Hi all,
I would like to use multiple instances of pulseaudio to capture applications' sound outputs (without redirecting it to the local speakers) and to stream them on the network, and I would like to hear your comments.
My idea is to create multiple pulseaudio daemons (one for each user on the host) and to start sound-aware applications for each user using a separate pulseaudio instance (maybe using padsp in the command-line)
Then I would like to "forward" all the audio streams to a second server, that could maybe store them to disk or play them through the speakers). I read that pulseaudio supports RTP multicasting (with the module-rtp-send module), but I think this will not work in my case.
How could I forward all the audio streams to a given server?
How could the listening server get the "user-id" for the streams? (the server has somehow to match the streams with the users)
In case I would like to program the backup server on my own (maybe in Java), what encoding/container options are supported? (For my needs uncompressed PCM would be enough, since the local LAN is fast enough for several streams)
Any advice or comment is welcome
Thanks a lot
Best regards
Andrea
