Tue Feb 18 16:56:11 2020
<9e126bf3> Well that’s a little more complex- but I’d get radio hardware to encode to wav then to ogg or mp3 then stream it with icecast.. then on the other side in the datacenter have processes listening to icecast then pass that audio stream into whatever voice rec / FFT spectrum analysis / deep learning tool to extract info you’re looking for to save