Teams Audio Test: Check Mic and Speakers, Make a Test Call

Before your next stand-up, run this Teams audio test to confirm your mic picks you up and your speakers or headphones play sound, without dialing anyone. You get a clear pass or fail for each device in under a minute, then you can finish the check inside Microsoft Teams so the first thing people hear in the meeting is your voice, not a muted mic.

How the Teams Audio Test on This Page Works

The page runs two checks once you allow mic access. The speaker check plays a tone through the left channel, then the right so you hear each side. The mic check reads your audio input live and reports two numbers in dBFS, where 0 is the digital maximum: the speech peak (the loudest point while you talk) and the idle floor (the level while you stay silent). The google meet audio test tool gives you a clear answer instead of guessing.

Worked example: you test your microphone by reading one sentence aloud and the peak lands at −16 dBFS. Silent, the floor settles at −47 dBFS:

$$\text{Gap} = \text{Peak} - \text{Floor} = -16 - (-47) = 31\ \text{dB}$$

A 31 dB gap puts your voice well clear of the room, and 16 dB of headroom below 0 dBFS means a laugh won't clip.

Reading the Audio Level Meter

The audio level meter is the bar that jumps as you speak. As a rough guide, speech peaks between about −24 and −10 dBFS read clearly on a call. A peak below −30 dBFS makes people strain to hear you, and one that touches 0 dBFS crackles. If the floor sits within 20 dB of your peak, background noise such as a fan or keyboard will carry into the call. If the speaker tone shows up on your mic reading, others may hear an echo, so switch to headphones.

Running a Teams Audio Test Before a Grant Panel Interview

A researcher has a grant panel interview over Teams in 25 minutes, using a USB condenser mic on a desk arm in a spare room with a dehumidifier humming in the corner. The question is simple: will the panel hear her clearly, or does something need to change first? The answer turns out to be no, not as set up: the voice-to-noise gap is only 24.4 dB.

She reads the first line of her opening statement at the volume she plans to use. The mic check shows a speech peak of −27.4 dBFS and, in silence, an idle floor of −51.8 dBFS. The gap is −27.4 − (−51.8) = 24.4 dB, and the peak sits in the soft band, below the −24 dBFS edge of clear speech.

Her first instinct is to raise the input gain by 9 dB. That lifts the peak to −18.4 dBFS, nicely inside the clear band, but it is the wrong fix on its own: gain raises the hum by the same 9 dB, so the floor rises to −42.8 dBFS and the gap stays at exactly 24.4 dB. Louder, yes. Cleaner, no.

Halving the distance to the mouth adds about 6 dB of voice (the inverse-square law) while the room hum barely changes, so the gap grows to 30.4 dB. Switching the dehumidifier off for the half hour drops the floor to −60.2 dBFS and lands the gap at 38.8 dB, with 21.4 dB of headroom, so even a nervous laugh during the interview won't clip on the Teams call.

My judgement: the number to watch is the gap, not the peak. A loud peak over a loud floor still sounds like a call from a laundry room. The caveat is that Teams applies its own processing after this page, so she plays back one sentence through Teams before joining to confirm the hum is gone there too.

Video Settings and a Solo Video Call Audio Check

For a longer audio check than the test call, record a session alone to hear yourself with Teams' own processing applied: create a calendar event in Outlook, turn it into a Teams meeting, join it, and choose Record and transcribe > Start recording from the ... menu. The video appears in that chat about a minute after you leave. The video settings on the same Devices page preview your camera.