Audio support

This seems like a good time to suggest that there should be some commonality and interoperability between AudioIO and VideoIO.

For instance, using VideoIO to strip out the audio should return the same sort of audio object as AudioIO gives you when you open a file. Currently that's an AudioFile but there should be something more abstract.

Another example would be if you set up a signal processing chain using AudioIO you could apply that to the sound track of a video, either during realtime playback or as an offline rendering operation.

There are lots of pieces in the AudioIO API that are either not yet designed or still changable, so this is a good time to make sure we're aligned.

There isn't any particular rush on this from my end, but if folks want to chat about how those interfaces should look I'm here.

The issue identifies AudioIO, VideoIO, and AudioFile as the relevant interfaces; start by comparing how each represents audio and how VideoIO exposes a video's soundtrack. Done would require an agreed abstract audio representation and compatible processing workflows, but the issue does not specify files, tests, or a final interface.