Why Phone Decibel Readings Differ From a Sound Level Meter

Contents
When you measure noise with a smartphone decibel app or web tool, the number often comes out quite different from what a professional sound level meter shows. That's not because the tool is inaccurate. It's because a phone microphone and a professional meter turn sound into numbers in fundamentally different ways. Once you understand how that gap is built, you'll have a clearer sense of how to take a decibel reading from a phone.
Summary ① A professional sound level meter is calibrated against a reference tone. A phone microphone has no fixed reference and runs on an arbitrary formula. ② There's no dB(A) weighting to match human hearing — the raw waveform hitting the mic is converted straight to decibels. ③ The automatic gain control and noise suppression that most browsers leave on by default have already altered the input before you see it.
The problem: same sound, different numbers on every device
When the noise from the floor above is getting to you, or the construction site down the street feels too loud, it's tempting to pull out your phone, measure the decibels, and treat that number as evidence. But in practice, different apps and different phone models give you different numbers for the same moment.
Borrow a professional sound level meter and measure from the same spot, and you may get a third value. No wonder it's confusing to tell which one is right.
That gap isn't a measuring mistake. It's a structural result of phone-mic measurement and professional equipment measurement being different methods from the ground up.
The fix: how to read a smartphone decibel reading correctly
First, the reference point (calibration) is different. A professional sound level meter is zeroed regularly with a calibrator that emits a fixed reference tone, such as 94dB or 114dB. A phone microphone has no such absolute reference. It calculates the strength (RMS) of the waveform coming in, adds an arbitrary offset, and reports that as a dB number. That offset isn't a physically measured reference — it's a "this seems about right" approximation. So when mic sensitivity varies from one device to the next, the same sound produces different numbers.
Second, frequency weighting is missing. Human hearing is more sensitive in the midrange (around the range of the human voice) than at very low or very high frequencies. That's why dB(A), the unit used in environmental noise standards, applies different weighting to each frequency to match that hearing curve. A phone-mic reading often reflects the incoming waveform as-is with no weighting, so when low-frequency noise is mixed in (an air conditioner, a refrigerator motor, and so on), the number can come out higher than what you actually perceive.
Third, the browser and OS have already processed the input. When a browser takes microphone input, it leaves automatic gain control (AGC), echo cancellation, and noise suppression on by default unless you turn them off. These features exist to flatten the audio for better call quality. So even when the actual sound pressure rises, the mic turns the input down on its own and the number doesn't climb as much as you'd expect.
Because of these three things, it's safer to treat a phone decibel reading as an indicator of relative change under the same conditions rather than as an absolutely accurate dB value. For repeated measurements on the same phone at the same spot, comparing before and after, it works well. But for situations that need an official standard — a neighbor-noise dispute or a workplace noise inspection — you need confirmation from a calibrated professional meter.
If you want to know how loud it is where you are right now, try our free Sound Decibel Meter.
Measure the noise right where you are, live, with your microphone.

Common snags
| Symptom | Cause | What to check |
|---|---|---|
| Nothing happens when you press [Start] | The browser has blocked the microphone permission request | Open the lock (or site information) icon at the left of the address bar and set microphone permission to Allow |
| The number won't drop in a quiet room | The mic also picks up low-frequency noise you can barely hear (air conditioner, refrigerator, etc.) | Turn off suspected noise sources one at a time and compare the numbers |
| The min, average, and max stats stay at 0 | Anything below 5dB is excluded from the statistics | Make a sound close to the mic so it goes above 5dB, and the stats will start filling in |
| Values differ from phone to phone in the same spot | Mic sensitivity, position, and whether there's a case all vary by device | Compare before-and-after readings on the same device instead of comparing absolute values |
| Max and min reset after you stop and restart | The stats reset to their initial state each time you start a new session | Only compare values from within a single measuring session |
The takeaway
A decibel reading from a phone is an approximation built from three layers: a different reference point than a professional sound level meter, no frequency weighting, and the browser's automatic audio processing. The safest use is to take repeated readings on the same device in the same spot and look at relative change, rather than holding up the absolute number as proof to someone else.
If you're curious how loud your room or office is right now, you can measure it right below.
Measure the noise right where you are, live, with your microphone.
Frequently asked questions
Can I use a smartphone decibel reading as official evidence, like for a noise complaint?
A phone microphone isn't equipment calibrated against a reference tone, so the reading itself is hard to use as official documentation. If you need something that meets an official standard, it's safer to get confirmation with a calibrated professional sound level meter.
Why do an iPhone and an Android phone give different readings for the same sound?
Mic hardware sensitivity and placement differ by device, and so does the amount of automatic gain control and noise suppression that the OS and browser apply by default. That's why comparing readings from the same device is more meaningful than comparing absolute values.
Is the dB shown here the same unit as the dB(A) I hear about in the news?
This tool converts the strength of the sound waveform entering the mic straight into decibels, with no weighting based on the human hearing curve. dB(A), the unit used in environmental noise standards, applies lower weighting to the frequency bands where human hearing is less sensitive. So for the same sound, the two values won't match exactly.
This article was adapted from our Korean original. Read the Korean original