There isn't a nice interface for setting audio parameters yet, but for the moment you can access the variables directly. Changes will take effect in the next synthesis network that you run.
ss> ss-sample-rate 44100 ss> (set! ss-sample-rate 48000) ss> ss-sample-rate 48000 ss> ss-buffer-frames 256 ss> (set! ss-buffer-frames 64) ss> ss-buffer-frames 64
As you can see, the only variable audio properties are ss-sample-rate and ss-buffer-frames. There are some other variables, but they are not currently exported as part of the interface.
audio-inSupport for reading from the sound card will be included in the next release of soundscrape. Until then, you can't access the sound card inputs from within a synthesis network.
Audio output is taken care of by the play and record routines. There should be no need to directly access the audio hardware to output sound.