Pink noise and the evidence behind coloured sound

When you listen to pink noise, you hear an acoustic spectrum defined by how its energy is distributed across different frequencies. While many people choose steady background sounds at night, understanding what the colour names signify and what laboratory evaluations have actually measured clarifies where the evidence stands.

How acoustic colours distribute power across frequencies

The colour names describe how a noise's power is spread across frequencies. A 2022 review of sounds tested for sleep describes white noise as a signal with equal power in any frequency band of a given width, pink noise, also called 1/f noise, as having a power spectral density inversely proportional to frequency, and brown noise, also called red noise, as having a power spectral density inversely proportional to the square of the frequency, falling by about 6 dB per octave, so that it has more intensity at low frequencies than white or pink noise and sounds like a low roar ^[1]. A 2025 study that used all three as listening stimuli describes pink noise as having equal energy per octave, with more low-frequency components than white noise, and brown noise, unlike white noise, as decreasing in power as the frequency increases ^[2]. The same 2022 review notes that the term white noise is often used for any general hissing sound, and that various sounds labelled white noise can in fact be white, pink or brown according to their power spectrum ^[1].

What systematic reviews concluded on auditory stimulation

A 2021 systematic review defined its subject as continuous white noise or similar broadband noise, and left out studies that used intermittent white noise to disrupt sleep or to enhance slow wave activity ^[3]. Across its 38 included articles it found heterogeneity in the characteristics of the noise itself, as well as in sleep measurement and control conditions, and under the GRADE criteria it rated the quality of the evidence for a sleep benefit from continuous noise as very low ^[3]. A 2022 systematic review of auditory stimulation during sleep in adults found 34 studies with 1,103 participants: 18 on white noise, 11 on pink noise and 6 on combinations ^[4]. It counted positive findings in 6 of the white noise studies (33 percent), 9 of the pink noise studies (81.9 percent) and 4 of the combination studies (66.7 percent), judged that there was no strong evidence to support auditory stimulation, and noted that none of the studies reported adverse effects with short-term use ^[4].

CIBDOL · What systematic reviews concluded on auditory stimulation

Pink noise in laboratory sleep experiments

Pink noise was the only noise colour in a 2026 seven-night laboratory study with polysomnography in 25 healthy adults, mean age 28.5 ^[7]. Played on its own at a constant 50 dBA, it was followed by on average 18.6 minutes less REM sleep than a noise-free night, time that went largely to stage N2; no other measure of sleep structure differed from the noise-free night, and neither did morning ratings of sleep, alertness and mood ^[7]. Pink noise at 40 dBA was tested only together with intermittent environmental noise ^[7]. The authors caution against the widespread and indiscriminate use of broadband noise and call for research on its colour, level and long-term use ^[7]. A second paper from the same 25 participants reported that pink noise reduced arousals and awakenings caused by aircraft noise in a dose-response manner, but was outperformed by foam earplugs ^[8].

In a 2026 crossover pilot study, 12 healthy people slept five consecutive nights in acoustically isolated bedrooms with polysomnography ^[9]. Traffic noise events at 45 to 65 dB caused acute sleep fragmentation even though total sleep time and overall sleep structure were preserved, and these disturbances were attenuated when continuous 45 dB pink noise was played through the night ^[9]. The authors ask for the results to be interpreted cautiously because of the small and homogeneous sample ^[9]. Earlier, a 2012 study played steady pink noise to 40 people during a night's sleep and to 10 during a nap, each also sleeping a quiet period, and reported a significantly higher percentage of stable sleep time with the noise ^[10]. Stable sleep was estimated from the electrocardiogram with a cardiopulmonary coupling method ^[10].

Some pink noise research uses short pulses timed to brain waves rather than a steady background sound ^[11]. In a 2017 study of 13 healthy adults aged 60 to 84, an algorithm tracked slow waves in the EEG and delivered pulses of pink noise timed to their upstate, in runs of five pulses followed by a pause, on one night compared with a sham night ^[11]. Slow wave activity over the whole night was similar between the two nights, it was higher during the pulse runs than in matched sham periods, and overnight word recall improved more after the stimulation night ^[11].

White and brown noise in sleep investigations

In a 2017 randomised crossover study, 18 healthy people spent two nights in a sleep laboratory about a week apart, one with normal environmental noise at about 40 dB and one with broadband sound from two speakers at about 46 dB, and went to bed 90 minutes earlier than usual to make getting to sleep harder ^[5]. With the broadband sound, the time from lights out to the first epoch of stage N2 was 13 minutes instead of 19, a median reduction of 38 percent ^[5]. In a 2005 polysomnography study, 4 healthy subjects slept under three conditions: a baseline night, a night with played recordings of intensive care unit noise, and a night with the same recording plus mixed-frequency white noise; a fifth subject completed only the first two ^[6]. The arousal index was 13.3 per hour at baseline, 48.4 with the recorded noise and 15.7 with white noise added ^[6]. The authors conclude that peak noise was not the main determinant of the disruption, and that the added white noise raised arousal thresholds by reducing the difference between background and peak noise ^[6].

A 2025 systematic review of sound machines for adults in hospital included 7 randomised trials with 496 participants, six of which used white noise ^[13]. The reviewers report findings in favour of white noise on sleep efficiency and good tolerability, with only 8 participants stopping because of discomfort or finding it less acceptable, but judged the evidence limited and too varied in duration of use, inclusion criteria and outcome scales to allow a meta-analysis ^[13].

A 2022 study played brown noise on its own on one of its study days and measured sleep ^[12]. In it, 54 patients in a hospital high dependency unit in India were divided into an app group and a control group of 27 each; the app group used a noise app through earphones twice a day for three days, with white noise on day 1, pink noise on day 2 and brown noise on day 3, and sleep was rated on a 4-point questionnaire scale ^[12]. The reported comparisons are between the app group and the control group, which differed significantly on each day; each colour was used on its own fixed day ^[12]. A 2026 uncontrolled pilot in 15 healthcare workers mixed brown noise with nature sounds and other acoustic components, so brown noise was not tested by itself there ^[15].

How awake listeners rate and respond to the colours

A 2025 study tested the idea that the colours differ in how arousing or calming they are, while people were awake ^[14]. In this pupillometry study analysing 31 listeners who heard white, pink and brown noise for 10 seconds each, pupil size did not differ significantly between the three noises ^[14]. In two 2025 rating studies with 84 and 1,280 participants, awake listeners rated brown noise as more sublime than white and pink noise and white noise as more uneasy than the other noises, and none of the three noises received the minimal emotional ratings the authors required of a neutral control ^[2]. Sleep was not measured in those rating investigations ^[2].