White, Pink, Brown: The Physics of Focus Noise

A colleague's phone call pierces silence; it vanishes under steady noise. That's masking — and the Ambient Noise Generator synthesizes the three classic noise colors from pure math, no audio files, no repetition seam.

The three colors

White noise packs equal energy per hertz. To the ear it's bright and static-like, and it masks best: every frequency band of the interruption finds noise already sitting on it.

Pink noise falls off at −3 dB per octave — equal energy per octave, matching how hearing itself works. It sounds balanced, rain-like, and is the one most people can listen to for hours. Research on sleep (where pink slow-wave audio improved next-day memory in small studies) made it famous beyond the office.

Brown noise (red) falls −6 dB per octave — the rumble of wind and waterfalls, deepest and most soothing for falling asleep.

Synthesized, not sampled

White is uniform random samples. Pink uses Paul Kellet's refined filter — seven running accumulators shaping the spectrum just right. Brown integrates white noise with a leak (0.02 coefficient), then scales. A few seconds of each are generated into looping audio buffers; long buffers make the loop point inaudible.

The mixer sums the three voices into a master gain, and sliders adjust levels live while playing — with setTargetAtTime smoothing so changes glide instead of clicking. The sleep timer (5–120 min) fades playback out automatically so brown noise doesn't run all night.

Using it well

Start around 40% pink; add white for serious masking, brown for sleep. Pair with the Pomodoro Focus Timer for the work/break structure the noise fills in — and keep volumes modest: masking needs surprisingly little level to work.