Museum Time Machine

Museum Time Machine is a full-screen art and poetry time portal. Pick a year from 1000 to 2000 and the page shifts into that era with museum artwork, period-matched poetry, ambient audio, historical context, an art inspection lens, poem reading mode, and a random walk through time.

Features

• Full-screen era artwork from The Met, Art Institute Chicago, Harvard Art Museums, and curated fallbacks.
• Fast artwork previews, contained background framing, and a detail placard for inspecting the selected work.
• Period poetry from PoetryDB for later eras, plus curated named medieval poets for the 1000s through 1400s.
• Per-session author history so a visitor does not keep seeing the same poet.
• Timeline controls, keyboard navigation, random walk mode, and responsive layouts for desktop, tablet, and mobile.
• Ambient radio lookup through Radio Browser, with graceful silence if no stream is available.
• Era-matched typography, page aging, dust particles, vignette, speech synthesis reading mode, and a museum-themed favicon.

Poetry Coverage

The app uses PoetryDB where it has reliable public-domain coverage. The 1000s through 1400s use a curated corpus because PoetryDB has very thin medieval coverage.

Years	Source Strategy
1000s-1400s	Curated named medieval and early Renaissance poets
1500s	PoetryDB author pool
1600s	PoetryDB author pool
1700s	PoetryDB author pool
1800s	PoetryDB author pool
1900s-2000	PoetryDB public-domain early modern pool

The current author and curated poem configuration lives in src/data/eraConfig.ts and src/lib/timeMachine.ts.

Voice Narration

Poem reading currently uses the browser Web Speech API: window.speechSynthesis with SpeechSynthesisUtterance. The app creates one utterance per poem line, prefers an en-GB voice when the browser provides one, falls back to any English voice, and highlights each line as it is spoken.

Current narration settings:

Setting	Value
Voice	Browser en-GB voice if available, otherwise any English voice
Rate	0.86
Pitch	0.92
Volume	0.9
Reading behavior	Line-by-line with active-line highlighting

Possible upgrades:

• Add an in-app voice picker from the browser's installed voices.
• Add era presets for rate, pitch, accent preference, and reading style.
• Add sliders for rate, pitch, and volume.
• Use true pause and resume with speechSynthesis.pause() and speechSynthesis.resume().
• Add optional cloud TTS for consistent cinematic narration across devices.

APIs

• /api/artwork tries The Met, Art Institute Chicago, Harvard Art Museums, then curated artwork fallbacks.
• /api/poem serves curated early poetry first for 1000-1499, otherwise fetches era-matched PoetryDB authors and falls back to curated poems.
• /api/audio searches Radio Browser for an HTTPS station that fits the active era.

Development

npm install
npm run dev

Open the local URL printed by Next.js. The usual local URL is:

http://127.0.0.1:3000

Verification

npm test
npm run typecheck
npm run lint
npm audit --omit=dev
npm run build
npm run e2e

Deploy

The app is Vercel-ready as a standard Next.js App Router project:

npx vercel deploy --prod --yes

Production URL:

https://museum-time-machine.vercel.app