Support

Sound Meter for iPhone

Tell us your iPhone model and iOS version, and — if it is about a reading — what the status pill said at the time (Ready, Approximate, Check mic or External). That one word explains most surprises.

Questions we expect

Why does it say “Approximate”?: Because no reference-measured profile exists for your iPhone model yet. The app applies a fixed provisional offset instead of a measured correction, and says so rather than presenting the number as calibrated. Relative changes — quieter versus louder — are still reliable.
Is this as accurate as a real sound level meter?: No, and we will not claim otherwise. It is a useful estimate, not a Type 1 or Type 2 instrument. Use certified equipment where a regulation or legal standard requires it.
It says “Check mic” on a perfectly clean phone.: The check needs a baseline and repeated evidence before it says anything, and it never claims certainty — the wording is always “may be blocked”. If you see it on a clean phone, please write to us: that is a false positive and we want the details.
The number looks wrong with my AirPods connected.: An external microphone has its own response, so the app labels the state “External” and does not apply the iPhone profile to it. Switch to the iPhone microphone from the banner on the Meter screen.
Where are my recordings, and do they leave my phone?: The app stores recordings locally and does not upload them. They are excluded from iCloud backup, and are deleted when you delete the session. That also means they do not transfer to a new iPhone.
How do I turn advertising off?: Buy or restore Sound Meter Premium to remove ads. This stops new ad requests once the entitlement is recognised; it does not erase data already processed while using the free version. Apple still handles purchases. Advertising choices, including the US “Do Not Sell or Share My Personal Information” control where applicable, are in Settings → Privacy.
I paid and it did not unlock.: Settings → Sound Meter Premium → Restore Purchases, on the same Apple Account that made the purchase. If that fails, email us with the date of purchase.

Getting a reading you can trust

Most surprising readings are not the app being wrong; they are the microphone being covered, the source being the wrong distance away, or the number meaning something other than what was expected. In order of how often it matters:

Find the microphone and leave it clear. It is on the bottom edge. A finger, a table surface or a thick case over it can cost many decibels.
Hold the phone away from your body, roughly at arm's length, pointing the bottom edge toward the sound. Clothing and a chest reflect and absorb.
Let it settle. Give it ten seconds before reading. A meter that has just started is still filling its time-weighting window.
Watch for wind. Outdoors, wind across the microphone reads as noise that is not there. There is no windscreen on a phone.
Record a session rather than screenshotting one instant. Sound moves; the average and the peak over a minute say far more than any single number.

What the metrics mean

dBA, dBC, dBZ: Frequency weightings. A approximates how human hearing responds and is the standard for environmental noise — use it unless you know why you want another. C keeps more bass, which is why it is used for peaks. Z is no weighting at all.
Fast and Slow: How quickly the displayed number reacts. Fast (125 ms) follows short events; Slow (1 s) smooths them out. Both are standard; they answer different questions.
LAeq: The energy average over the session — the steady level that would carry the same total energy. This is the number most noise guidance is written in terms of, and it is not the same as the arithmetic average of what you saw on screen.
MAX, and why it may look low: The highest time-weighted level, which is the standard LAFmax. Some apps report an instantaneous peak hold under the same word, which catches a single sample and therefore prints a bigger number for the identical sound. Neither is wrong; ours states which it is, and reports the instantaneous figure separately as LCpeak.
LASmin: The quietest moment, Slow-weighted. Useful for describing a background level between events.
Noise exposure: A plain-language band, not a dose calculation and not a health assessment. It is A-weighted, so it follows the dBA level even when the big number is showing dBC or dBZ.

Exporting a session

Open a session in History and export it as PDF (a readable report with the level curve and the summary) or CSV (every sample, for a spreadsheet). Both carry the device, the settings and the honesty line stating the reading is uncalibrated, so the document cannot be mistaken for a certified measurement later. Long sessions are downsampled in the curve — the caption says so, and the summary figures stay authoritative.