API access to progressively decoded audio data

[Splendide-Imaginarius]

As best I can tell, chunks expose the decoded audio data to API consumers, but must decode the entire file first; music can decode the audio in real-time but doesn't seem to expose the audio data to API consumers. It would be nice to get something in-between, where an audio file could be partially decoded, and API consumers could extract the so-far-decoded audio data. Sort of like what SDL_sound does.

Maybe there's already a way to do this, but if so, I failed to find it in the documentation.