audio techniques

Pink Noise for Sleep: What the Slow-Wave Research Actually Found

Pink noise has the strongest sleep evidence of the three noise colors. What the slow-oscillation studies found, and why an app playlist is not the same thing.

Sample · Lunaria Rain at a distance — settling into the ambient layer 39s
A short Murmora whisper. Make your own →

Pink noise is the middle child of the noise colors, and the only one with a genuine research story attached to it. Search results tend to flatten that story into a claim: pink noise deepens sleep and improves memory. The underlying studies are real. What they did is narrower and stranger than the claim suggests, and the gap between the two is the most useful thing this page can give you.

The short version: pink noise delivered in short bursts, timed by an EEG rig to the exact phase of a sleeping brain’s slow oscillations, has been associated with more slow-wave activity and better morning recall. Pink noise streaming continuously from your phone is a different intervention with a different, more modest case. Both are worth understanding, and they are frequently sold as the same thing.

What pink noise is, in one paragraph

Pink noise is a random signal whose energy falls off as frequency rises, at a rate of 3 decibels per octave. Every time the frequency doubles, the power halves. That slope happens to match how a lot of natural sound distributes its energy, which is why pink noise is usually described in natural terms: steady rain, a shower, wind through leaves, a river some distance away. It has less high-frequency hiss than white noise and less low-end weight than brown noise, sitting almost exactly between them.

The general reason background sound helps sleep at all is auditory masking, and that mechanism belongs to all three colors equally. We have covered it in detail in the guide to white noise for sleep, so this page won’t repeat it. What follows is the part that is specific to pink.

The slow-oscillation research, described accurately

During stage-3 NREM sleep, the cortex produces large, slow oscillations of roughly one cycle per second, alternating between an active up-state and a quiet down-state. These are the delta-range waves that define the deepest part of the night, and their up-states are when a lot of overnight memory consolidation appears to happen.

Ngo and colleagues published a study in Neuron in 2013 testing whether that rhythm could be nudged. They monitored sleeping participants’ EEG in real time and played very brief pink-noise bursts locked to the rising phase of each detected slow oscillation. The stimulated nights showed enhanced slow oscillations and better retention of word pairs learned before bed than the sham nights. Papalambros and colleagues extended the approach to older adults in Frontiers in Human Neuroscience in 2017, with a similar pattern of results: more slow-wave activity during stimulation, and improved recall the following morning.

Two things about these findings are easy to lose. The first is that the timing was the intervention. Sound played at the wrong phase of the oscillation does not help and can disrupt. The second is that replication across the wider literature has been uneven, with effects that vary by age, by baseline slow-wave activity, and by how the stimulation is tuned. This is a promising line of work rather than a settled one.

Sample · Lunaria Rain at a distance — settling into the ambient layer 39s
A short Murmora whisper. Make your own →

Why a pink noise app is not that experiment

The equipment is the difference. Closed-loop acoustic stimulation needs electrodes on the scalp, live slow-oscillation detection, and a system that can fire a fifty-millisecond burst within a narrow window of a one-second cycle. A continuous pink-noise track does none of this. It plays the same sound at the same level regardless of what your brain is doing, which means it lands at every phase of every oscillation, favorable and unfavourable alike.

This is worth being blunt about, because the citation gets borrowed constantly. A product page describing pink noise as clinically shown to deepen slow-wave sleep is usually pointing at Ngo or Papalambros while selling something those studies did not test. The same pattern shows up in binaural beats marketing, where a narrow laboratory finding gets stretched across a category.

There is a smaller and less publicized line of work on steady, non-phase-locked pink noise, which has found modest increases in the proportion of stable sleep. The framing there is stabilization rather than enhancement: the sound may help sleep hold its shape, not push it deeper. That is a reasonable thing to hope for from a track you press play on. It is not the memory result.

What steady pink noise can reasonably do for you

Strip out the overclaiming and a decent case remains.

It masks well, and it masks the frequency range most household disruptions occupy. Traffic, a partner turning over, a building’s pipes, a door two floors down. Those intermittent sounds are what break the light transitions between sleep cycles, and raising the room’s floor blunts them.

It is easy to tolerate for eight hours. This sounds trivial and isn’t. The most common reason people abandon white noise is that the hiss becomes irritating once the novelty passes. Pink noise’s gentler high end survives long exposure better, which matters more for a nightly habit than any marginal difference in mechanism.

It may stabilize more than it deepens. If you are tracking anything, track awakenings rather than depth. That is where the plausible effect lives.

And it does nothing for the internal layer. Masking has no purchase on a mind rehearsing tomorrow’s conversation. If your room is already quiet and you are still awake at midnight, the acoustic environment was never the variable. Sleep hygiene covers the setup conditions, and practices that put actual content into the sleep-onset window address the rest.

Pink, white, or brown — deciding in one night

The honest ranking is short. Pink noise has the better research pedigree, though most of it belongs to a technique you cannot run at home. White noise has the broadest masking evidence in noisy real-world settings. Brown noise has the thinnest direct research and the strongest subjective following among people who find the other two fatiguing.

Since the working mechanism you will actually get is the same across all three, choose by ear. Play each for ten minutes at the volume you would sleep at, and notice which one you stop hearing first. The one that disappears fastest is the one that will still be tolerable at three in the morning. That test is more predictive than any of the frequency charts.

Sample · Benjamin Toward the slow part of the night 35s
A short Murmora whisper. Make your own →

The version of this Murmora is built around

We publish this as a company that makes sleep audio, so it is worth saying plainly what we think ambient sound is for. Pink noise handles the room. It cannot handle what you brought into the room with you, because it carries no content, and content is the thing that reaches the receptive state at sleep onset rather than merely protecting it.

Murmora’s sessions are built for that second layer: personalized affirmations written around your actual situation, spoken in a guide voice you choose or in a clone of your own, and delivered sparsely through the night — a phrase, then several minutes of quiet, rather than a continuous voice competing with your sleep. Whatever ambient track you already run can sit underneath it. The mechanism behind the verbal layer is worked through in our guide to whether affirmations work while sleeping. You can hear the voices at /#voice.

What to try this week

Run a plain seven-night test, and hold everything else constant. Pink noise at conversation volume or below, speaker across the room rather than on the pillow, on from lights out through the night. Each morning write down one number: how many times you remember waking. Not how deep it felt, which nobody can judge from the inside, and not how long it took to fall asleep, which is confounded by too much else.

If the count drops, masking is doing something for you and it is worth keeping. If it doesn’t move, you have learned that sound was not the constraint, which saves you from buying a machine to solve a problem you don’t have. In that case the next thing to test is the layer noise can’t reach: a bedtime routine that ends the day properly, or audio with words in it aimed at the twenty minutes when you are drifting and unusually willing to believe them.

Common questions

Is pink noise better than white noise for sleep?

For masking, the two are close to interchangeable, and preference for the sound matters more than the label. Pink noise has the better research record, but almost all of it comes from studies delivering precisely timed bursts rather than a continuous track. If you are simply covering street sound, pick whichever one you can stop noticing. The full masking case is covered in our guide to white noise for sleep.

Does pink noise actually improve memory?

Under laboratory conditions, in some people, modestly. Ngo and colleagues (2013) and Papalambros and colleagues (2017) both found improved recall of material learned before bed when pink-noise bursts were phase-locked to slow oscillations during deep sleep. Replication has been uneven, and effects vary with age and baseline slow-wave activity. No study has shown the same result from an ordinary overnight playlist.

What does pink noise sound like?

Steady rain, a shower running in another room, wind through trees, a river at a distance. It has none of white noise's high hiss and none of brown noise's deep rumble, which is why many people describe it as the most natural of the three. That naturalness is not a mechanism, but it does make it easier to leave on for hours without irritation.

Should I play pink noise all night or only at sleep onset?

All night is reasonable if your reason is masking, since the light transitions between sleep cycles recur several times and those are the moments a sudden sound is most likely to break. Onset-only makes more sense if you are using audio with content in it, like affirmations or a guided session, where the receptive window is the drift into sleep rather than the whole night.

What volume should pink noise be played at?

Low. Around 50 decibels at the bed, which is quiet conversation across a room or below, is enough to raise the ambient floor. Louder does not improve masking in any meaningful way and sustained loud sound increases arousal instead of reducing it. Place the speaker in the room rather than beside your head, and for infants keep the source well away from the crib.

Can I use pink noise and sleep affirmations together?

Yes, and they divide the problem cleanly. Pink noise addresses the acoustic environment and can run through the night. Sleep affirmations address what the mind is carrying and belong to the sleep-onset window. The one rule is levels: if you cannot follow every word of the affirmations over the noise, the noise is too loud and the content is being lost.