AudioLab.tools — authority platform for audio + AI

A content and tools platform at the intersection of audio, AI, development, Android, DSP, accessibility and live workflows. No marketing funnel with "join the waitlist" buttons behind every page. Each cluster ships a working browser-side demo plus deep docs.

Hand-drawn sketch: AUDIOLAB.TOOLS at the top, three clusters MIXLAB / HEARLAB / ENGINEERING, each with TOOLS / DOCS / INSIGHTS pillars, caption 'real tools · deep docs · no slop'. — Early whiteboard sketch · since grown into six labs

Three principles (from the site itself)

Tools you can actually use. Each cluster ships a working, browser-side demo. Nothing hidden behind a waitlist.
People-first writing. No keyword-stuffed AI slop. Every page is written by someone who has shipped audio in production and is willing to be wrong in public.
Built where it matters. Android, realtime DSP, accessibility, developer ergonomics. The places where audio AI actually has to work, not just marketing demos.

Six labs — honest about what works

The site revolves around six browser-side labs. Each lab is honestly labelled by what it actually does — working, in progress, or a simulated concept:

Lab	Status	What it does
MixLab	Beta	Loudness/dynamics/stereo — verified to ±0.08 LU against the EBU R128 reference
VoiceLab	Beta	Voice QA: clarity, pacing, filler density, clipping
SignalLab	Beta	Audio indexer: silence/clipping detection, region tags
Ψ Lab	Building	Phase vocoder · time-stretch, freeze, reverse, smear (fully client-side AudioWorklet)
SkillLab	Building	Practice challenges with feedback (works, being expanded)
HearLab	Building	Non-medical hearing companion (dashboard separate as a concept)
CueLab	Simulated	Live-routing concept with simulated data

Everything client-side: audio never leaves the browser. Alongside that, docs, methodology, glossary, roadmap — indexable content, no JS-only routes.

Stack

Framework

Astro 6 (content-first, static with React islands where needed)

React 19, Tailwind v4 (CSS-first config)

Content

MDX (@astrojs/mdx), per-cluster content collections

OG cards

Satori + Resvg.js — dynamically generated at build time

Fonts

Geist + JetBrains Mono (variable)

Performance

Sharp for the image pipeline, build-time pre-rendering

RSS

Feed.js per cluster, per tag

Sitemap

@astrojs/sitemap

Astro was chosen over Next.js for this use case: content-first, lots of static pages, React islands for interactive demo tools only where it's truly needed. No 100% client-side rendering for a glossary.

What's in it

Hero with motion fallback (full-bleed video + static image fallback for reduced-motion)
Per-cluster landing pages with their own tools, docs and insights stream
Glossary, methodology, roadmap, changelog as public artefacts
Showcases (case studies) and Labs (experimentation space)
API page (internal tools others can integrate)
Pricing + contact + privacy

Status

Live at audiolab.tools. Six labs are running; four do real analysis (Beta), one is being expanded, one is a simulated concept — and it says so right there. MixLab's loudness is verified against the EBU R128 reference (±0.08 LU vs ffmpeg). The AI interpretation layer is the next step, not a current claim.