Steinberg SpectraLayers 12 leverages AI audio separation to recreate individual instruments and vocals from music for editing within a spectrograph-based interface. It can help audio editors improve song mixes, especially for live or poor recordings.