How to Measure Microphone Self-Noise

Identify the symptom and environment

Microphone self-noise is hiss or broadband noise that the microphone itself produces even when the room is quiet. It is separate from background noise such as fans, traffic or room tone. You usually notice it as a soft rushing sound in recordings, or as a noise floor that stays visible in audio software when the microphone is connected but no one is speaking.

Before measuring, record what you are using: the microphone model, the audio interface or USB connection, the operating system, the recording app or browser-based tool, and the current input level. Also note the sample rate if the app exposes it. These details matter because a higher input level can make the same self-noise appear louder, while changes to the sample rate should not change the physical noise of the microphone itself.

Isolate where the fault occurs

First decide whether you are measuring the microphone or the recording chain around it. Unplug other input devices that could add noise, or disable them in the app. Mute speakers and headphones to avoid acoustic pickup or monitoring loops. Close unnecessary apps that may use the microphone, but avoid disabling system-wide security protections as a way to reduce noise.

If possible, compare the microphone with another input device on the same computer and with the same input level. A very different noise reading may point to gain, connection or driver settings rather than the microphone capsule. For USB microphones, try another port if the reading contains hum, clicks or bursts; if the sound is stable broadband hiss, the port is less likely to be the cause.

Record a sample with the microphone covered but not touching the capsule. A padded box, a thick cloth or a quiet closed space can help, but the result is only useful if background noise is clearly lower than the microphone noise you are testing. Note the distance and covering method because that changes how much external sound reaches the mic.

Try reversible fixes in order

Set the input level to a repeatable position and leave it fixed during the measurement. In desktop audio software, use a recording track with no processing such as compression, noise reduction or automatic gain. In a browser tool, allow microphone access only for the site you are testing. Check the operating system privacy settings to confirm the correct mic is selected and that the app has permission.

Record at least ten to twenty seconds of silence. Then select a section that contains no clicks, mouth sounds, bumps or handling noise and read the RMS level in decibels, often labelled dBFS. This is not the same as a manufacturer self-noise specification, which is usually reported in dB SPL or dBA, but it lets you compare two setups under identical conditions.

For a more useful comparison, repeat the test at the same sample rate and input level with each microphone or setting you want to evaluate. Keep the recording conditions as close as possible. If the app offers a spectrum view, look for a smooth noise floor rather than spikes at mains frequency or its harmonics; strong narrow spikes may indicate electrical interference instead of self-noise.

Confirm the result

Play back the original recording and confirm that the hiss or noise floor described at the start is reduced or gone after any change. If you changed only the input level, repeat the original setup and check that the observed noise changes in proportion to gain. A measurement is meaningful only when the before-and-after conditions are recorded.

Do not treat one dBFS reading as an official self-noise measurement unless the setup is calibrated. Keep the comparison result: microphone, interface or port, input level, sample rate, covering method and RMS reading. That record is enough to support a practical decision even when it cannot replace a manufacturer specification.

If the problem remains

If the noise stays identical across very different input levels, the recording chain or monitoring path may be the source. Recheck the selected input and output devices, app permissions and cable or connection. If the noise changes strongly with gain and remains in a quiet comparison recording, the microphone or its built-in preamp may be producing significant self-noise.

Save the comparison observations: which microphones were tested, the exact settings, and the recorded RMS or spectrum result. Contact device support if a new microphone is unexpectedly noisy, if the noise includes strong hum or crackle, or if the self-noise level makes the microphone unusable for its intended purpose. A symptom alone is not proof of failed hardware, but a repeatable comparison record gives support a useful starting point.