noise-suppression-for-voice
noise-suppression-for-voice copied to clipboard
Add ALSA config example
After some trial and error I finally got the LADSPA plugin to work with ALSA (without PulseAudio).
Here's a snippet for the capture PCM device, to use in ~/.asoundrc or your preferred alsa conf file:
[...]
capture.pcm {
# Add an ALSA plug for LADSPA
type plug
slave.pcm {
# Add the LADSPA noise filter
type ladspa
slave.pcm {
# Convert from float to int
type lfloat
slave {
format "S16_LE"
pcm "hw:1,0" # Use card 1 (e.g. USB webcam soundcard), device 0 (the default)
}
}
# LADSPA configuration for the noise filter
# See https://github.com/werman/noise-suppression-for-voice
path "/usr/lib/ladspa"
capture_plugins [
{
label noise_suppressor_mono
input { controls [ 50 ] } # VAD Threshold %
}
]
}
}
[...]
Maybe it could be helpful to add this to the How-to, under Linux.
Thank you! I'll add this snippet.
The above is an absolutely golden example as had been battling away and would never of got type lfloat
Had to rename capture.pcm to pcm.capture compiled on a Pi4 running @ 2.0Ghz.
Had to drop the VAD sensitivity to really low values though.
no-rnnoise https://drive.google.com/open?id=1pIH5O_TP6YoNrp9ql2rr_t5LmnpMB9QE rnnoise https://drive.google.com/open?id=1_Qr-XaaaxEy-nQeiS8GaTVCWCd1egde_
Its actually a good attempt by the pi4 running pios 64bit lite @ 2.0Ghz There is prob some optimisation and neon use that could provide much better results as its very good for what it is. Have a hunch though the Aarch64 compile may not work as well as x86_64
PS pcm.capture {
was all I had to change