Grey noise

The equal-loudness contour

The human ear's frequency response was systematically mapped by Fletcher and Munson in the 1930s and revised by Robinson and Dadson in the 1950s. The current international standard is ISO 226:2003 (revised 2023), which publishes equal-loudness contours: curves showing the sound pressure level required at each frequency to produce equal perceived loudness.

The contours have a characteristic shape: they dip in the middle (the ear is most sensitive around 2 to 4 kHz, the speech band) and rise at the extremes (less sensitive to very low and very high frequencies). A 1 kHz tone at 60 dB sounds about as loud as a 100 Hz tone at 70 dB or a 50 Hz tone at 80 dB. The shape of the dip varies slightly with overall loudness, which is why the standard publishes multiple curves at different loudness levels.

Grey noise is built by taking white noise and applying a filter that is the inverse of the equal-loudness contour. Where the ear is most sensitive (3 kHz), the filter reduces energy. Where the ear is least sensitive (sub-100 Hz and above 10 kHz), the filter boosts energy. The result is a signal whose physical spectrum is highly uneven but whose perceived spectrum is flat.

What grey noise sounds like

Subjectively, grey noise sounds neutral, soft, and present without being intrusive. Listeners often describe it as “the noise that does not have a colour” or “noise without character.” The lack of a dominant band makes it easy to ignore once it has been playing for a few minutes, which is part of why some users prefer it for sleep.

Compared to brown noise, grey is less rumbly. Compared to white, it is less hissy. Compared to pink, it lacks the natural-water quality. It is the noise spectrum closest to “perceptual silence with substance,” if that paradox is allowed.

The lack of a dominant band has a practical cost: grey noise is less efficient at masking specific intrusive sounds than the more focused colours. Traffic noise lives in the low to mid band where brown noise puts most of its energy; brown is therefore the better masker for traffic. Snoring lives in the very low and mid band; again, brown wins. Grey's strength is comfort, not masking performance.

Audiology and research applications

The original use case for grey noise was audiological testing where a stimulus needed to be perceptually neutral. If you are testing hearing thresholds across frequencies and you want the background noise to bias the results as little as possible, grey noise is the appropriate spectrum because it does not artificially elevate the masking pressure on any particular frequency band.

Grey noise also appears in psychoacoustic research as a control stimulus, in hearing aid fitting protocols where a perceptually flat background is preferred, and in some tinnitus assessment workflows. None of these are mainstream consumer applications; they are clinical and research uses that have leaked into the consumer sleep sound conversation through audio engineering communities and sleep researchers who happen to also be hobbyist audiophiles.

For the typical sleep user, grey noise is one of the more obscure options. Most mainstream sleep apps include it (Calm, BetterSleep, Sleep Sounds, the in-page generator on this site), but it rarely tops the popularity charts. The audience is technically curious users and a subset of audiology-aware tinnitus sufferers who find the perceptually flat profile easier to habituate to than spectrum-emphasised alternatives.

Grey for sleep: when it makes sense

Three scenarios where grey noise is a reasonable choice.

One, the listener finds brown noise too rumbly and white noise too harsh. Grey sits between them in perceived character. For a small but consistent subset of users, this middle position is the most comfortable for long-night exposure.

Two, the environmental noise to be masked is itself spectrally varied (a mix of traffic, conversation, HVAC, neighbour activity) rather than concentrated in one band. Grey noise provides broad-spectrum masking without overcommitting to any one frequency range, which can match a mixed environment better than a single-band-focused colour.

Three, the listener has variable hearing sensitivity across frequencies (mild age-related high-frequency loss is common) and wants a sound that does not feel imbalanced. Grey's perceptual flatness can paper over moderate hearing variation in a way that white or violet noise does not.

For babies: technically fine, practically rare

Grey noise at AAP-compliant volume is safe for infants on the same terms as any other colour. The 50 dB head-level ceiling and the 7 ft placement rule apply uniformly. There is no spectrum-specific concern.

Practically, grey is rarely chosen for nursery use because the intrauterine resemblance argument (which favours pink and brown) does not apply, and the masking efficiency argument also favours brown for typical home environmental noise. Grey is neither preferred nor contraindicated for infants; it is simply not the choice anyone reaches for first.

For toddlers and older children, individual preference matters more. Some children find grey calming where they find brown too dark. Pediatric sleep guidance does not differentiate by spectrum, only by dB.

Frequently asked

Sources

• ISO 226:2003 (revised 2023) Acoustics, Normal equal-loudness-level contours, iso.org
• Fletcher H, Munson WA. “Loudness, its definition, measurement and calculation.” Journal of the Acoustical Society of America 1933; 5(2): 82-108
• Robinson DW, Dadson RS. “A re-determination of the equal-loudness relations for pure tones.” British Journal of Applied Physics 1956; 7(5): 166
• Acoustical Society of America, acousticalsociety.org
• Sleep Foundation, noise colour reference