| name | mic-level-check |
| description | Sample a microphone for a few seconds and report peak/RMS level, clipping, and silence. Use to confirm a mic is actually picking up audio and roughly at the right gain. Quick diagnostic — for full dictation tuning use dictation-mic-tune. |
Mic Level Check
Capture a short clip from a source and report the level so the user can tell whether the mic is alive, muted, too quiet, or clipping.
Steps
- Resolve target source: default with
@DEFAULT_SOURCE@ unless the user specified a device (then resolve the ID/name via list-audio-devices).
- Confirm it's not muted:
wpctl get-volume <source-id>
- Record ~3 seconds:
parecord --device=@DEFAULT_SOURCE@ --record-time=3 --file-format=wav /tmp/miccheck.wav 2>/dev/null
(Or parec | sox if parecord isn't available.)
- Analyse with
sox:
sox /tmp/miccheck.wav -n stats 2>&1
Capture: RMS lev dB, Pk lev dB, Min/Max amplitude.
What to report
- Peak dB: ideal for speech is around -6 dB to -12 dB. Above -3 dB → risk of clipping. Below -30 dB → too quiet, raise gain.
- RMS dB: ideal -18 to -24 dB for speech.
- Silence detected: if peak is below -50 dB, mic is probably muted, broken, or has wrong device routing.
- Clipping: if
Max amplitude ≥ 0.99, lower gain.
Suggested fixes (offer, don't auto-apply)
- Too quiet → raise volume with
wpctl set-volume <source-id> 1.2 (Pipewire allows >1.0 software gain up to ~1.5 cleanly).
- Clipping → reduce volume; also check whether a hardware boost is enabled in
alsamixer.
- Silent → check mute state, check
pavucontrol Recording tab for app routing, check whether the right Bluetooth profile (HSP/HFP) is active.
For a richer dictation-tuned workflow (RNNoise, gain staging, persisted source), use dictation-mic-tune.
