Check the selected input above the instrument, then speak close enough for a normal call. If another input is listed, select it and start a fresh run. A granted site permission only allows a request; it does not prove that the intended microphone was selected or that its signal is active.
Mic test
Run a mic test to confirm that the selected input reaches this browser, then read peak level, average level, clipping and quiet gaps before recording a short local playback.
Also known as: mic test, mic test online, microphone test, test microphone, test mic, audio input test, mic test playback. Free, online, and no download required.
microphone, permission requiredNothing uploaded
Start the mic test online, allow microphone access and speak at your normal call distance.
01
What the mic test checks
| Check | How it is measured | Fails when |
|---|---|---|
| Input signal | Float time-domain samples produce peak and RMS values in dBFS. | The rolling sample remains at digital silence. |
| Clipping | Three or more consecutive samples at an absolute value of 0.99 are counted. | Sustained full-scale samples distort the captured waveform. |
| Quiet gaps | The analyser records digital-silence windows of at least 120 ms during activity. | The input repeatedly drops to exact silence during a steady signal. |
How to run the mic test
- Choose the microphone you intend to use and place it at your normal call distance.
- Press Start, allow access and speak in a steady voice for several seconds.
- Read peak, average level, clipping and quiet gaps before changing a setting.
- Record 5s locally, stop speaking when the countdown ends and play the sample back.
- Change one input, gain or processing setting, then repeat the same phrase at the same distance.
If something failed
Keep the same speaking distance and compare the live average before changing input volume in the operating system. Raise one control in a small step, restart the mic test and listen to a new recording. Do not raise playback volume and call that an input improvement.
Turn monitoring off and use headphones before recording again. Speakers can return the playback to the active microphone. The five-second recording stays in this tab, but an open speaker-to-mic path can still create acoustic feedback in the room.
02
How to read the result
Mic test with playback
Start with the live meter. The microphone test reads the audio stream selected by the browser and shows level in dBFS, where zero is the digital ceiling. A moving waveform confirms that samples reached this page. It does not confirm that Zoom, Teams or another application is using the same input.
Need to hear the result? Use mic test playback after the meter responds. The Record 5s locally control holds one short recording in memory and creates a playback control in this tab. It is separate from monitoring, which routes the live input toward the output and can feed back through speakers.
Keep the input test and playback test in order. The mic test with playback begins only after a live signal appears. If the meter is still, a silent recording adds no new evidence. If the meter moves but the recording sounds faint, compare speaking distance and the average level before changing system input volume. That keeps the audio input test tied to one clear question.
Echo, noise and sensitivity
Do you hear your own voice twice? Turn off live monitoring and close other software that may be monitoring the microphone. Echo in this page can come from sound leaving speakers and returning to the microphone; the browser cannot inspect echo cancellation used later by a call application.
Background noise needs a quiet reference. Stay silent for a few seconds, note the displayed floor, then speak from the same position. This is where echo, noise and sensitivity need separate observations. Moving the microphone closer can increase voice level relative to room noise without changing digital gain. Do not describe one room reading as a permanent microphone specification.
How sensitive is the current setup? Compare two controlled runs with the same phrase and distance while changing one supported input control. The page reports the stream after operating-system and browser processing. An auto gain control value in the track settings describes the negotiated stream; it does not reveal every adjustment made by the device or driver.
How the mic test works
The page requests microphone access after the Start action. Once a stream opens, Web Audio supplies float time-domain samples to the analyser. The meter records peak and RMS values, counts clipping samples and watches for exact-silence gaps after activity has begun.
Peak shows the strongest sample in the current frame. RMS represents signal energy across that frame and is used here as an average level. The numbers are useful for comparing the same setup; they are not calibrated sound-pressure measurements and cannot state how loud the microphone is in a room.
A quiet gap is counted only when activity was already seen and exact digital silence lasts at least 120 ms before signal returns. A noise gate or interrupted stream can create that pattern. Ordinary quiet audio above exact silence may not enter the gap counter, so read the waveform and playback with the count.
What the result can establish
A responsive meter establishes that this browser received an audio stream during the run. A clean local recording establishes that the captured sample can be played back in this tab. Neither result establishes which input a separate application will choose when it opens later.
Does one application remain silent while this test works? Open the microphone not working guide and compare that application's selected input, mute state and permissions. Use the same named device. Changing several controls at once makes it harder to identify which layer changed the result.
Clipping or a low average is an observation, not a repair instruction. Repeat the same phrase once after a single level or distance change. Keep the better result only if the playback improves without new clipping, gaps or feedback.
Measurement limits
The browser cannot inspect analogue gain stages, cable condition, driver internals or processing inside another application. It receives a negotiated media stream. The mic test cannot certify frequency response, latency, speech quality or whether a device meets a manufacturer specification.
No recording or measured value is uploaded by this tool. The short recording remains in page memory and is discarded when it is replaced or the page closes. Device labels become available after permission so you can choose an input; this page does not store a hardware inventory on a server.
Use test microphone results as a controlled browser observation. If the same problem persists in the operating system and several applications, record the device name, connection and exact symptom before seeking model-specific support. A browser cannot select the required repair from those symptoms alone.
03
Mic test questions
How do I run a mic test online in this browser?
Choose the intended input, press Start and allow microphone access. Speak at your normal call distance until the waveform and level move. Read the values before changing anything, then use Record 5s locally if you need playback. This mic test online checks the stream received by the current browser tab.
What do peak and RMS mean in this audio input test?
Peak is the strongest sample in the current analysis frame. RMS describes energy across the frame and serves as the average level shown here. Both use dBFS, with zero at the digital ceiling. Compare repeated runs of the same phrase; neither number is a calibrated sound-pressure measurement.
Why does the microphone test show gaps of exact silence?
The counter begins after activity and records a gap when exact digital silence lasts at least 120 ms before signal returns. A gate, stream interruption or connection problem can produce that pattern. Quiet room sound above exact zero may not count, so listen to local playback and inspect the waveform too.
How do I stop echo in a mic test?
Turn off live monitoring and use headphones before trying again. If speakers play the microphone signal, the same sound can return through the active input and create feedback. Close other monitoring software as well. Echo cancellation inside a call application is outside this browser test and needs a separate comparison.
Why does my microphone pick up so much background noise?
Record a quiet reference, then repeat from your normal speaking position. Move the microphone closer before adding gain, because distance can improve voice level relative to the room. Check fans and speakers nearby. The displayed floor applies to this setup and moment; it is not a fixed specification for the microphone.
How do I check microphone sensitivity?
Keep the phrase, distance and room unchanged while comparing one input-volume setting at a time. Read average level, peak and playback together. This shows how the delivered stream changes. It cannot measure analogue sensitivity directly or reveal every automatic adjustment made before audio reaches the browser.
How should I repeat the mic test fairly?
Use the same microphone, browser, distance and phrase. Record the first average, peak and gap count, change one input or processing setting, then start a fresh run. Keep playback volume unchanged when judging captured loudness. If results disagree, collect another sample before choosing which setting to keep.
What can the mic test not prove?
It cannot prove analogue microphone quality, driver health, acoustic loudness, call quality or the settings used by another application. It confirms only the negotiated stream and samples received by this tab. Use the how to test a microphone guide to plan a controlled follow-up.