Skip to content

Free browser tools that respect your files — everything runs locally, nothing is uploaded.

Generative Focus Soundscape

Five layers — rain, deep drone, wind, stream, and pentatonic chords — synthesized live in your browser from a seed. Free, no sign-up, nothing uploaded.

0:00 elapsed

Press play — everything is synthesized in this tab.

Layers3/5 on

  • Rain

    Smoothed noise bed with droplet plinks

    22

  • Deep drone

    Three detuned oscillators, slow swell

    55

  • Wind

    Brown noise through a drifting bandpass

    30

  • Stream

    High-passed pink noise, gently pulsing

    30

  • Chords

    Pentatonic notes every few seconds

    42

Presets

Root note

The drone sits on this note and the chords draw from its major pentatonic scale.

Every random choice — noise buffers, droplet times, chord picks — derives from this one number, so the same seed rebuilds the same generative patch. Changing it while playing rebuilds the soundscape.

60

Fades out gently over the last 30 seconds, then stops.

Quick Answer

A focus soundscape generator builds an endless ambient bed — rain, a deep drone, wind, a stream, and sparse chords — algorithmically instead of looping a recording. This free tool synthesizes all five layers live with the Web Audio API in your browser: pick a preset, press play, and set a 25-, 50-, or 90-minute session that fades out on its own. Nothing is uploaded and no account is required; every sound is generated on your device.

How It Works

1

Pick a preset — Deep work layers rain, a drone, and slow chords — or toggle the five layers and set each level yourself

2

Press play: audio starts after that one click (browsers require it) and every layer fades in smoothly

3

Set the root note (A1 to D2) and the seed — the same seed rebuilds the same generative patch, and Copy link shares it

4

Choose a session length of 25, 50, or 90 minutes — playback fades out over the last 30 seconds and stops itself

5

Adjust layer levels and master volume while listening; every change ramps smoothly with no clicks

Key Facts

  • Five independent layers — rain with droplet plinks, deep drone, wind, stream, and generative pentatonic chords
  • All sound is synthesized live with the Web Audio API — no audio files, no streaming
  • One integer seed drives every random choice: the same seed rebuilds the same generative patch
  • Four presets: Deep work, Rain focus, Wind and stream, Everything
  • Five selectable root notes (A1 to D2); chords draw from the root’s major pentatonic across two octaves
  • Session timer (25, 50, or 90 minutes) fades out over the final 30 seconds, then stops
  • Shareable seed links (?seed=123) reproduce an exact patch; settings persist in your browser
  • Runs entirely client-side — free, no sign-up, nothing uploaded

Frequently Asked Questions

How does it make sound without audio files?

Every layer is synthesized live with the Web Audio API. The rain bed is white noise smoothed sample-by-sample by a one-pole lowpass filter into a looping 4-second buffer; wind is brown noise (a leaky integrator over white noise) through a slowly sweeping bandpass; the stream is pink noise (Paul Kellet’s filter) with a highpass and gentle amplitude modulation. The drone is three detuned oscillators under a 0.05 Hz swell, and droplet plinks and chord notes are short sources scheduled a couple of seconds ahead by a lookahead scheduler.

What does the seed do?

The seed is a number from 1 to 999999 that feeds a mulberry32 pseudo-random generator, split into one independent stream per generator. Those streams produce the noise buffers, the droplet timings, the chord picks, and the slow LFO phases — so the same seed rebuilds the same generative patch every time it plays. Press Randomize for a fresh patch, type a number to return to a known one, or use Copy link to share the exact patch as a ?seed= URL.

Is it free, and how does it compare to Brain.fm or Endel?

This tool is completely free with no account, trial, or limits. Brain.fm and Endel are subscription apps — as of late 2026 Brain.fm lists about 7 to $20 a month depending on plan and app store (check their sites for current pricing). Those products adapt to inputs like time of day or heart rate and license patented audio work; this tool is a simpler, transparent generative patch you control directly, which is exactly why it can be free and local.

Is it private, and does it work offline?

Yes. All synthesis happens in your browser tab — there are no audio streams and nothing you play is uploaded anywhere. The page itself loads from the network like any web page, but once it is open the soundscape keeps generating even if your connection drops. Your seed, mix, and settings are saved only in this browser’s localStorage.

Will it keep playing when I switch tabs or lock the screen?

Switching tabs is fine — playback deliberately keeps running in the background tab, which is the normal focus workflow. On desktop, audio continues while you work elsewhere. Some mobile browsers pause web audio when the tab goes to the background or the screen locks; if that happens, keep the tab foregrounded or use a dedicated browser window.

How is this different from a white noise generator?

White noise is one static sound; a generative soundscape evolves. Droplets plink at pseudo-random intervals, chord voices fade in over 4 seconds and release over 8, the drone swells on a 20-second cycle, and the wind filter drifts — so your ear keeps finding gentle novelty instead of locking onto a loop. The changes are slow and sparse by design: variation without distraction.