Hi all, new user here and pretty new to pi stuff. I have a couple of pi’s running for smart home stuff but i’m pretty much a ‘follow the instructions’ guy at present.
I thought i might have some fun with Rhasspy so i got a Pi3b+ and 2-mic pi hat, followed the manual and have it running, i can catch intents in my Home Assistant setup as well. I had to do the downgrade thing with the mic setup as it failed to work without it.
The issue i have is that its pretty much deaf - if i’m not within about 50cm of the pi then it hears nothing most of the time, playing back the audio reveals a muffled whisper.
Can this be fixed or do i need a better mic board??
You can play around with the alsamixer and try different volume settings, that might help. Also, try the hintak branch of the drivers, they don’t need a downgraded kernel and contain a few fixes. I don’t have a 2mic but from what I read on this forum it should perform better than what you are describing.
Thanks, i’ll try the F6 thing then if no luck will start reading up on how to uninstall stuff i guess, this Linux is like trying to learn Japanese to me
All linux commands -h / --help give you basic paramters.
Often there is the Pi built in audio that for many just confuses and with a hat just turn of by editing /boot/config.txt and put a hash infront so that it becomes a remark
To further confuse even if a device is shared its settings and conf files are totally separate when in a docker container as docker containers are isolated as thats the point.
You need to docker exec -it <container_name> /bin/bash and check and set inside the container
Also don’t use hw:0 indexes use plughw:0 indexes as that enables software autoconversion so any sampling rate or format error will just get converted rather than fail
Don’t use the desktop version of Raspberry as the addition of pulseaudio as default further confuses.
#pcm default to allow auto software plughw conversion
pcm.!default {
type asym
playback.pcm "play"
capture.pcm "cap"
}
#pcm is pluhw so auto software conversion can take place
#pcm hw: is direct and faster but likely will not support sampling rate
pcm.play {
type plug
slave {
pcm "plughw:0,0"
}
}
#pcm is pluhw so auto software conversion can take place
#pcm hw: is direct and faster but likely will not support sampling rate
pcm.cap {
type plug
slave {
pcm "plughw:0,0"
}
}
#sudo apt-get install asound2-plugins
#will use lower load but poorer linear resampling otherwise
defaults.pcm.rate_converter "speexrate"
#Set default mixer controls
defaults.ctl.card 0
sudo apt-get install asound2-plugins to install the alsa-plugins as really the default rate convertor is speexrate if installed or linear if not.
save as /etc/asound.conf for systemwide or ~/asoundrc for user based
Change the index :0,0 to your card
You should see PCMs play & cap listed… But they will be defaults so just use sys default
thanks, will have a play later, the best thing i have learnt so far is that its so easy to wipe the OS and start again on a Pi if I make a mess in there