At 2 a.m., a tired parent does something almost everyone does. The baby is still fussing, the shushing isn't working, so they reach over and turn the sound machine up one more notch. The crying softens. The room feels calmer. It feels like proof that louder is better.
That feeling is worth examining, because a newborn's ears can't tell you when the dial has gone too far. Your baby has no way to say "that's too much." The only person in the room who can judge the volume is the one who is exhausted and desperate for quiet.
Why the sound works in the first place
The appeal of white noise isn't mysterious. Before birth, a baby spends months inside a surprisingly noisy place. Blood moving through vessels, the mother's heartbeat, her digestion and her muffled voice all reach the womb as a steady wash of sound. A silent bedroom is, in a sense, the unusual environment.
Steady broadband sound also does a second job. It raises the background level of a room so that sudden noises, like a door closing, a dog barking or a sibling shouting, stand out less against it. This is auditory masking: a constant sound reduces how much a sudden one contrasts with its surroundings. A sleeping baby is less likely to be jolted awake by a spike if the spike doesn't rise as far above the floor.
So the idea isn't flawed. The question is how loud the floor should be.
The number worth knowing: 50
The American Academy of Pediatrics has long recommended that hospital nurseries keep ambient noise at or below 50 A-weighted decibels. That figure was set for intensive care and newborn units, where sound exposure has been studied for its effects on sleep and development. It isn't a legal limit for your home, and it isn't a bright line between safe and harmful. But it's the most commonly cited yardstick for how loud a baby's sleeping space should be.
Fifty decibels is about the level of a quiet conversation across a room, or a running refrigerator heard from the next room over. It's softer than most people expect "white noise" to be.
A 2014 study in Pediatrics tested a set of infant sleep machines at their maximum settings. Researchers placed them at different distances from where a baby's head would be. The finding that made headlines was that all of the machines could exceed that 50 dBA guideline, and some at their loudest settings were capable of producing levels in the range associated with hearing risk for adults over long exposures. The study measured machines, not babies. It did not show that any child's hearing was harmed. What it showed was that the ceiling on these devices is far higher than the level anyone would choose on purpose.
Why hours matter as much as loudness
Hearing damage from noise isn't just about how loud. It's about dose, meaning level multiplied by time. Occupational safety guidance for adults, such as the NIOSH recommendation, uses 85 dBA for an eight-hour day as a reference point. Every 3 dB increase roughly halves the time that's considered acceptable, because 3 dB is about a doubling of sound energy.
Now think about a baby's pattern. A sound machine isn't running for eight hours. It may be running for twelve or more, covering overnight sleep and naps, for months. Adult workplace limits don't translate directly to infants. Nobody has established a precise safe dose for a baby's cochlea, and anyone who quotes you one is overreaching. But the logic of dose still applies: the longer the exposure, the lower the level you'd want it to be.
There's also a quieter question that researchers raise without claiming to have answered. Infants' ear canals are smaller than adults', which can change how sound is delivered to the eardrum, and the auditory system is still maturing. Many experts treat that uncertainty as a reason for caution, not alarm.
Distance does more than the volume dial
The single most useful variable is where the machine sits. Sound spreads out as it travels, and in open space, doubling your distance from a source lowers the level by roughly 6 dB. A room with hard walls bounces some of that energy back, so the drop is smaller in practice, but the principle holds: a machine on the crib rail delivers much more sound to the baby's ears than the same machine across the room.
This is the part many parents get backwards. A machine clipped to the crib or tucked beside the mattress, set to a low-feeling volume, may deliver more energy to the baby's head than a machine on a dresser, set one notch higher.
The practical lesson is that you can lower the dose without lowering the benefit. Move it farther away and the baby gets the masking effect of a steady room-level hum without sitting inside the loudest part of the speaker's field.
What "sounds fine" tells you, and what it doesn't
Your own judgment of loudness isn't a reliable gauge here, for three reasons.
First, you're adapted. After a few minutes your brain stops registering a constant sound, a process called auditory habituation, so a machine that seemed loud when you switched it on can feel like "nothing" by the time you leave the room.
Second, you're standing in a different spot. The level at your standing height and distance from the door is not the level at the mattress.
Third, loudness perception depends on frequency. Some machines emphasize bright, hissy sound, others a deep rumble. Two devices that sound equally loud to you can have different measured levels, and the meter, not the ear, is what tells you how much energy is arriving.
The fix is to measure instead of guess.
Your next moves
- Measure at the mattress, not at the door. Download a free sound level meter app that reports A-weighted decibels, hold your phone where the baby's head would rest, and run the machine at its usual setting. Phone microphones aren't calibrated instruments, so treat the reading as a good estimate. If it's well above the mid-50s, turn it down or move the machine.
- Move the machine as far from the crib as the room allows. Put it across the room on a dresser or shelf, not on the crib rail or mattress. Then re-measure. You'll often see a clear drop with no change to the dial.
- Choose a soft setting, not a loud one, and keep it consistent. The goal is a gentle floor, not a wall. If a quiet level only works for a week before you feel pressure to raise it, that's the cue to adjust the routine, not the volume.
- Use the timer if your device has one. Many babies mainly need the sound while falling asleep. If yours does, set it to switch off after a stretch instead of running all night, which cuts the total dose.
- Mark the safe volume with tape. Once you've found a level that measures comfortably, put a dot of tape at the dial position so a midnight hand doesn't creep upward.
- Mention it at the next pediatric visit. If you notice your child doesn't startle to sudden sounds, doesn't turn toward voices, or seems slow to babble, tell the pediatrician. Hearing is typically checked in newborn screening, and concerns later on are worth raising and following up.
The larger habit hiding in this question
It's easy to think of hearing safety as something that begins with headphones and ends with concerts. In fact the habits that shape a lifetime of listening start early, and they're mostly set by adults who never thought to measure anything. The same instinct that makes you check a bath's temperature with your wrist can be pointed at a room's sound level with your phone.
It also holds for your own ears. The sound floor you build around a baby, you probably also build around yourself: the podcast turned up over traffic, the show left on to fill silence, the earbuds nudged one click higher on a noisy train. Most of us have never actually seen our own exposure as a number.
That's the thread we care about at Audra. Audra is a free at-home hearing screening that plays pure tones through your phone, and it's built to help you notice how you hear over time instead of guessing. It isn't a medical device and it can't diagnose anything, and it has nothing to say about an infant's hearing, which needs a pediatric professional. But for the adults in the house, a baseline of your own hearing, taken on a calm evening after the nursery is finally quiet, is a small, useful way to start paying attention. If you'd like to try it, it's at https://audra.lumenlabs.works.