Interpret dBFS, SNR, latency and distortion only within the route and processing settings shown in Evidence.
Microphone Level and Spectrum Test
Inspect live RMS dBFS, peak, noise floor, waveform, FFT and processing flags.
A microphone is silent, too quiet, clipped or sounds processed.
This focused entry uses the shared Audio protocol. The measurement logic and Evidence version remain identical to the complete lab.
Remove the easiest false alarms first.
Close calls and recording apps, select the intended route and keep system gain fixed.
Use a moderate safe volume and avoid changing microphone distance during a protocol.
Treat acoustic and confirmed fixed-gain loopback sessions as different measurement paths.
Run one repeatable protocol.
- 01
Grant permission and select the intended input.
- 02
Speak at normal distance while watching dBFS and waveform.
- 03
Inspect processing flags and the local spectrum.
Evidence the result should expose.
Separate observed, review and inconclusive.
Repeat with fixed gain, then compare wired, Bluetooth and direct loopback paths one variable at a time.
Acoustic measurements include speaker, room, microphone, OS and browser; they are not device specifications or hearing diagnoses.
Start with the likely path, then retest.
Wrong input
Gain staging
Browser processing
Room noise