Pink noise is a random sound whose power falls as the pitch rises, in proportion to 1/f: minus 3 decibels per octave, so every octave carries the same share of power. It sounds softer and rounder than white noise, often compared to a waterfall. The famous "pink noise for deep sleep" results come from short pulses of pink noise timed to the sleeping brain by EEG equipment, not from a track playing all night. Studies of continuous pink noise at night are small and mixed, and the largest recent one found it reduced REM sleep.
This page goes deeper on pink noise alone. For how white, pink and brown noise compare, see our guide to white, pink and brown noise.
Pink noise, physically
The Sound On Sound glossary defines pink noise as "a random signal with a power spectral density which is inversely proportional to the frequency. Each octave carries an equal amount of noise power." In numbers:
- 1/f: double the frequency and the power per hertz halves.
- Minus 3 dB per octave, which is the same as minus 10 dB per decade (a decade is a tenfold step in frequency).
- Equal power per octave: the octave from 100 to 200 Hz carries as much power as the octave from 5,000 to 10,000 Hz, even though the second is 50 times wider.
That's why pink noise sounds more even to the ear than white noise: we hear pitch in octaves, and pink noise gives each octave the same weight.
Where pink noise shows up. It's also called 1/f noise or flicker noise. An engineering textbook chapter on noise (Woolf, Engineering LibreTexts) notes that it is "evident in electrical circuits," for example as the excess noise in carbon composition resistors above their thermal noise. In a 1975 paper in Nature, Richard Voss and John Clarke wrote that the power spectrum of many fluctuating physical variables is "1/f-like," and reported that loudness fluctuations in music and speech, and pitch fluctuations in music, show 1/f spectra. That's about how music changes over time, not about the sound of music itself, but it's one reason pink noise is sometimes called the most "natural" color.
How a pink noise generator works
You can't record pink noise from nature on demand, so apps and generators make it from white noise:
- A pinking filter. White noise goes through a filter that takes away 3 dB per octave. In practice this is several simple filters added together, each covering part of the range.
- The Voss-McCartney algorithm. Several random sources are added up, each changing at half the rate of the one before, so the lower octaves get their share.
AI Meditation: Sleep & Calm uses the first method: its code runs Paul Kellet's "refined" pinking filter, which Robin Whittle's page on pink noise generation describes as "a weighted sum of first order filters" and dates to the music-dsp code archive in 1999. According to that page, Kellet stated that it is "accurate to within +/-0.05dB above 9.2Hz (44100Hz sampling rate)." The app plays its sounds at exactly that rate, 44,100 samples per second (from the app's code). Our reading: at the app's sample rate, the filter meets Kellet's stated accuracy from 9.2 Hz upward. We have not measured the sound from a phone speaker, which also shapes what you hear.
"Pink noise restorative sleep study": what was actually tested
Searches for "pink noise restorative sleep study" lead to one line of research: closed-loop acoustic stimulation. Here's how it works and what it found.
- The method. Equipment reads your brain waves (EEG) while you sleep and plays short pulses of sound at the exact moment a slow wave of deep sleep is rising. A 2020 paper in Nature and Science of Sleep summarizes it: phase-locked pink noise stimulation "can amplify slow oscillatory activity." A 2013 study from the University of Tübingen in Neuron found that stimulation in phase with the slow oscillation enhanced it and improved memory consolidation, while stimulation out of phase did not.
- In older adults. In a 2017 study of 13 adults aged 60 to 84, pulses of pink noise timed to slow waves increased slow-wave activity while the pulses played, though not over the whole night, and improved word recall the next morning. Our comparison guide describes this study.
- The meta-analysis. A 2021 systematic review and meta-analysis in Sleep pooled 10 studies with 177 participants on acoustic slow-wave stimulation and memory. The combined effect was small (Hedges' g 0.25) and did not reach statistical significance (p = 0.07); only the subgroups with phase-locked stimulation reached significance. The authors concluded: "Currently, the evidence is not sufficient to recommend the use of commercially available devices."
- A retraction. A 2022 meta-analysis in the Journal of Sleep Research that reported larger memory effects in young adults was retracted in 2026, according to its PubMed record. If you see it cited, it no longer counts.
The key point: all of this used pulses timed to your brain, not a continuous sound. Even a 2020 study of non-phase-locked pink noise, which increased the share of deep (N3) sleep in 8 participants, had a sleep technologist watching the recordings live and choosing when to play the sounds.
Continuous pink noise at night: what the studies found
When pink noise simply plays all night, as it would from an app, the results are small and point in different directions:
| Study | Who | What they did | What they found | Limits |
|---|---|---|---|---|
| Kawada and Suzuki, 1993 | 1 young man, then 4 students | Steady pink noise at 60 dB(A) | Time to fall asleep tended to be shorter with the noise | 5 people; 60 dB is about the level of normal conversation according to the WHO |
| Zhou et al., 2012 | 40 adults (night sleep), 10 (naps) | One night or nap with steady pink noise, one quiet | More "stable sleep time," estimated from heart and breathing signals | Indirect sleep measure (from an ECG-based method) |
| Vickrey and Lerner, 2023 | 72 adults | Wake, silent sleep, or sleep with open-loop pink noise at home, then an insight task | No gain in insight with pink noise; less time in the first light-sleep stage (N1) after falling asleep | One task; home headband recordings |
| Vincens et al., 2026 | 12 healthy adults | Five nights in sound-isolated bedrooms, with traffic noise, pink noise at 45 dB, or both | Pink noise reduced the sleep disruption and blood changes linked to traffic noise | Pilot; the authors ask for cautious interpretation |
| Basner et al., 2026 (Sleep) | 25 healthy adults | Seven lab nights with environmental noise, pink noise at 40 or 50 dBA, earplugs and combinations | Pink noise reduced REM sleep; adding it to environmental noise worsened sleep structure overall; earplugs prevented nearly all noise effects until 65 dBA | Young, healthy group; lab setting |
| Basner et al., 2026 (Sleep Health) | Same 25 adults | Compared noise sources | Pink noise reduced aircraft noise awakenings in a dose-dependent way but was outperformed by earplugs; a crying baby and a fire alarm were the most likely to wake people | Same study, second analysis |
Taken together: continuous pink noise may soften the effect of noise you can't control, like traffic. On a quiet night it isn't neutral, and earplugs did better in the largest study. The authors of the 2026 Sleep paper caution against "the widespread and indiscriminate use of broadband noise."
Pink noise vs white noise, and pink vs brown
In one line each (our comparison guide has the full table):
- Pink vs white: white noise has equal power per hertz and sounds like a bright hiss; pink noise has equal power per octave and sounds softer. See white noise.
- Pink vs brown: brown noise falls twice as fast, minus 6 dB per octave, and loses almost all of the hiss. See brown noise.
No study we found has shown one of the three to be better for sleep than the others.
What the app's pink noise sounds like
The app describes its pink noise as "softer and more balanced than white noise. A gentle, steady wash that many people find calming for sleep." That last part is a description of how people use it, not a study result. It's generated live on your phone, so there's no loop point, and you can set a timer from 5 to 90 minutes, after which it fades out and stops, or let it play until you stop it.
How to use pink noise for sleep
- Use it to mask something, like traffic or a snoring partner. In a quiet room, the 2026 lab study suggests it may do more harm than good.
- Keep it low. The recent studies used 40 to 50 dB. The WHO puts a library at about 40 dB and normal conversation at about 60 dB.
- Use a timer so it fades out after you fall asleep. In the 2026 lab study, the REM reduction came from pink noise playing through the night.
- Try earplugs if the noise you want to block is loud. They did better than pink noise in the largest study.
- Mind the volume for your hearing. The US National Institute on Deafness and Other Communication Disorders (NIDCD) says sounds at or below 70 dBA are unlikely to cause hearing loss even after long exposure; long or repeated exposure at or above 85 dBA can. The WHO suggests keeping your device's volume at no more than 60% of maximum.
Pink noise in the app
AI Meditation: Sleep & Calm has pink noise alongside white, brown, blue, violet and grey noise, and four soft tonal sounds: Deep tone, Warmth, Crystal and Resonance. There are no nature recordings like rain or ocean in the app. If you'd rather fall asleep to a voice, there are guided sleep meditations and sleep stories, read by AI voices. See how the app works, try it for free, no account needed.
Your health
This app is a wellness app and does not replace medical advice. If sleep problems last for weeks or affect your day, talk to a doctor. If you notice changes in your hearing, see a doctor as well. The NHS insomnia page explains when to see a GP about sleep.
Sources
All pages accessed on September 25, 2026.
- Sound On Sound: Pink Noise (glossary). 1/f, equal power per octave, waterfall.
- Woolf P. et al.: Noise Modeling: White, Pink, and Brown Noise, Pops and Crackles, in Chemical Process Dynamics and Controls, Engineering LibreTexts (CC BY 3.0). Pink noise in electrical circuits and resistors.
- Voss RF, Clarke J: ‘1/f noise’ in music and speech. Nature 1975;258:317-318 (abstract).
- Whittle R: DSP generation of Pink (1/f) Noise (firstpr.com.au). Pinking filters, Voss-McCartney algorithm, Paul Kellet's refined filter and its stated accuracy.
- Ngo HV et al.: Auditory closed-loop stimulation of the sleep slow oscillation enhances memory. Neuron 2013;78(3):545-553.
- Papalambros NA et al.: Acoustic Enhancement of Sleep Slow Oscillations and Concomitant Memory Improvement in Older Adults. Frontiers in Human Neuroscience 2017;11:109.
- Wunderlin M et al.: Modulating overnight memory consolidation by acoustic stimulation during slow-wave sleep: a systematic review and meta-analysis. Sleep 2021;44(7):zsaa296.
- Stanyer EC et al.: The impact of acoustic stimulation during sleep on memory and sleep architecture: A meta-analysis. Journal of Sleep Research 2022;31(3):e13385. Retracted (J Sleep Res 2026;35(1):e70111).
- Schade MM et al.: Enhancing Slow Oscillations and Increasing N3 Sleep Proportion with Supervised, Non-Phase-Locked Pink Noise and Other Non-Standard Auditory Stimulation During NREM Sleep. Nature and Science of Sleep 2020;12:411-429.
- Kawada T, Suzuki S: Sleep induction effects of steady 60 dB (A) pink noise. Industrial Health 1993;31(1):35-38.
- Zhou J et al.: Pink noise: effect on complexity synchronization of brain activity and sleep consolidation. Journal of Theoretical Biology 2012;306:68-72.
- Vickrey B, Lerner I: Overnight exposure to pink noise could jeopardize sleep-dependent insight and pattern detection. Frontiers in Human Neuroscience 2023;17:1302836.
- Vincens N et al.: Pink noise reduces impact of traffic noise on sleep and the blood metabolome: a cross-over pilot study. Communications Medicine 2026;6(1):114.
- Basner M et al.: Efficacy of pink noise and earplugs for mitigating the effects of intermittent environmental noise exposure on sleep. Sleep 2026;49(5):zsag001.
- Basner M et al.: The efficacy of pink noise and earplugs for mitigating sleep disruption induced by different environmental noise sources. Sleep Health 2026;12(4):702-706.
- NIDCD: Noise-Induced Hearing Loss (last updated April 16, 2025). 70 and 85 dBA.
- WHO: Deafness and hearing loss: Safe listening (questions and answers, March 6, 2026). 60% volume, decibel examples.
