Pink noise and sleep

Published 2026-05-18 · 8 min read

Pink noise occupies a particular position in the sleep-sound conversation: it has the best-developed peer-reviewed evidence base of any noise colour, but the evidence is for a fairly specific application (slow-wave-coupled memory consolidation in adults), not for general sleep onset. Pink also has a strong subjective claim because its spectrum resembles the acoustic environment in utero, which is part of why it feels familiar to newborns. This page walks through the spectrum, the Papalambros 2017 Northwestern study and its follow-ups, the intrauterine analogy, and how pink fits alongside white and brown for both adult and infant sleep.

What pink noise is

Pink noise is broadband acoustic noise whose power spectral density rolls off at -3 dB per octave. In plain terms, this means each octave (each doubling of frequency) contains the same total energy. White noise, by contrast, has equal energy per Hz, which makes upper octaves (which span many more Hz) dominate the perceived sound. Pink noise spreads the energy more evenly across the human-audible spectrum from roughly 20 Hz to 20 kHz.

Pink noise sounds like steady rain, distant ocean surf, or wind through trees. It has less hiss than white noise and less rumble than brown noise. Many sound design textbooks describe pink as the “natural” noise spectrum because it matches the statistical structure of many environmental sounds (waterfalls, rivers, foliage in wind), as well as the spectrum of biological signals like the human voice averaged across speakers.

The naming comes from the visible-light analogy: white noise is named for white light (all frequencies equally represented), pink noise is shifted toward the lower-frequency “red” end of the audio spectrum, hence the pink label. Brown noise is the deeper variant, sometimes called red noise, with -6 dB/octave roll-off (see brown noise).

The intrauterine resemblance

Acoustic measurements taken from the womb (intrauterine microphone studies dating back to the 1980s) consistently show a sound environment dominated by low-frequency content from maternal vasculature, heartbeat, breathing, and intestinal motility, with mid-frequency content from maternal voice and external environmental sound attenuated by tissue. The aggregate spectrum is broadly pink-shaped, with a slight emphasis on the lower frequencies.

This is the basis for the “fourth trimester” framing popularised by pediatric authors and used widely in newborn care. Newborns calm reliably to sounds with intrauterine-like spectra, which includes pink noise, brown noise, and many naturalistic recordings (waves, rain, vacuum cleaners). The mechanism is partly familiarity (the auditory system has been exposed to similar input for the third trimester) and partly the masking effect, which lets the developing infant's startle response stay quieter.

None of this changes the AAP volume guidance. Pink noise at 60 dB at 30 cm is still too loud for an infant, intrauterine resemblance notwithstanding. The 50 dB head-level ceiling applies to pink the same as to white (see 50 dB ceiling).

The Papalambros 2017 study and what it actually showed

The most-cited pink noise sleep study is Papalambros, Santostasi, Malkani, Braun, Weintraub, Paller, and Zee, “Acoustic Enhancement of Sleep Slow Oscillations and Concomitant Memory Improvement in Older Adults,” published in Frontiers in Human Neuroscience in 2017. The study, from the Cognitive Neurology and Alzheimer's Disease Center at Northwestern University, enrolled 13 older adults (mean age around 75) and used a closed-loop acoustic stimulation system. During NREM sleep, the system detected the rising phase of slow-wave EEG oscillations in real time and played short pink noise bursts timed to coincide with those rising phases.

The result: the closed-loop pink noise nights produced larger slow-wave amplitudes during sleep and statistically significant improvements in word-pair recall the following morning, compared to sham nights without acoustic stimulation. The effect size was modest but consistent and has been replicated in subsequent work from the same lab and from independent European groups.

What this does not show: that running ambient pink noise from a sound machine at bedtime improves memory. The Papalambros effect depends on closed-loop timing to the EEG, which requires a worn device that can detect slow-wave phase. Consumer sound machines do not do this. The clinical takeaway is real but narrow: there is a mechanism by which pink noise can enhance sleep-dependent memory consolidation, and there is published evidence it works in a controlled setting. The translation to consumer use is not straightforward.

What pink noise reasonably does for sleep without a closed loop

Without EEG-timed bursts, the case for pink noise rests on the same masking argument as any colour: it makes environmental sound less disruptive by raising the ambient floor, which reduces the contrast between background and intrusive events. Pink does this more efficiently per dB than white, less efficiently than brown (which puts even more energy in the low band where most disturbances live).

For sleep onset, the spectrum probably matters less than the consistency. Habituation to a reliable acoustic environment is the primary mechanism, and any reasonably-shaped broadband noise will do. For night-long sleep maintenance, lower-frequency spectra (pink, brown) win on the subjective comfort axis: they are less fatiguing across hours.

For specific tinnitus masking, the audiology guidance (American Tinnitus Association, British Tinnitus Association) treats pink as one of several useful spectra; individual response varies. See the existing tinnitus page.

Pink vs white for babies

For infants, the AAP 50 dB ceiling is colour-agnostic. Practical considerations favour pink and brown:

One, the intrauterine resemblance gives pink and brown a head start on familiarity. Many newborns calm faster to pink than to white in the first few weeks.

Two, the lower-frequency emphasis means pink masks effectively at slightly lower dB than white, giving you margin on the 50 dB cap. A pink machine reading 45 dB at the head often masks comparably to a white machine reading 49 dB.

Three, the gentler high-frequency content is less likely to disturb a baby's developing auditory system over many hours. This is anecdotal rather than evidence-based at the consumer level, but the audiology consensus broadly supports it.

Frequently asked

Can I get the Papalambros benefit from a sound machine?

Almost certainly not. The published effect depends on closed-loop EEG-timed acoustic bursts, not on ambient pink noise. Some research wearables (Dreem, Philips SmartSleep, both since discontinued) attempted to commercialise the technique; none are currently mainstream consumer products.

What's the best pink noise app?

The major sleep apps (Calm, Headspace Sleep, BetterSleep, Sleep Sounds) all include pink as a preset. Free in-browser generators (including the one on this site’s homepage) work fine. Spectrum quality is broadly comparable across mainstream sources.

Is pink noise safe for premature babies?

At AAP-compliant volume, yes. NICU acoustic guidance restricts ambient sound to <45 dB average regardless of source, which is more conservative than the home AAP infant ceiling. Speak to your NICU team before introducing any sound source for a hospitalised preterm infant.

Does pink noise help with ADHD focus?

The Söderlund line of research on broadband noise and inattentive children used white noise specifically. Pink noise has similar acoustic properties and likely a similar mechanism, but is less directly studied in that population. See the existing /for-adhd/ page.

Sources

Brown noise for sleepGreen noiseGrey noiseWhite vs pink vs brownWhite noise for newborns

Updated 2026-04-27