Output
Waveform; spectrum; digital dBFS; local clip
UserBenchmark / All Tools / Mic Noise Floor Test
Audio
Mic Noise Floor Test
Observe quiet input level, speech level and clipping with processing disabled when supported.
Free to useNo downloadNo sign-up
Mic Noise Floor Test · At a glance
Quick answer: Observe quiet input level, speech level and clipping with processing disabled when supported. The lowest observed level depends on gain and processing.
Mic Noise Floor Test
The test loads when you open it. Device access is requested only after you choose to start.
Detailed guide
How to use this tool
- Keep the room quiet, start input and observe a baseline before speaking.
- Retain the same input gain for the comparison.
Input
Microphone permission and an input device
Browser requirement
HTTPS; getUserMedia; Web Audio
Processing
Runs on your device; no tool-result upload
How it works and what the result means
Signal level is shown in dBFS, where 0 is digital full scale. It is not a calibrated sound pressure reading. The noise floor is the lowest observed level, not a laboratory SNR measurement.
Reading the result
The lowest observed level depends on gain and processing. A suppressed noise floor does not necessarily mean a quieter microphone capsule.
What to check next
Disable automatic processing when supported and compare the same mic position. Fans, handling and electrical hum can affect the spectrum.
Use HTTPS or localhost, check the selected operating-system input, and close applications that hold the microphone. Stop releases microphone access; recordings remain in your browser.
Frequently asked questions
Is the displayed difference a laboratory SNR value?
No. Signal and noise are not measured with a calibrated reference. Treat it as a setup comparison within the browser.
How do I start the Mic Noise Floor Test?
Start the microphone, approve access, and speak at a normal level. Watch the waveform and spectrum. Use Record and Finish recording to listen to a local clip.
Does this tool require a download or an account?
No. Open the interactive tool above in a modern browser. The instructions and explanations are available even with JavaScript disabled; running a test needs JavaScript.
Why can two attempts give different results?
Power settings, background tasks, browser scheduling, operating-system input processing and device settings affect measurements. Compare repeated attempts with identical settings rather than the single highest score.