Guide · Audio in

Microphone Too Quiet? Fix Low Mic Volume

By Device Bench · Published · Reviewed

A quiet microphone does not always need more gain. First check which microphone is selected and how it is positioned, then work through input levels and processing. Use the same phrase and distance for each comparison so you can tell whether a change helped.

Working band−30 … −12 dBFS Meter floornear −60 dBFS First checkposition + angle

Keep the test open while you work through the checks below. Its meter and local playback help you compare adjustments; the text summary lets you save the readings with your notes. The test runs in your browser, and no audio leaves your device.

  1. Open the Microphone Test and allow the input you intend to compare.
  2. Speak the same phrase from the same distance before and after one change.
  3. Copy the local diagnostic summary. It contains the observed readout values and limits, not audio or the microphone name.

Work through the signal chain

Your voice reaches the microphone's capsule, then the input stage and converter turn it into digital samples. Device controls and operating-system settings affect the level; the browser or call app may also process the stream. More than one stage can contribute to a faint or noisy result. Raising gain later in the chain can make both voice and noise louder without improving their balance.

Start by checking the selected microphone and its physical controls. Set a repeatable speaking position, then try the operating system's input level and check the browser or app settings. If the level is still too low, compare another device or input path. Change one thing at a time: a better reading shows which adjustment helped, not which hidden component caused the problem.

Check distance and direction

Start with the microphone in its intended position and follow the manufacturer's guidance. Moving closer usually captures more voice relative to room sound, but useful spacing depends on the microphone type, pickup pattern, mounting, and plosive protection. Try a small change in distance or angle, then compare the meter and local playback before moving it again.

Check which side of the microphone is meant to face you; many microphones are less sensitive from other angles. Keep a headset boom out of the direct breath stream. Laptop microphones are fixed, so a change in posture or room sound may matter more than it would with a movable microphone.

Adjust the operating system's input level

Once the position is settled, adjust the operating system's input level while watching the meter. On Windows, open Settings, then Sound, choose the input device, and raise the input volume slider. Some devices also offer boost in their properties. On macOS, open System Settings, then Sound, and use the input level slider in the Input tab while speaking.

Keep the microphone test open in a second window and speak normally as you adjust the level. Confirm that the operating system and test are using the microphone you intend to check, especially if both a built-in mic and a headset are available. Leave boost until you have tried position and the main input slider: it raises the noise floor along with your voice. Compare playback as well as the meter before keeping the new setting.

What the meter should read

The mic test measures digital amplitude against its full-scale reference, with the displayed level capped at 0.0 dBFS. Device Bench marks −30 to −12 dBFS as a practical working band for this browser meter, not as a universal recording, conferencing, or broadcast standard. Watch the moving level and held peak, then listen to local playback rather than choosing a setting from the number alone.

If the signal stays near the meter floor, go back to distance, the selected input, physical controls, and input gain. A Peak displayed at 0.0 means the stream reached this meter's near-full-scale threshold. It does not identify where clipping or limiting occurred upstream. For readings between those cases, repeat the comparison at the same speaking level and distance.

Browser and app processing

Browsers and call apps may apply automatic gain control (AGC), noise suppression, or echo cancellation. If the level drifts while you hold your voice as steady as practical, automatic processing is one possibility; it does not prove AGC is active. Voice variation, another processor, a connection issue, or hardware behavior could also explain it. Check the browser-reported settings and the call app's controls before deciding what to change.

Aggressive noise suppression can cut off the quiet edges of speech; echo cancellation can affect loudness too. Processing can differ between apps, so a microphone that measures well here may still sound faint in a particular call app. Check that app's audio settings separately. The mic test records and plays back the stream the browser gives this page, after browser-level processing. It does not include a call app's later constraints, codec and network path, or the remote listener's playback. Use the call app's own test call or preview as a separate check.

Check the hardware path

If the signal stays near the floor after the software checks, look at how the microphone connects. A 3.5 mm analog microphone relies on the computer or interface's input stage; a USB microphone has its own preamp and converter. That difference alone does not tell you which will give a better level. Models, controls, drivers, processing, and connections vary, so compare the actual devices rather than choosing by connection type.

Before buying a replacement, check any gain dial, mute switch, inline control, cable, adapter, hub, or vendor utility in that path. A low browser reading alone cannot tell you which of them is responsible.

If those checks still leave too little level or too much noise, different hardware or a suitable interface may be worth trying. Repeat the same meter and local-playback checks with the same phrase and position after the change.

Sources and standards

The references below explain browser capture constraints, settings, and processing controls. Device Bench uses the troubleshooting order and −30 to −12 dBFS working band here to compare changes, not to diagnose hardware or set a conferencing standard.

See a mistake or a changed standard? Report a correction.

Related