About us
MicTester is an independent, single page online mic test. It is maintained by one person, not by a microphone manufacturer, a headset brand or a browser vendor. The person behind it has years of experience building and running products, spent a long time as a product manager at a leading internet company, and has always been into technology, gadgets and the web. Over the years that has meant helping to plan and ship hundreds of projects, as well as developing and running projects alone. We like trying new technology and new products, we start from what people actually need, and we keep working to make things useful, easy to use and worth trusting.
Last updated: 2026-11-14What this site is
This page exists to answer one question quickly: is the microphone delivering a usable signal, and what does that signal look like? Everything needed to answer it is on the homepage. There is no blog, no account area and no separate tutorial pages for each operating system.
The tool is free to use. It runs entirely in the browser using the Web Audio API and the media capture interfaces that your browser already provides. No server component receives audio.
Why we built it
Built in checks inside meeting and recording apps usually stop at "microphone detected". That tells you almost nothing when the actual problem is a low input level, a noisy room, a Bluetooth headset stuck in call mode, or auto gain quietly changing the signal. A mic test that shows the level, the waveform, the spectrum and the noise floor gives you something concrete to act on.
We also wanted the check to be private. A microphone permission is a sensitive one, so the shortest honest answer to "where does my voice go" should be "nowhere outside this tab".
How the readings are produced
Peak and speech level come from time domain samples of the incoming stream. The noise floor is estimated from the quietest stretches, signal to noise ratio is the gap between speech level and that floor, and the clip counter marks samples at digital full scale. The spectrum is a fast Fourier transform of the same samples. Sample rate, channel count and input latency are reported by the browser audio context.
All values are in dBFS, a digital scale relative to full scale. They are practical guidance, not calibration. Recommended ranges on this page are -30 to -6 dBFS for speech peaks and below -45 dBFS for the noise floor in a quiet room.
What this site is not
It is not a sound level meter. Browsers expose dBFS, not dB SPL, and a web page has neither the microphone sensitivity nor a calibrated reference needed to convert between them. It is also not a frequency response measurement, a channel balance test, or a substitute for lab equipment.
We do not describe this site as an official product of any hardware or software vendor. Device names shown in the microphone selector come from your operating system, not from us.
Contact
Questions, corrections and bug reports are welcome. Write to ha0719#gmail.com (replace # with @). We do not run a support ticket system, so a short description plus the browser and operating system you used is the fastest way to help us reproduce an issue.
We do not ask for audio files. Please do not send recordings; the readings and a description are enough.
Corrections and updates
If a published explanation turns out to be wrong, we correct the page and update its last updated date rather than silently editing it. The changelog keeps a short note on the first release.
The content production and measurement method behind this page is summarised in the "Method and references" section on the homepage.