Doorpi - audio from raspberry pi really silent on other side (webrtc)

  • Hi all,


    I hope English is not a problem since my German isn't that good.


    I succesfully installed doorpi and asterisk. I have two-way audio working, but for some reason, the audio on my intercom station is really silent (but coming through). My setup is as follows:

    • asterisk in docker container running on NAS
    • webrtc client (jssip) on a tablet (the intercom)
    • Doorpi installed on raspberry pi zero w Raspiaudio hat


    I can succesfully record with arecord and replay with aplay, the sound quality of the raspiaudio is excellent. However, during SIP calls, the sound on the intercom station (a tablet with jssip) is too silent to be usefull. I doublechecked if sound was coming through by installing CAVA (Console-based Audio Visualizer for Alsa) and it is. Any idea's what could be the cause?


    My trace logs are on Pastebin: https://pastebin.com/cdXexHzM


    I would love to use my Doorpi! Thanks again!

  • Problem solved (but I don’t understand it though)!!


    Putting the answer here as perhaps it may help fellow googlers:
    So, in the doorpi configuration one must specify the capture device (so usb microphone, …), which is then used by linphone.
    If I do cat /proc/asound/cards this gives me the following:

    Code
    pi@raspberrypi:/usr/local/etc/DoorPi/conf $ cat /proc/asound/cards 
    0 [sndrpigooglevoi]: snd_rpi_googlev - snd_rpi_googlevoicehat_soundcar                           
        snd_rpi_googlevoicehat_soundcard


    According to the documentation, it should then be as follows:
    capture_device = ALSA: snd_rpi_googlevoicehat_soundcar


    This sometimes works, but if it works, the recored audio is way too silent. Off course I double, tripple and 15-pple checked alsamixer if my recording level was ok.


    The solution was to put:
    capture_device = ALSA: default device.


    Now, I’m having audio from the pi zero w to my browser running jssip, confirmed by CAVA. I’ll test the full-duplex audio tomorrow.


    Can anybody tell me the difference between: ALSA: snd_rpi_googlevoicehat_soundcar and ALSA: default device?


    Some keywords so german speaking people will find the solution: microfon leise still