I'm trying to generate and stream audio+video via aconv to a remote location such as Youtube from a headless linux server, which will then be multicasted out to the internet at large. I have successfully been able to stream audio and video from a headless Ubuntu 13.10 server hosted locally in VMWare. I use Xvfb to create a virtual frame buffer for the video and the audio was done typically with ALSA and Pulse.
However I now need to do the same on a hosted VPS (Dreamhost) that does not have virtualised hardware such that VMWare implements. I have been able to successfully stream the video but as there are no audio devices in the VPS I cannot input or output any audio.
Is it possible to create a virtual software-based sound device that applications can send audio to and aconv can read from to add to the stream? If so, how? If it can be done with Pulse that would help too.