Is there much progress out there in audio processing algorithms to remove lead vocal from music tracks to produce a karaoke, or the inverse, stripping out everything but the lead vocal to produce an acapella?