Noise has a color. Not metaphorically — literally, in the same sense that light has a color: a power spectrum. White noise has equal energy at every frequency. Pink noise rolls off at higher frequencies. Brown noise rolls off twice as steeply. The difference shows up immediately when you listen: white sounds bright and hissy, pink sounds like a steady rain, brown sounds like a low river or distant surf.

The choice between them isn't aesthetic. Each has a different masking profile and a different evidence base for sleep, focus, and tinnitus work. AmberRoom's recipes use different colors for different jobs — and the recipe inspector's NSE value links back here.

White · flat power across the spectrum

White noise has equal energy per Hz — a flat spectrum from 20 Hz to 20 kHz. The result sounds bright, almost hissy: most of the perceived energy is in the high frequencies, because human hearing has a lot of resolution up there.

White is good as a generic sound mask — a window AC unit, a fan. It's not great for sleep at moderate volume because the high-frequency content keeps the auditory cortex more activated than warmer noise colors. It's also rarely the right choice for tinnitus, where the goal is to reduce the salience of a high-frequency tone — feeding more energy into that band can make the ringing louder, not quieter.

Use case in AmberRoom: tinnitus recipe only — notch-filtered white noise at −18 dB with the notch centered on the user's tinnitus frequency. For sleep, focus, calm, and grief we use pink or brown instead.

Pink · 1/f, the natural-sounding one

Pink noise rolls off at −3 dB per octave — half the energy at each higher band. This 1/f spectrum is everywhere in nature: rainfall, distant traffic, the harmonic envelopes of music, even the firing patterns of certain neuron populations. The brain recognizes it as "background environment" rather than "novel signal," which is exactly why it's good for sleep masking.

Pink is the noise color with the strongest evidence base for sleep. The Papalambros (2017) study used pink-noise pulses, phase-locked to detected slow waves, to enhance slow-wave activity and improve next-morning memory in older adults. The broader sleep medicine literature also supports continuous pink noise for shortening sleep onset latency.

Use case in AmberRoom: sleep recipe (continuous pink at −20 dB underneath the 2.5 Hz delta binaural).

Brown · 1/f², the warm one

Brown noise (also called Brownian or red noise) rolls off at −6 dB per octave — steeper than pink, putting most of the energy in the lower frequencies. It sounds deeper, warmer, more "rumbly." If pink is steady rain, brown is a low river or distant ocean.

Brown is the better choice for focus work because it masks ambient distraction (HVAC, conversation, traffic) without adding bright high-frequency content that itself becomes attentionally salient. Many people who "don't like white noise" find they like brown immediately. There's less rigorous trial data on brown specifically — most of the "noise for focus" literature studies pink — but the masking-by-spectral-overlap mechanism is well-established.

For tinnitus, brown is often the safer default than white because it puts less energy into the high frequencies where most tinnitus presentations sit.

Use case in AmberRoom: focus (−28 dB), calm (−22 dB), grief (−26 dB), and meditation (−24 dB) all run a brown-noise floor underneath the band-targeted binaural. The energy recipe deliberately runs without a noise floor — bright crystal bowls only — to keep the morning lift spare.

The masking principle

All three noise colors work, when they work, by spectral masking: the broadband sound covers the irregular bumps, taps, voices, and whirrs in your environment that would otherwise pull your attention. The brain has to spend cognitive resources tracking unpredictable sounds; a steady noise floor takes that load off. This is why the same person who can't fall asleep on a quiet street falls asleep instantly on a plane: the cabin noise is doing the work.

AmberRoom's recipes use noise as a foundation, not a cover-up — sitting underneath the binaural and bowl layers at moderate volume, low enough that you're not straining to hear the rest of the recipe, high enough to do its masking job.

The active noise color shows in the recipe inspector — the NSE value. Each session tells you exactly what color and at what dB.