Better sandboxing, basically. A system was needed for video, turns out it was a good fit for audio. Audio and video aren't that different, TBH (audio just has

Better sandboxing, basically. A system was needed for video, turns out it was a good fit for audio. Audio and video aren't that different, TBH (audio just has more alpha/blending rules, and lower tolerance on missed frames; video has higher bandwidth requirements). Wouldn't surprise me if both pipelines eventually completely converge. Both "need" compositors anyways.