Credits, licence and provenance

Gnaural

This studio is a browser re-implementation of Gnaural, Bret Logan's open-source binaural-beat scheduler. The scheduling model — voices, entries that ramp into the next entry, the last entry wrapping to the first — the .gnaural XML format, and the synthesis approach all follow the upstream program (BinauralBeat.c, gnauralXML.c, ScheduleXML.c, release 1.0.20110606).

Engine (engine.js): a re-implementation written from the published upstream source. Upstream is LGPL-2.1-or-later; this port keeps those terms. Full text: COPYING.LESSER.
Everything else on this site — the UI, the panels, the library generator, the preset packs, the help pages: GPL-2.0-or-later. Full text: COPYING.
No upstream preset files are bundled: the upstream preset archive states no licence. Use Open .gnaural for files you already have.
Not affiliated with, endorsed by, or connected to the upstream project.

What is original to this site

The whole library — every session, pack and preset is a tone schedule written here. No audio, no narration, no scripts and no manual text from any recording, programme or product is included anywhere on this site.
The focus-level presets are original schedules built from publicly documented tone frequency sets (carrier-and-beat measurements republished by enthusiasts since the 1990s), named descriptively rather than after any exercise, and shipped with no affiliation of any kind to the Monroe Institute, Hemi-Sync, or any other organisation.
The speaker pack is designed here: isochronic pulses and one monaural beat, carriers limited to 200–500 Hz, no interaural pair anywhere in the pack — and the project's own tests assert that property rather than promising it.
The evidence discipline — the claim ledger, the statuses on every number, and the refusal to make a medical claim — is this site's own editorial choice.

What this site claims, in full

Claimed: the studio produces, edits, renders and shares precisely scheduled tone layers; the synthesis matches the upstream engine's behaviour; a periodic tone at rate R produces a measurable neural response phase-locked to R while it plays (the frequency-following / auditory steady-state response), and that response grows with level.

Not claimed: that any session here changes your brain state, improves focus or sleep, treats any condition, or has any health effect at all. The 2023 systematic review of 14 EEG entrainment studies found 8 contradicting the entrainment hypothesis, 5 supporting it and 1 mixed — no clear trend either way. Where a number comes from a folk tradition, the page says so. Where it comes from physics, the page says how far the physics goes. The evidence page lists every claim with its status.

Built with, and without

With: plain HTML, CSS and JavaScript; the Web Audio API (AudioWorklet where available, with a main-thread fallback); canvas for the graph, meters, spectrum and phase wheel.
Without: no frameworks, no build step at runtime, no CDNs, no external fonts, no analytics, no accounts, no cookies, no server-side processing of anything you do here. The share links encode a session in the URL fragment, which browsers never send to a server.
Privacy, plainly: your sessions, favourites and skin choice live in your browser's local storage and nowhere else. Nothing is transmitted anywhere, ever.

Colophon

Written for this site. The library and the help pages are generated and checked by build scripts, so the numbers in the prose are the numbers the audio engine plays. Pack archives are byte-deterministic. The site is tested headlessly — synthesis, file parsing and packing all have Node test suites, and the page itself is driven by a browser harness before anything ships.

Licence texts: COPYING (GPL-2.0-or-later) · COPYING.LESSER (LGPL-2.1-or-later, the engine) · NOTICE.txt · moddys.net (attribution and provenance, authoritative).