I've been using darkice with a soundcard source but am moving to a setup where I'll be recording video while broadcasting live audio to shoutcast. Ideally I'd take the audio from the video to save additional connections. My plan is to use dvswitch for the video, and use dvsink-command to take a stream out. The example talks of streaming to icecast using ffmpeg2theora|oggforward. With darkice I've two questions: 1) how do I set up darkice to take the piped input (no need for named pipes of course). 2) What audio format should I get out of ffmpeg? Any help much appreciated. Nathan