MicTester

How many decibels your microphone is missing

Automatic gain control is switched off on this page on purpose. It is the setting that hides a quiet microphone by propping it up, and you cannot fix a level you cannot see. Say a sentence, read the hold figure, and look it up in the table below.

Sample trace

A drawing, not a recording — nothing here came from a microphone yet.

Spectrum

low → high frequency

Level (dBFS)

Peak -12.3|RMS -21.3|Hold −∞

−50−20−9−0.5−600

Reading: On target — that is the sample trace’s own peak, not yours.

Before the browser asks

  • · Your browser will ask for permission. Nothing happens until you allow it.
  • · The page then reads three things: the waveform, how loud you are, and the device list.
  • · No recording starts on its own. The recorder has its own button below.
  • · Audio is processed in this tab and never uploaded — there is no server to upload it to.
  • · Automatic gain control hides the fault you are trying to measure: it props a weak signal up and the meter looks fine. Off is the only honest starting point.

Why “too quiet” is a number, not a feeling

“You’re very quiet” is the most-repeated sentence in remote work, and it is almost never acted on properly, because nobody converts it into a quantity. A microphone that peaks at −38 dBFS is not vaguely quiet — it is precisely eighteen decibels below where speech should sit, which is a linear factor of about eight. That is a number you can go and find on a knob.

Band Peak range What MicTester tells you
Nothing arriving below -50 dBFS Below −50 dBFS there is no usable speech in the signal. Either the wrong input is selected, the device is muted in hardware, or nobody is talking.
Too quiet -50 to -20 dBFS Audible, but the far end will strain. Raise the input gain until your loudest speech peaks land above −20 dBFS.
On target -20 to -9 dBFS Speech peaks belong in this window. It is the band between the alignment level and the permitted maximum, so you have headroom for a laugh or a cough.
Hot -9 to -0.5 dBFS Loud enough that a sudden peak can hit the ceiling. Fine if your speech only touches this band; back the gain off if it lives here.
Clipping risk -0.5 dBFS and above At the ceiling. Anything louder cannot be represented and comes out as distortion that no amount of processing downstream will remove.

Where these four boundaries come from. −50, −20, −9 and −0.5 dBFS are the thresholds OBS Studio publishes for its own input-level indicator, in the knowledge-base article “Audio Mixer Technical Details” dated 2022-02-13, read directly on 2026-08-15. MicTester reuses them so a reading here means what a reading there means. The band names and the advice in the last column are this site’s own reading of those levels — they are a scale we declare, not a survey of anyone’s microphones, and we have measured no population to support them. The target band here is the same one every page on this site uses; nothing on this page is a measurement of your hardware except the reading the tool takes from it.

How much gain each reading needs

Every column below is arithmetic on the reading in the first one. Find the row nearest your peak hold and read across.

Gain to target is -20 minus the reading; gain to ceiling is -9 minus the reading. The multiplier is 10gain/20, the linear factor that gain represents. Bits used is 16 + reading ÷ 6.02, i.e. how much of a 16-bit word a signal at that level occupies — every 6.02 dB of shortfall costs one bit. Generated at build time from src/lib/dbfs.ts and asserted cell by cell in this project’s tests; no microphone was measured to produce any of it.
Your peak reads Gain to target Gain to ceiling That is a factor of Bits used of 16 Band
-60 dBFS +40 dB +51 dB 100× 6.0 Nothing arriving
-54 dBFS +34 dB +45 dB 50.12× 7.0 Nothing arriving
-48 dBFS +28 dB +39 dB 25.12× 8.0 Too quiet
-42 dBFS +22 dB +33 dB 12.59× 9.0 Too quiet
-36 dBFS +16 dB +27 dB 6.31× 10.0 Too quiet
-30 dBFS +10 dB +21 dB 3.16× 11.0 Too quiet
-24 dBFS +4 dB +15 dB 1.58× 12.0 Too quiet
-20 dBFS +0 dB +11 dB 1× 12.7 On target

The last column is the part people underestimate. Volume can be fixed later; resolution cannot. A signal peaking at −48 dBFS is using roughly half the bits of a 16-bit word, so amplifying it afterwards stretches a coarse signal across the full range and brings its quantisation noise up with it. This is why a recording made too quietly and “fixed” in an editor sounds hissy while one recorded at a sensible level does not.

Fix it in this order

1. Distance, before anything electrical

Sound pressure falls off with distance from a source, so in a free field halving your distance roughly doubles the amplitude reaching the capsule — about 6.02 dB, the same 6.02 dB that one bit of resolution is worth. Going from 60 cm to 30 cm is therefore worth as much as a substantial gain increase, with one decisive advantage: it does not move the fan, the street or the neighbour’s dog any closer. Real rooms are not free fields and reflections blunt the effect, so 6.02 dB is the ideal case and a real room will hand you less than that — but the direction is always right.

2. The gain control on the device

An audio interface or a USB microphone with a physical knob should be set here first. OBS Studio’s audio guide puts it plainly: always start at the device in question, and check whether it has a physical knob usually labelled “Gain”. Everything downstream can only amplify what this stage produced.

3. The operating system’s input level

Both paths quoted from Microsoft Support, “Fix microphone problems” (Applies To: Windows 11, Windows 10), read directly on 2026-08-15. The two rows are genuinely different instructions in different tabs of that article, not two names for one screen — Microphone Boost appears in the Windows 10 tab.
PathDetailDocumented for
Start → Settings → System → Sound → Input → the device’s properties page Adjust the Input volume slider under Input settings, retesting as you go. Windows 11
Settings → System → Sound → Input → Device Properties → the Levels tab Microsoft’s Windows 10 instructions put Microphone and Microphone Boost sliders here. OBS Studio’s own guide adds a useful trick for this dialog: right-click the percentage value and choose decibels, then aim for about 0.0 dB for the least clipping. Windows 10

4. The application, last

Only now touch the input slider inside Zoom, Discord or Teams. If you have to push an app’s slider to maximum to be heard, the three steps above were skipped, and the app is amplifying a signal that was already too small — noise and all.

Eighteen decibels, found in ninety seconds

The readings below are an illustration of how the arithmetic is used, not a measurement of any particular microphone — yours will differ, which is the whole reason the meter is on the page.

Someone is told they are quiet on every call. On this page, with gain control off, they say a normal sentence and the hold reads −38 dBFS. That reading falls between the table’s rows, so run the same arithmetic on it directly: +18 dB to target, a factor of about 7.9, and about 9.7 of 16 bits in use. So the shortfall is real and large.

They move from the back of the chair to a normal seated distance, roughly halving the gap, which the free-field arithmetic says is worth about 6.02 dB: the hold moves to about −32. Raising the Windows input volume takes it to −24, and a turn of the interface’s gain knob lands it at −18, inside the target band. The app’s own slider was never touched — which means it still has headroom in both directions instead of being pinned at maximum.

Worth checking afterwards: turn automatic gain back on and confirm nothing now clips. A microphone set correctly with gain control on top is fine; a microphone set too hot with gain control fighting it is the other failure, and the meter above shows it immediately as peaks in the clipping band.

When it is not a gain problem at all

  • The meter reads −∞. Digital silence. Nothing to amplify — go to the not-working page.
  • The level is fine here but not in one app. That app’s own input slider or processing. The Zoom page covers the automatic-volume trap specifically.
  • You are quiet only sometimes. Gain control hunting, or a headset boom that swings away when you turn your head. Record and listen while moving.
  • You are loud but unintelligible. That is not level, it is noise or processing — try the headset test with processing off.

FAQ

Quiet microphone questions

Why does my mic sound fine in some apps and too quiet in others?

Because some apps apply automatic gain control and some do not. Gain control multiplies a weak signal up until it looks normal, so a badly set microphone can sound acceptable in a meeting and hopeless in a recording. That is why this page turns the browser’s gain control off by default — it is the only way to see the level you actually have rather than the one an algorithm is propping up.

How much gain do I actually need?

Read the peak hold above and subtract it from -20. If your loudest normal speech peaks at −38 dBFS, you need +18 dB, which is a linear multiplier of about 7.9×. The table on this page does that arithmetic for a range of readings so you can look yours up without reaching for a calculator.

Is it better to raise the gain or move closer to the microphone?

Closer, every time, and the arithmetic says why. Sound pressure falls with distance, so in free field halving your distance adds about 6.02 dB — and it adds that to your voice only, not to the fan behind you, because the fan’s distance has not changed. Turning the gain up adds the same decibels to everything, noise floor included. Move first, amplify second.

What is “microphone boost” and should I use it?

It is amplification applied in the driver, and Microsoft documents Microphone and Microphone Boost sliders on the Levels tab in its Windows 10 instructions. It works, but it belongs at the end of the chain: it amplifies whatever the device already captured, including its noise. Use a physical gain knob if you have one, then position, then the operating system’s input volume, and reach for boost only if you are still short.

Why does my meter read −∞ instead of a low number?

Because that is a different fault, and the distinction matters. Minus infinity means the samples arriving are exactly zero — digital silence, which no amount of gain will amplify. That is a muted device, a hardware mute switch, or the wrong input selected, and the not-working page is where it belongs. This page is for a signal that exists and is too small.

Does a quiet microphone actually lose quality, or just volume?

Both, and the table on this page shows why. A 16-bit sample has 16 bits of resolution at full scale, and every 6.02 dB below that leaves one of them unused. A signal peaking at −48 dBFS is using about 8 of the 16 — so when you amplify it later you are stretching half the available resolution across the whole range, and the quantisation noise comes up with the voice. That is arithmetic, not opinion.

More from MicTester

When the level is finally where you want it

  • the full mic test with every switch on one page — waveform, spectrum, dBFS meter, recorder and device picker together, with all three processing toggles exposed.
  • microphone latency test — Reads the delay your own browser reports and shows the buffer arithmetic every one of those figures comes from.
  • mic test for Discord — Runs the same record-and-play-back loop Discord’s own Mic Test does, with Discord’s processing defaults switched on, plus the exact menu path to Voice & Video.
  • Keyboard Tester — The same idea for the other input device: press a key, watch it light up, and see the exact event your browser received.
  • Click Speed Test — Counts clicks per second against the browser’s own timestamps, so the deadline holds even when the frame rate does not.