how accurate are snoring apps? an honest look at how they work

It’s a fair thing to be suspicious of. An app claims you snored for 47 minutes and peaked at 64 decibels, and you slept through every second of it — so how would you ever know if that’s true? The honest answer is that a phone is genuinely good at some of this and genuinely can’t do other parts at all. The useful thing is knowing which is which.

what a phone actually measures

A snoring app has one sensor that matters: the microphone. It listens to the room for the acoustic signature of snoring — a low, rhythmic, repetitive sound that rises and falls with your breathing — and separates it from silence, traffic, a fan, or the person next to you shifting under the duvet.

Two things it reads well:

• Loudness. A phone microphone measures sound pressure, which maps to decibels. It isn’t a calibrated meter, so the absolute number can be off by a few dB depending on distance and surface — but if you record the same way each night, the comparison holds. A peak that climbs from 58 to 68 dB over a month is a real change, whatever the true absolute value.
• When and how much. Picking snoring out of silence and timestamping it is something software does reliably. “You snored in three stretches, the longest from 2 to 3 a.m.” is well within what audio can tell you.

So for the questions most people actually have — do I snore, how loudly, how often, and is it changing? — yes, an app is accurate enough to be useful, especially as a relative measure over time.

where accuracy gets fuzzy

Now the honest limits.

• The absolute decibel number is approximate. Your phone isn’t a lab instrument. Treat the dB as a reliable relative figure — great for comparing your own nights — not a clinical measurement you’d quote to three decimal places.
• It can mistake other sounds for snoring, and the reverse. A heavy fan, a partner, a passing truck — a good app filters most of it, but no detector is perfect. This is exactly why being able to play back the loud moments matters: your own ears settle any case the algorithm got wrong.
• It hears one room, not two people. If your partner snores too, that’s on the recording. No app reliably tells two sleepers apart.

None of this makes the data useless. It makes it what it is: a solid, consistent, comparable record of the sound of your nights — not a medical instrument.

the line no honest app should cross

Here’s the part to be genuinely wary of: any app that claims to detect sleep apnea from audio alone. Apnea is defined by breathing events — your airway closing, your blood oxygen dropping — and a microphone measures none of that. It can’t see your airflow or your oxygen, so it cannot diagnose or rule out apnea, full stop. Anything promising a diagnosis from sound is overselling, and that’s the moment to trust the app less, not more.

What audio can do is catch the pattern — long silences followed by gasping restarts — that’s worth showing a doctor. That’s a prompt, not a diagnosis. Snoring and sleep apnea aren’t the same thing, and the honest role of an app is to help you notice when it’s time to ask a professional.

how we think about it

We’d rather undersell than overclaim. SnoreWise detects snoring by amplitude and frequency, gives you your peak in decibels with a timestamp, and keeps the loud clips so you can hear them yourself — and it says plainly that it’s not a diagnosis. The accuracy that matters here isn’t a percentage on a marketing page; it’s whether the same method, used the same way every night, gives you a trend you can trust. That it does.

If you want to test it yourself, that’s the right instinct: record a few nights, listen back to a loud stretch, and check whether what you hear matches what the summary said. It usually does — and where it doesn’t, you’ll know exactly how much salt to take with the number.