Quering an audiofile: What do you hear? What do you see? What do you feel? What don't you hear?