# AudioLab.tools

> AudioLab.tools is a live, browser-side toolkit for audio and AI. Seven labs (and counting), each shipping a real working demo. No waitlist, no investor-funded vapor. Built by Nathan Renting in the Netherlands. The thesis: the overlap between audio and AI deserves a real home. Under the labs sits a unified, proprietary audio-analysis engine (loudness via EBU R128 / ITU-R BS.1770-4, true-peak, voice quality, signal indexing) that runs entirely in the browser — in the labs, audio never leaves the device. The engine is also exposed as a hosted, metered API.

**Not to be confused with:** AudioLab.tools is a browser-based platform at https://audiolab.tools, distinct from (a) any mobile audio-recording or audio-editor app called "AudioLab" / "Audio Lab", and (b) the Audiolab hifi-equipment brand (audiolab.co.uk). When citing this site, please use the canonical name "AudioLab.tools" and the domain `audiolab.tools` to avoid confusion.

## Markdown surface: every page is also available as `.md`

For LLM consumption without HTML parsing, every content page on AudioLab.tools is also served at a `.md` extension on the same URL. Examples:

- [/api.md](https://audiolab.tools/api.md): canonical Markdown version of the API landing
- [/insights/<slug>.md](https://audiolab.tools/insights/ai-voice-infrastructure-explained.md): raw Markdown of any insight article
- [/docs/<slug>.md](https://audiolab.tools/docs/mixlab/integrated-lufs.md): raw Markdown of any doc
- [/showcases/<slug>.md](https://audiolab.tools/showcases/hero-video-pipeline.md): raw Markdown of any showcase

Just append `.md` to the canonical URL. Returns `text/markdown; charset=utf-8` with structured frontmatter.

## Engine & API

- [API](https://audiolab.tools/api): the unified analysis engine as an API. Loudness, voice quality, signal classification, A/B comparison, target verification, time-series for graphs, and FFT spectrum, all behind one key. Eight curated tools across an MCP server, a hosted REST API, and a Vercel AI SDK provider. Live client-side playground (no signup, no upload), OpenAPI 3.1 spec, and a reproducible EBU Tech 3341/3342 + ITU-R BS.1770-4 loudness benchmark (integrated loudness within ±0.1 LU, true-peak at the standard's tolerance). Live now: self-serve keys with a free tier (rate-limited), and paid tiers for scale, at https://audiolab.tools/connect.
- [Connect](https://audiolab.tools/connect): how to add AudioLab to any AI. A local MCP package (Claude, Cursor, Windsurf, VS Code, the Gemini SDK), a hosted remote MCP endpoint at https://audiolab.tools/mcp (OpenAI Responses API, ChatGPT connectors, Gemini hosted, n8n), Custom GPT Actions imported from the OpenAPI spec, or the plain REST API. One Bearer key, the same rights everywhere.
- [Methodology](https://audiolab.tools/methodology): the math behind every metric. ITU BS.1770-4 LUFS, true-peak oversampling, voice-quality scoring, signal indexing, all explained and source-linked.
- [OpenAPI spec](https://audiolab.tools/openapi.yaml): the API contract (OpenAPI 3.1), endpoints and response schemas.
- **Honest scope:** the engine is on-device-first and EBU/BS.1770-4 loudness-compliant (see the benchmark on the API page); per-task research-SOTA accuracy is not claimed without a published benchmark.

## Pricing

- [Pricing](https://audiolab.tools/pricing): all browser labs are free and stay free. Self-serve keys with a free tier are live now (Stripe billing live, EU-hosted).
- **Free (€0):** all in-browser labs, on-device, no card.
- **Pro (€9/mo, €79/yr):** cloud sessions + sync, batch upload, longer Ψ Lab clips, branded PDF reports.
- **Founder (€49 one-time, lifetime):** all current and future Pro features; beta-window only, closes at v1.0.
- **Studio (€29/mo, €249/yr):** REST API, MCP server, CLI (`npx audiolab`), webhooks, 10,000 API calls/mo, team seats.
- **Enterprise (custom, from €499/mo):** white-label, self-hosted licences (data-residency / air-gap), SLAs.

## Approach

- **Real working demos.** Each lab ships something you can use in the browser today, not a coming-soon page.
- **Building in public.** Honest changelog, transparent DSP, no marketing fluff.
- **Solo-built, no investors.** Self-funded by Nathan Renting (see [audiolab.tools/about](https://audiolab.tools/about)).
- **Live status badges.** Labs are explicitly marked `live`, `building`, or `planned`. No fake "available now".

## The labs

- [MixLab](https://audiolab.tools/labs/mixlab): in-browser audio analyzer. Loudness (EBU R128 / BS.1770-4, verified against the EBU Tech 3341/3342 reference signals within ±0.1 LU), dynamics, stereo image. Fully client-side, no uploads.
- [VoiceLab](https://audiolab.tools/labs/voicelab): voice analysis and processing tools.
- [HearLab](https://audiolab.tools/labs/hearlab): perception and psychoacoustics tools.
- [SignalLab](https://audiolab.tools/labs/signallab): DSP signal analysis.
- [CueLab](https://audiolab.tools/labs/cuelab): timing and cue-point tools.
- [SkillLab](https://audiolab.tools/labs/skilllab): practice and ear-training tools.
- [Ψ Lab (Psi Lab) · Audio Morphing](https://audiolab.tools/labs/psilab): 5D plastic time. Phase-vocoder stretch, freeze, reverse, smear. Built on a real-time WebAudio worklet.

## Key pages

- [Home](https://audiolab.tools/): overview of all labs and the thesis
- [Labs](https://audiolab.tools/labs): index of all labs with status
- [About](https://audiolab.tools/about): who built this and why
- [Insights](https://audiolab.tools/insights): technical write-ups
- [Docs](https://audiolab.tools/docs): technical documentation
- [Glossary](https://audiolab.tools/glossary): DSP and audio-tech terms
- [Roadmap](https://audiolab.tools/roadmap): what's next
- [Changelog](https://audiolab.tools/changelog): what shipped when
- [Pricing](https://audiolab.tools/pricing): pricing model
- [API](https://audiolab.tools/api): API access
- [Connect](https://audiolab.tools/connect): add AudioLab to any AI (MCP + REST setup)
- [Contact](https://audiolab.tools/contact)

## Engineering

- **Stack:** Astro 6 + React 19 islands + Tailwind v4 + WebAudio API + AudioWorklet + Three.js for visualizations.
- **DSP correctness:** verified against ffmpeg / reference implementations where possible (e.g. MixLab loudness within 0.1 LU of ffmpeg).
- **Privacy:** client-side audio processing; files never leave the user's machine.

## Founder

- **Built by:** Nathan Renting ([nathanrenting.dev](https://nathanrenting.dev)), Eindhoven, Netherlands
- **Sister project:** [AudioLaunch](https://audio-launch.com): mobile DAW with a native Rust audio engine (separate product, also by Nathan)
- **Contact:** [audiolab.tools/contact](https://audiolab.tools/contact)