Calming Sounds for Anxiety: What to Play, and Why.
When anxiety is up, silence is loud: every creak and voice and engine gets checked, and the checking is the problem. A steady, featureless sound gives the ear something that will not change. This page puts one within reach — pink or brown noise with three honest starting points, a slow beat if you want one, and a breathing pacer, which has the better evidence — and says plainly what any of it can and cannot do.
Read this first: the sound on this page is an aid for the moment you are in. It is not a treatment for anxiety.
Steady sound can take the edge off a bad half-hour. It does nothing to the pattern that keeps producing bad half-hours. The approaches with the strongest trial evidence for anxiety disorders are cognitive behavioural therapy and, where a doctor advises it, medication; if anxiety is shaping your days, a GP or a psychologist is the place to start. Never stop or change a prescribed medication without talking to the doctor who prescribed it.
If you are in the US and in crisis, call or text 988. If you have chest pain, trouble breathing, or a feeling of pressure that does not ease when you sit down, call emergency services — a panic attack and a heart problem can feel alike, and only a clinician can tell them apart.
Nothing you choose here is saved. Audio starts only when you press Play and stops after twenty minutes at most.
Three places to start
Pink noise, volume 45. Like steady rain: it covers the small, unpredictable sounds that an anxious ear keeps checking, without the hiss of white. Pick a starting point above for the reasoning, or choose a colour directly.
Or pick a colour
No headphones needed for noise — a phone speaker on the desk or the bedside table is fine, and for sleep it is safer than earbuds. Keep the level where you could still hold a conversation: the aim is to soften a room, not to fill it with something louder. The audio is a four-second loop generated in your browser, so nothing streams, nothing is uploaded and nothing is remembered between visits.
Why a steady sound can help in the moment.
One part of this is physics, and two parts are reasoning. It is worth keeping them apart.
The physics is masking. A steady noise with energy across the range where voices, doors and traffic live raises the floor under those sounds, so their peaks stand less far above their surroundings and each one is harder to pick out. That is the signal-to-noise ratio, and it is not in dispute; it is why a fan covers a neighbour and why rain makes a street go quiet. On an anxious evening, much of what reaches you is small and intermittent — a floorboard, a car slowing, a voice through a wall — and a bed of pink or brown noise simply takes those below the level at which they register. Nothing about anxiety has changed. There are fewer things to react to.
The first piece of reasoning is about attention. Anxiety narrows attention onto threat and keeps it there; the ear scans, finds something, and the body answers before you have decided whether it mattered. A sound that is always the same gives the scanning nothing to land on. Rain does not have a sudden bit. People describe this as the sound “holding” the room, and it is a fair description of what a continuous, featureless signal does to a system that is looking for change. It is not something a trial has isolated and measured, and this page does not pretend otherwise.
The second is predictability. A sound you chose, at a level you set, that will stop when you told it to, is one of the few things in a bad hour that is entirely under your control. That matters more than it sounds. Much of what makes anxiety exhausting is the sense that things are happening to you; a timer you set and a volume you pick are small acts of the opposite. Again: reasoning, not a finding, and the honest version of why people reach for a familiar sound rather than a new one.
What none of this does is touch the thought that started the loop, or the pattern that keeps starting it. Masking is applied to the room, not to you. If you want the thing with trial evidence for the loop itself, it is on the breathing page and in a therapist’s office, and the next section says so with the numbers attached.
What the trials actually show.
Four bodies of evidence are relevant, and they rank in a particular order: slow breathing, music, nature sounds, binaural beats. The one this site is named after comes last. Every paper below was checked against PubMed on the date at the top of this page, and each link goes to its record.
Slow breathing
Zaccaro and colleagues (Frontiers in Human Neuroscience, 2018) reviewed the studies on breathing at fewer than ten breaths a minute in healthy people and found a consistent picture across the fifteen that met their criteria: slower breathing shifted the balance of the autonomic nervous system toward its parasympathetic side, changed the rhythm of the heart in step with the breath, and was accompanied by reports of greater comfort, relaxation and alertness and less anxiety and anger. Laborde and colleagues (Neuroscience & Biobehavioral Reviews, 2022) then pooled 223 studies of voluntary slow breathing and found that the heart-rate variability measure which indexes that parasympathetic activity rose during a session, immediately after one, and after a course of them. Neither paper is a trial of breathing as a treatment for an anxiety disorder, and heart-rate variability is a physiological marker rather than a feeling. But it is the best-replicated effect on this page, it costs nothing, and it is why the pacer is linked from inside the player rather than buried below it. The site’s own pages on breathing exercises for anxiety and resonance breathing go through the rate and the ratio.
Music
The strongest evidence for sound and anxiety is for music, and most of it comes from one setting: people waiting for surgery. Bradt, Dileo and Shim (Cochrane Database of Systematic Reviews, 2013) is a Cochrane review of 26 randomised trials with 2,051 participants in which patients listened to recorded music before an operation. On the state scale of the State-Trait Anxiety Inventory, the music groups came out several points lower than standard care, with a small accompanying drop in heart rate and diastolic blood pressure. The authors are blunt about the weakness: almost every trial was at high risk of bias because a person cannot be blinded to whether they are listening to music, and anxiety was self-reported. The effect is real as far as it goes, and it goes as far as a waiting room.
Nature sounds
Buxton and colleagues (Proceedings of the National Academy of Sciences, 2021) pulled together 36 publications on the health effects of natural sound — water, wind, birds — and ran meta-analyses on 18 of them. Across those, exposure to natural sound was associated with lower stress and annoyance, and with better health and mood outcomes. Two cautions, both from the paper itself. The studies are heterogeneous, mixing laboratory playback with time spent outdoors, so the pooled figure covers very different experiences. And the confidence interval on the health-and-mood result is wide enough to run from almost nothing to very large, which is what a small, varied evidence base looks like. It supports “rain and water are a reasonable thing to reach for”; it does not support a specific claim about anxiety.
Binaural beats
Garcia-Argibay, Santed and Reales (Psychological Research, 2019) is the meta-analysis this site uses as its anchor. Twenty-two studies, 35 effect sizes, an overall effect of medium size across cognition, anxiety and pain together, with anxiety one of the four outcomes pooled. Longer exposure and exposure before a task rather than during it did better. The trials are small, the participants mostly students, and the comparison conditions vary; the fair summary for anxiety is a small-to-moderate effect on state anxiety, the kind you have right now, with nothing to say about trait anxiety or about any diagnosed condition. The one well-designed clinical trial is Padmanabhan, Hildreth and Laws (Anaesthesia, 2005), which randomised day-surgery patients to binaural-beat audio, the same soundtrack without the beats, or nothing, and measured anxiety on the same state scale as the Cochrane review. The binaural group’s score fell by clearly more than either comparison group’s. It is a single trial in one hospital, and it has not been replicated at that quality since. Whether that is encouraging or damning depends on your temperament; the honest reading is that a slow beat under the noise is a reasonable thing to try with headphones on, and that the noise and the breathing are carrying more of the load. Safety is its own page.
What is missing
No trial compares pink noise with silence for people who are anxious. No trial compares noise colours against each other for anxiety. The sleep literature on noise is small, short and inconsistent, and continuous overnight noise at higher volumes is a hearing consideration rather than a benefit. If you read elsewhere that a colour of noise reduces anxiety by a given amount, the number was invented; this site published figures like that once, and the corrections page lists them.
Pink or brown, and how loud.
The player offers two colours on purpose. White noise is the one most people have heard of and the one most people stop using, because its flat spectrum sounds bright and hissy and tires the ear over an hour. The two that remain cover nearly every calming use.
- Pink falls off gently with frequency and sounds like steady rain. It covers voices, doors and traffic, which sit in the middle of the range, without the harshness of white. It is the default here and the better choice in daylight. The pink noise page has the spectrum and the sleep evidence.
- Brown falls off more steeply and sounds like surf or a far-off waterfall. Its energy sits low, which is why people describe it as settling rather than airy, and why it is the usual choice for night and for anyone who finds pink too thin. The brown noise page goes further.
The volume rule is the same for both: you should still be able to talk over it without raising your voice. The instinct when anxious is to turn the sound up until the room disappears, and the result is a room that is now louder than it was. The player’s three starting points sit at 15, 40 and 45 out of 100 for that reason; the lowest one exists because for many people the useful amount of noise is barely audible.
Speakers are fine. Noise is not a binaural effect, so it does not need to reach each ear separately, and a phone on the bedside table is more comfortable and safer than earbuds through the night. Headphones only become necessary if you add the beat.
What does not help.
Louder. Masking depends on covering the sounds you are reacting to, not on volume as such. Past the level that blurs them, every increase is noise you are now sitting inside, and at sleep volumes across a whole night it becomes a hearing question. If you keep reaching for the slider, the problem is not the slider.
Longer. The player stops at twenty minutes and does not offer more. A sound that is always on becomes the room’s default, and silence — which you will meet again — becomes the thing that feels wrong. Use it for the hour that needs it; let it stop.
Waiting for it to work. Listening for an effect is a form of scanning, and scanning is what the sound was supposed to interrupt. Put it on, set the timer, and do something else with your attention: the breathing pacer, a task, the next page of a book.
Using it instead of the thing it is meant to sit under. Noise is a bed. The evidence-backed activity — slow breathing, or the exercise a therapist gave you — is what goes on top of it. A bed with nothing on it is fine for a bad night and not a plan.
Choosing sound over sleep hygiene, movement, or talking to someone. None of those is on this page because none of them is a sound. All of them have better evidence than anything that is.
What this page has, and what the app has.
The web tool is deliberately small. Two noise colours here, five on the noise generator; a binaural, isochronic or pure tone with a white, pink or brown bed on the main generator; a twenty-minute cap on everything; no account and nothing stored. All of it is generated in the browser from a few seconds of buffer, so it loads on a bad connection and costs no data. If what you need is twenty minutes of steady sound and a breathing pacer, you do not need anything else.
The mobile app is where the recorded ambience lives: 34 ambient sounds — rain, ocean, campfire, café, fan and the rest — across 44 presets, with no session cap, so it can run through a night. That is a different product for a different use, and this page does not pretend the browser version is it. Neither one is a treatment, and the app’s store listing says so in the same words.
Things people reasonably ask.
What is the best sound for anxiety?
There is no trial that ranks sounds for anxiety against each other, so the honest answer is the steady, featureless one you can tolerate at a low level. For most people that is pink noise in the day and brown at night. Music has better evidence than noise in the one setting it has been tested properly, people waiting for surgery, and slow breathing has better evidence than either.
Does pink noise help with anxiety?
Nobody has run the trial. What pink noise demonstrably does is mask the small intermittent sounds that an anxious ear keeps checking, which removes things to react to; whether that lowers anxiety on a scale has not been measured. Treat it as a way to make a room easier to be in, not as something that acts on anxiety itself.
Is brown noise or pink noise better for anxiety?
Neither has been tested for anxiety, so choose by ear. Brown is deeper, sits lower and tires people least over a long stretch, which makes it the usual choice at night. Pink covers voices and mid-range sounds better and is the usual choice in daylight. Try each for a minute at low volume and keep the one you stop noticing.
Do binaural beats help with anxiety?
The 2019 meta-analysis by Garcia-Argibay and colleagues found a small-to-moderate effect on state anxiety, the kind you feel right now, across mostly small studies. One randomised trial in day-surgery patients found a clearly larger drop in anxiety with binaural audio than with the same audio without the beats. That is a reasonable basis for trying a slow beat under the noise with headphones on, and not a basis for calling it a treatment. The binaural beats for anxiety page goes through the evidence.
What should I do if a panic attack is starting?
Lengthen the out-breath and keep doing it; the panic attack page gives the minutes in order and the breathing pacer gives you something to follow. Sound is optional background. If there is chest pain, trouble breathing or pressure that does not ease when you sit down, call emergency services, because a panic attack and a heart problem can feel the same and only a clinician can tell them apart.
Can I use these sounds instead of medication or therapy?
No. The sound on this page is an aid for a bad half-hour. The approaches with trial evidence for anxiety disorders are cognitive behavioural therapy and, where a doctor advises it, medication. Never stop or change a prescribed medication without the doctor who prescribed it. If anxiety is shaping your days, a GP or a psychologist is where to start.
Do I need headphones?
Not for noise. It is not a binaural effect, so a phone speaker on a desk or a bedside table works and is safer for sleep than earbuds. Headphones are needed only if you add a binaural beat, which is why the link to the beat says so.
Is anything on this page recorded or stored?
No. The noise is generated in your browser from a four-second loop, so nothing streams and nothing is uploaded. Which starting point or colour you pick is not saved and not sent to analytics. Audio starts only when you press Play and stops after twenty minutes at most.