The best sound meter apps, on the right phones, can read within about 2 dB of a professional meter for everyday noise. NIOSH’s own app is tested to within ±2 dBA on iPhones. Many apps do much worse, results vary from one phone model to the next, and even a good app on an uncalibrated phone gives you an estimate rather than a certified measurement.
That’s accurate enough to tell a 70 dB room from an 85 dB one, which is what most people need. It isn’t accurate enough for workplace compliance or legal evidence.
What did the NIOSH study of sound meter apps find? #
The best-known research comes from Chucri Kardous and Peter Shaw at the National Institute for Occupational Safety and Health (NIOSH), who tested apps against a reference-grade sound measurement system in their acoustics lab.
The 2014 study (Kardous and Shaw, Journal of the Acoustical Society of America) screened more than 130 iOS apps and found only 10 that met their selection criteria for testing. Only 4 of 62 Android apps were tested. The two best apps had mean differences of 0.07 dB (unweighted) and -0.52 dB (A-weighted) from the reference, and two more were within ±2 dB. In other words, a handful of apps were very good, and most weren’t worth testing.
The 2016 follow-up (Kardous and Shaw) repeated the exercise with external calibrated microphones plugged into the phones. Measurements came within ±1 dB of the reference, and the authors concluded that external calibrated microphones greatly improve accuracy and precision.
A separate 2016 study (Roberts, Kardous and Neitzel) tested combinations of microphones and apps with pink noise from 60 to 100 dBA. With external microphones, certain apps did not differ significantly from a Type 1 (precision) measurement system.
NIOSH then built its own free app, the NIOSH Sound Level Meter. NIOSH says it tested the app in a reverberant chamber and found it accurate within ±2 dBA, and that with a calibrated external microphone it meets the Type 2 requirements of the IEC 61672-3 sound level meter standard. The app is iOS only. NIOSH’s stated reason is that the Android device market is too fragmented across manufacturers to verify.
Why are phone sound meters less accurate than real ones? #
A professional sound level meter is built around a measurement microphone with a known, flat response, and it’s checked against an acoustic calibrator. A phone microphone was designed for phone calls. The gap comes from a few places:
- Every model hears differently. Mic capsules, their placement and the openings in the case vary across phones, so the same app can read a few dB apart on two devices.
- Voice processing. Phones apply automatic gain, noise suppression and filtering to make calls sound good. Those help a conversation and hurt a measurement. Good meter apps try to switch them off where the operating system allows. On iOS, for example, an app can request Apple’s measurement audio mode, which disables automatic gain.
- Frequency response. Phone mics tend to roll off deep bass and can be uneven in the treble, so a rumbling truck and a hissing fan may be misjudged in different directions.
- The ends of the scale. Very quiet rooms can fall below what the mic can resolve over its own electronic noise, and very loud sources can overload the mic or the audio chain. Readings near the extremes deserve extra doubt.
- The scale itself. A phone reports a digital level, not pressure. The app has to translate that into dB SPL, and unless that translation is calibrated for your exact phone, it’s an educated assumption.
How accurate does a reading need to be? #
Whether a phone is “accurate enough” depends on what you’ll do with the number.
| What you’re doing | Accuracy you need | Phone app good enough? |
|---|---|---|
| Checking if a concert or gym class is dangerously loud | A few dB | Yes |
| Comparing two rooms, two appliances, before and after soundproofing | Consistency more than accuracy | Yes, with the same phone in the same spot |
| Keeping a log of a noisy neighbor | A few dB, plus timestamps | Yes, as supporting notes |
| Workplace noise compliance | Type 2 (about ±2 dBA) or better | No, use a certified meter or dosimeter |
| Legal disputes, ordinances, engineering sign-off | Class 1 or Class 2, calibrated | No |
For compliance work, readings from a Type 2 meter are generally treated as accurate to about ±2 dBA and Type 1 to about ±1 dBA. We explain the classes in Class 1 vs Class 2 sound level meters, and when a phone is fine versus when to buy hardware in sound meter app vs SPL meter.
Relative readings are more trustworthy than absolute ones. If the same phone in the same spot reads 8 dB lower after you close the window, the window cut the noise by roughly 8 dB, even if the phone’s absolute reading is off by 3.
How can you get more accurate readings from your phone? #
- Use A-weighting (dBA) unless you have a reason not to. It’s what hearing limits and most noise rules use. See dB vs dBA.
- Uncover the microphone. Thick cases and fingers over the bottom edge can shave off several dB.
- Hold the phone away from your body at about the height of your ears, with the microphone end facing the noise. Your body reflects and blocks sound.
- Measure for a while and read the Leq, the energy average, rather than eyeballing a jumping number. Thirty to sixty seconds is a good start for steady noise. What is Leq? explains why.
- Avoid wind and handling noise. A breeze across the mic or a thumb shifting on the glass can add big false peaks.
- Calibrate if you can. Borrow a calibrated meter or a known sound source and set an offset so your phone matches it. Our guide to calibrating a sound meter app walks through it.
- For iPhone and serious work, pair the NIOSH SLM app with a calibrated external microphone.
Where does Sound Meter dB fit? #
Sound Meter dB is upfront about this. The app does the math a real meter does, applying A-weighting to the raw microphone signal, Fast (125 ms) or Slow (1 s) time weighting and a true Leq instead of a simple average. On iOS it uses the system’s measurement audio mode. It still maps your mic’s digital level to an approximate dB SPL scale, so its own footnote calls readings “good estimates for awareness and comparisons, not a substitute for a Class 1 or Class 2 sound level meter.”
If you have access to a reference meter, Settings, then Calibration, lets you add an offset of up to ±30 dB so your phone lines up with it. The offset is saved on the device and applied to every reading from then on.
Frequently asked questions #
Are iPhone sound meter apps more accurate than Android ones? #
In the NIOSH testing, iOS apps performed more consistently, and NIOSH built its app for iOS only because Android phones vary so much between manufacturers. A good Android app on a given phone can still be consistent; you just can’t assume one phone’s accuracy carries over to another model.
Can I use a phone decibel reading in a noise complaint? #
You can use it to document when and roughly how loud a noise was, which helps a complaint. Official enforcement usually relies on an inspector’s calibrated meter and your local ordinance’s measurement rules, so don’t expect a phone reading to be the deciding evidence. See how to measure neighbor noise.
Do phone sound meters read too high or too low? #
Either, depending on the phone, the app and the kind of sound. Low rumble is often under-read, and handling noise or wind can make readings jump too high. Calibrating against a reference meter removes most of the constant offset.
What is the most accurate free sound meter app? #
The most thoroughly tested free option is the NIOSH SLM app for iPhone and iPad, which NIOSH reports as accurate within ±2 dBA. On Android, no app has an equivalent official validation, so calibration against a reference meter matters more there.