Every guide to white noise says some version of the same thing: keep it around 50 dB. The number is repeated so consistently across sleep advice articles and product pages that it has taken on the weight of settled fact. But here's what those guides leave out — 50 dB is the ceiling, not the target. It's the maximum level the American Academy of Pediatrics permits for infant sound machines, established to prevent hearing damage. It was never intended as the recommended volume for a good night's sleep.
The distinction matters because the number you actually want — the level where white noise provides genuine sleep benefits with minimal disruption to your brain's natural cycling — sits considerably lower. And reaching it has less to do with finding the right dial position than with understanding that the most important variable isn't the device setting at all. It's where you put the device.
This article separates three fundamentally different thresholds that most sources conflate, explains the acoustic physics that determine what your ears actually receive versus what the device emits, and gives you a practical framework for calibrating your bedroom — not to a generic number, but to your specific environment.
The Problem with "Just Keep It Low"
The advice to keep white noise "low" is not wrong. It's just not useful in the form it's usually given. Without knowing what you're measuring, where you're measuring it, and what you're comparing it to, "low" is a concept that most people translate as "not so loud that it bothers me" — which can still be significantly louder than what the research identifies as beneficial.
The core issue is that there are three separate noise thresholds in the sleep-and-sound conversation, and they address three completely different concerns. One is about permanent hearing damage. One is about sleep fragmentation. One is about deep sleep optimization. These thresholds are set by different bodies of research, measured under different conditions, and apply to entirely different physiological phenomena. When they all get collapsed into a single "keep it around 50 dB" instruction, the result is a number that is simultaneously too high for ideal sleep quality and too low for hearing protection concerns — satisfying neither purpose precisely.
Understanding each threshold separately is what allows you to stop guessing and start calibrating.
Three Different Thresholds — and Why People Mix Them Up
The hearing safety threshold: 70–85 dB
The National Institute for Occupational Safety and Health (NIOSH) and the CDC identify sustained exposure above approximately 70–80 dB over an eight-hour period as the zone where cumulative noise-induced hearing loss begins to accumulate. This is the threshold that governs workplace noise regulations, limits on concert venues, and the warnings on power tools. It has essentially nothing to do with whether your white noise is helping you sleep well. The fact that your sound machine is well below industrial noise levels doesn't tell you anything meaningful about its effect on your sleep architecture.
This threshold is relevant to one specific concern: people who run their sound machines at maximum volume directly next to their bed. A 2014 study from the American Academy of Pediatrics tested 14 commercially available infant white noise machines and found that all of them exceeded safe volume limits at maximum settings, with some reaching 85 dB at close range — genuinely into the hearing-damage zone. For most adults using a machine at moderate settings, this is not the relevant concern. But it matters for anyone who compensates for a loud environment by simply cranking up the volume.
The sleep fragmentation threshold: above 40 dB ambient
The World Health Organization's Night Noise Guidelines for Europe identify 40 dB as the approximate threshold above which environmental noise begins to cause measurable biological effects — increased sleep fragmentation, elevated cortisol, reduced time in deeper sleep stages. Below 40 dB, the sleeping brain generally processes ambient sound without triggering arousal responses. Above it, particularly with unpredictable or fluctuating noise, the brain increasingly registers the acoustic environment as something to evaluate rather than ignore.
This threshold is the most directly relevant for white noise use. If your sound machine is pushing your bedroom ambient level above 40 dB, you may be solving the problem of external disruptions while introducing a different problem: a sustained auditory load that keeps the brain in a lighter state of sleep processing. The masking benefit and the arousal cost start to cancel each other out.
The deep sleep optimization zone: 30–35 dB ambient
Research published in Sleep Medicine Reviews points to an ambient bedroom level of approximately 30–35 dB as the range where slow-wave sleep — the deepest, most physically restorative stage — proceeds with minimal acoustic interference. At this level, the brain's auditory system is receiving enough consistent input to ignore environmental fluctuations, but not enough to constitute a meaningful processing load. This is the zone where white noise functions as it should: a seamless acoustic backdrop rather than a noticeable presence.
Achieving a 30–35 dB ambient level in most real-world bedrooms is not difficult. In a room with moderate background quiet (a typical suburban interior at night runs around 25–30 dB), adding white noise at this level is a modest uplift that accomplishes the masking goal without pushing into fragmentation territory.
| Threshold | Level | Source | What it governs |
|---|---|---|---|
| Hearing safety | 70–85 dB | NIOSH / CDC | Permanent hearing damage risk over 8 hours |
| Sleep fragmentation | above 40 dB ambient | WHO Night Noise Guidelines | When noise begins disrupting sleep stages |
| Deep sleep optimization | 30–35 dB ambient | Sleep Medicine Reviews | Level where slow-wave sleep is least disrupted |
Why the Number on Your Device Means Almost Nothing
The most overlooked variable in all of this is not volume — it's distance. The number you set on your device determines what comes out of the speaker. What matters for your sleep is what arrives at your ear. These are not the same number, and the gap between them is determined almost entirely by how far you place the device from your head.
The inverse square law in plain language
Sound intensity follows a simple physical principle: every time you double the distance from a sound source, the sound pressure level at the listener drops by approximately 6 dB. This is not a setting you can adjust or a feature of any particular device — it is how sound propagates through air. The implication for bedroom sound machines is significant and predictable.
A machine placed on your nightstand at 30 centimeters from your head delivers a substantially different acoustic dose than the same machine producing the same output placed across the room. The device hasn't changed. The volume dial hasn't moved. But what your ears and your sleeping brain receive has changed dramatically.
What your white noise actually sounds like at your pillow
Applying the inverse square law to the AAP's 2014 measurements — which found that infant noise machines commonly output around 65–70 dB at close range at their medium-to-high settings — produces a practical table for understanding what different placements actually deliver:
| Distance from device | Approx. SPL drop | Estimated level at ear | Assessment |
|---|---|---|---|
| 30 cm (nightstand) | — | ~65 dB | Too loud — above fragmentation threshold |
| 60 cm | −6 dB | ~59 dB | Still above recommended range |
| 120 cm | −12 dB | ~53 dB | At the AAP ceiling — not the target |
| 200 cm (AAP minimum) | −18 dB | ~47 dB | Approaching optimal range |
| 300 cm (far wall) | −22 dB | ~43 dB | Good — within masking range with low load |
These figures are approximations based on free-field acoustic propagation. Real bedrooms have walls, soft furnishings, and absorption that complicate the math — but the directional truth holds: distance reduces the acoustic dose at your ear, predictably and without requiring you to touch the volume control. Moving a device from your nightstand to the far side of a typical bedroom typically reduces the SPL at your head by 18–22 dB, which is the difference between a level that works against deep sleep and one that supports it.
Why the nightstand is working against you
The nightstand placement is intuitively appealing — close to you means you can hear it, which feels like it should be more effective. But it inverts the entire logic of sound masking. The goal is not for the white noise to be loudly audible to you. The goal is for it to raise your bedroom's acoustic floor just enough that external disruptions lose their contrast against the background. That works at low ambient levels. It also works — at lower cost to your sleep architecture — when the device is at a moderate distance across the room.
A second problem with the nightstand placement is asymmetry: the white noise source is much closer to your head than the disruptive sounds you're trying to mask, which typically enter from windows, walls, or hallways. This means you end up with a loud close sound competing with a moderately loud distant sound, rather than a consistent ambient field surrounding you. The acoustic result is less effective masking at more cost. Place the device at a distance to create a more uniform ambient field.
Start with maximum distance — the far wall, dresser, or shelf on the opposite side of the room. Then bring the volume up from zero until you notice the masking effect on background sounds. That level is your baseline. The device should still be audible from across the room, but it should not feel like a dominant presence in the space.
If your room is small and "across the room" means 150 cm, place the device in a corner directed away from the bed. Reflective surfaces will diffuse the sound, effectively increasing the acoustic distance.
The SNR Principle: Matching Your Volume to Your Room
The single most useful reframe in this entire article is this: the right volume for white noise is not an absolute number — it is a relationship. Specifically, it is the relationship between your white noise level and the ambient noise level of your bedroom environment.
How the brain decides whether to wake up
Your sleeping brain does not respond to loud sounds per se — it responds to acoustic events. An acoustic event is a sudden change in the sound environment, what researchers call the acoustic startle response. The brain's arousal threshold during sleep is calibrated not to absolute volume but to the contrast between the baseline and a new sound. This is why a parent can sleep through ordinary household noise but wake instantly to their baby's cry — the cry is acoustically distinctive, not just loud.
A car alarm sounding at 65 dB against a background of 20 dB silence represents a 45 dB jump — a large contrast that the sleeping brain almost always registers as a threat. The same car alarm sounding against a background of 42 dB white noise represents a 23 dB jump — below the typical arousal threshold for most people in lighter sleep stages. The white noise has not blocked the sound. It has reduced the contrast, and that is what matters.
How to find the minimum effective volume for your bedroom
This approach treats white noise as a calibration tool rather than a comfort setting. The process has two steps.
First, measure your bedroom's actual ambient noise level at nighttime. Free apps provide a reasonable approximation for this purpose — the NIOSH Sound Level Meter app (iOS) is a notable option developed by the National Institute for Occupational Safety and Health itself. Consumer-grade smartphone apps have a margin of error of roughly ±2–5 dB depending on the device's microphone quality, which is acceptable for this calibration task. Measure on several different nights to get a representative reading. Most quiet suburban or urban bedrooms register between 25 and 40 dB at night, with the variation driven primarily by traffic patterns, HVAC systems, and building noise.
Second, set your white noise to approximately 8–10 dB above that baseline reading. This is the minimum level needed to meaningfully reduce the contrast effect of typical environmental disruptions. If your bedroom ambient is 28 dB, you're aiming for roughly 36–38 dB at your position — achievable with a device at moderate settings placed at the far side of the room. You do not need to go higher unless the disruptive sounds you're masking are significantly louder than typical.
What to do if real noise is genuinely loud
For bedrooms that face genuine acoustic challenges — a busy road, upstairs neighbors, bass from adjacent apartments — the masking approach has limits. White noise can narrow the contrast between your acoustic floor and a disruptive event, but it cannot make a 75 dB intrusion disappear behind a 40 dB background. The gap is simply too large.
In this scenario, there are two options, and the better strategy combines both. The first is to address the source: acoustic curtains, door sweeps, and window seals can reduce incoming noise by 10–20 dB before it enters the room, shifting the problem to a solvable scale. The second is to accept that some disruptions in genuinely loud environments may require a slightly higher white noise floor — in the 45–48 dB range — but to achieve that through careful distance placement rather than by placing a loud device close to your head. The goal is always to use the minimum effective level at the maximum effective distance.
What the 2026 Penn Medicine Study Actually Found — and What It Didn't
The Penn Medicine study published in Sleep in February 2026 found that continuous pink noise at 50 dB reduced REM sleep by nearly 19 minutes per night in healthy adults over seven nights. The coverage of this finding generated significant concern about the safety of sound machines in general, and white noise in particular — since both are broadband sounds subject to the same underlying mechanisms.
The finding is important. But understanding what the study actually tested, and what it did not, is essential for applying it correctly to your own practice.
What the study tested and what it didn't
The study used 25 healthy adults in a controlled laboratory setting, playing broadband noise at 50 dB for seven consecutive nights. It did not test volumes below 50 dB. It did not test subjects who regularly use sound machines and may have adapted to the auditory environment. It did not test the effect of proper placement — whether the sound was played from a nearby source or from across the room.
50 dB is the AAP's ceiling for infant machines — a threshold designed to prevent hearing damage, not an endorsement of that level as comfortable for sleep. Most experienced sound machine users report listening at levels well below that: a gentle ambient presence rather than a clearly audible sound. The study provides no data on what happens at 35 or 40 dB, which is the range that careful placement and moderate settings actually deliver.
Why 50 dB is the ceiling, not the target
Dr. Mathias Basner, the study's lead researcher, did not recommend eliminating broadband noise use. His recommendation was explicit: anyone who wants to continue using sound during sleep should do so at the lowest level that still produces the desired effect. That is precisely the calibration framework this article describes. The study's contribution is not "stop using white noise" — it is "stop assuming more volume is better, and stop placing devices where the volume at your ear is higher than you think."
The 2026 findings are a strong argument for the SNR approach: find the minimum effective level for your specific room, achieve it through distance rather than high output, and use audio designed for extended low-volume playback. At those parameters, the research does not indicate harm — and the masking benefits remain intact. For a broader look at how this study fits the existing research landscape, see our detailed guide to white noise sleep benefits.
Practical Setup — Getting the Volume Right
The principles above translate to a concrete setup process. Here is how to implement them in a standard bedroom.
Distance first, volume second
Begin by choosing your placement. The far side of the room — a dresser, a shelf, the windowsill on the wall opposite your bed — is almost always the best option. If your bedroom is genuinely small and "across the room" is under 150 cm, point the device into a corner or toward the ceiling to encourage sound diffusion rather than direct transmission toward the bed. Aim for at least 200 cm of distance in any direction from your head. This placement will typically reduce the SPL at your ear by 18–22 dB compared to a nightstand placement without changing anything else.
How to verify your actual bedroom level without professional tools
Once the device is placed, open a free sound level meter app. The NIOSH SLM (available on iOS) was developed specifically for occupational noise measurement and provides readings that are more calibrated than general-purpose apps, though any app can give a useful approximation for bedroom calibration. Hold the phone at roughly pillow height at your sleeping position, start the white noise, and read the ambient level. Aim for a reading between 35 and 42 dB. If you're above 45 dB, reduce the volume on the device. If the masking effect disappears entirely below 35 dB — meaning you can now clearly hear the sounds you were trying to mask — raise it slightly.
The "whisper test" is a quick informal alternative: if you can hold a conversation at normal speaking volume while the white noise plays and still hear your own words clearly, the volume is probably in a functional range. If the white noise drowns out speech at conversational volume, it is likely too loud.
All night vs. timer: what actually matters
The most important factor determining whether all-night white noise is beneficial or costly to your sleep is volume, not duration. The concern raised by the Penn Medicine 2026 study is a volume concern — it operated at 50 dB throughout the night. At the lower ambient levels produced by proper placement and calibrated settings, continuous playback provides masking protection throughout your full sleep cycle, including during the transitions between cycles when you are most vulnerable to acoustic disruption.
If your primary sleep challenge is falling asleep rather than staying asleep, and your bedroom is quiet after midnight, a timer that runs for the first 60–90 minutes after bedtime is a reasonable option. But the framing matters: if you are keeping the volume appropriate and the device at a proper distance, duration is a secondary consideration. The goal is a gentle, consistent acoustic field — not something that needs to be switched off to allow your brain to rest.
A note on speaker quality
The quality of your playback device has a modest but real effect on how the sound behaves, particularly for noise colors with significant low-frequency content. The integrated speakers in many compact noise machines are optimized for the mid-range frequencies where white noise is most audible, but they can struggle to reproduce the deeper bass content of brown or pink noise accurately. This sometimes leads users to raise the volume to compensate for perceived thinness — inadvertently driving up the SPL without improving the masking quality. Dedicated noise machines with wider frequency response, or a small speaker with reasonable bass extension placed at distance, tend to deliver more accurate reproduction at the levels that matter.
Featured audio partner The Blackout Room produces 10-hour white noise sessions specifically calibrated for low-volume, all-night use. Place the device at distance, start at a conservative volume, and let the track do its work:
Does Noise Color Change the Volume Rules?
The fundamental principles — distance first, minimum effective level, below the fragmentation threshold — apply equally to all noise colors. But there is one practical nuance worth noting.
Lower-frequency sounds like brown noise are perceived as softer at the same SPL because of the ear's frequency sensitivity curve, which peaks in the 2,000–5,000 Hz range. A brown noise track playing at 40 dB SPL often feels quieter than white noise at the same measurement because less of its energy falls in the range where human hearing is most acute. This is a feature — it means brown noise tends to be less fatiguing over extended listening. The caution is the opposite: because it feels quieter, there is a natural tendency to raise the volume to "feel like it's working," which can push the actual SPL higher than intended. Measure with an app rather than relying on subjective loudness perception when setting up brown or pink noise.
For a full comparison of how different noise colors perform for sleep — including which one to choose based on the type of disruption you're masking — see our guides on the best noise color for sleep and brown noise vs white noise. For specific guidance on the REM sleep findings and their implications for different colors, see our article on pink noise sleep benefits.