Cool! Wonder if that solves your issue with ALSA / Portaudio. Linux and audio back-ends has (in my experience) never been really good or stable.
From what I can see the issue is related to pulling (recording) audio. A good fix could be to disable audio input completely. I just submitted a patch (nap_source) that lets you set the channel audio input count to 0, which in turn disables audio input completely.
You can configure the audio service using a
'config.json' file next to your executable. For example:
"InputChannelCount" : 0,
"OutputChannelCount" : 2,
"AllowChannelCountFailure" : false
"HighDPIMode" : true,
"DoubleBuffer" : true,
"EnableMultiSampling" : true,
"MultiSamples" : 4
As you can see I configured both the sound and render engine using 2 configuration objects. Right now you have to create the config.json file manually, but we want to embed it in Napkin for easy visual configuration later on.
InputChannelCount to 0 should result in no audio input stream and a possible fix. That way you don’t have to remove video, since it only outputs stereo audio.