in final-3: the captions split spelled-out years across chunks ("En mil novecientos cincuenta" / "y tres"), so numbers are now one caption word in digits; an ungrammatical counter label ("5,049 veces / ganó la mano espejo"); and two lines spoken too fast (243 and 211 words a minute), voiced again at x1.0.

Files#

• final.github.mp4, final-es.github.mp4: the copies under 10 MB; final.mp4, final-es.mp4: the masters, release assets; poster.jpg, poster-es.jpg: frame 0 of each.
• project/ (English) and project-es/ (Spanish): storyboard.md and storyboard.json, the narration (narration.md; narration.es.md in Spanish), showtime.json, index.html, scenes/ (one script and style per shot), assets/ (the toy model's runs and histograms), audio/mix.json, voice/ (timeline, word times, captions; the WAVs are not committed). The English showtime.json title was the default "Project"; it now carries the film's title for exports.
• review/: English round 1 (pairwise: VERDICT.md, each order's FINDINGS.md), English round 2 (FINDINGS.md, RESPONSE.md, audio.txt), Spanish round 1 (FINDINGS.md, RESPONSE.md, audio.txt).
• receipt.md, receipt-es.md: what each job took, from what showtime recorded; credits.txt.

To rebuild: showtime voice script project/narration.md -o project/voice (and project-es/narration.es.md), then showtime render project and showtime render project-es.

Sources and license#

• The films: made with showtime for nobel-2026-lab (github.com/FavioVazquez/nobel-2026-lab, MIT): educational demos of the 2026 Nobel Prizes, not research. Facts from nobelprize.org (the press release and the popular and scientific backgrounds); the simulation shots are the lab's own toy model.
• Molecule: L-alanine from PubChem CID 5950 3D coordinates (NCBI, public domain).
• Music: "I'm glad you are here with me" by Loyalty Freak Music (CC0), from the album "LOFI AMBIENT SONGS !".
• Voices: Kokoro (Apache-2.0), voices af_bella (English) and ef_dora (Spanish), made on the rendering machine.
• Fonts: Fraunces and IBM Plex Sans (the Paperback look) and IBM Plex Mono, SIL OFL 1.1, from the fonts showtime installs.

The project files on GitHub