Text this: Computational auditory scene analysis :