Resonance

An ambient room, composed while you listen.

How it works ↓

Silent — the ring is flat because nothing is playing

Everything you hear is being written now

There is no recording on this page. When you press play, your browser builds a small synthesizer with the Web Audio API — oscillators, filters, a delay line, and a reverb whose impulse response is generated on the spot — and then a scheduler starts choosing notes from the current mood's scale, at unhurried random intervals, forever. The piece you hear has never been played before and will not be played again.

The visual is the signal

The ring is not an animation timed to look musical. An AnalyserNode sits at the end of the audio graph and reports the actual waveform and spectrum of what is reaching your speakers; the ring's shape is the waveform bent into a circle, and the glow behind it follows the energy in the low bands. Silence draws a flat circle — which is what silence looks like.

Three moods

Dawn leans on a bright major pentatonic scale with bell-like tones. Drift floats long, detuned pad swells over a two-note drone. Undertow moves to a low minor scale, darker filtering, and slower events. Switching moods retunes the scale, the drone, and the filter while the room keeps breathing.

A short, honest history

Music made by systems rather than performances has a long lineage. Brian Eno's Discreet Music (1975) used a tape-delay feedback loop as a composition machine, and his liner notes for Ambient 1: Music for Airports (1978) named the genre: ambient music should be "as ignorable as it is interesting." In the mid-1990s he popularized the term generative music for pieces that are rules, not recordings — this page is a small descendant of that idea.