This is definitely possible as I’ve done it with a Tascam US428 4ch usb Interface. A good place to start is to run the commands aplay -l [this lists your interface options] and alsamixer [which can I help you relate names to interfaces as well]
They are usually sub indexed, something like 0,0 or 0,1 Depending on your configuration and what audio packages you have installed already, this can be a little bit of a jumping around game to figure out, but it will definitely work. As far as the audio quality goes, I found some quirks with the ALSA System settings when you access them at a lower level. For example, on a built in laptop soundcard I was able to set the audio input to sample at 48 kHz from the flowgraph, but on a raspberry pi 2 with a USB audio input card I could only stream at 96K. The thing is, flowgraph would still run on the raspberry pi, it just sounded bad because of the mismatch between the flowgraph and hardware. If I remember correctly (bear in mind that it’s been a while and I’m pretty sure there is documentation somewhere in the GNURadio docs for this) the syntax for audio cards has a relationship to how much of the intermediary driver you’re using (bare metal/kernel, simple system audio control, audio control with effects and routing) I think the “hw:0,0” refers to the lowest level stream on the hard work hard, Whereas “plughw:0,0” I think goes through your operating systems internal Audio mixer. Then you elevate to pulse from there. In the end, for me, some of it was just a matter of figuring out what my hardware supported and trying different things. The good news is that a new USB audio card costs orders of magnitude less than a new USB Software Defined Radio! Hope this helps! <end transmission> > On Aug 28, 2021, at 14:12, Barry Duggan <ba...@dcsmail.net> wrote: > > > I would like to make a flowgraph with two separate (independent) audio input > ports and two separate (independent) audio output ports. Is it possible using > Audio Source and Audio Sink blocks and/or using direct file I/O to /dev names > (bypassing PulseAudio)? I've tried a scheme on two computers using ZMQ > blocks, but the audio quality was very bad (I don't know why). > > Thanks! > --- > Barry Duggan KV4FV > https://github.com/duggabe > > >
