Steinberg SpectraLayers 12 allows musicians to separate music into individual tracks and vocals to improve poor recordings

Steinberg has launched SpectraLayers 12, audio editing software that can separate music back into individual instruments and vocals, according to the company. The tracks can be reused to mix new versions, edited to correct audio issues, and processed to improve poor vocal recordings.

SpectraLayers leverages AI to help editors select desired or problematic sounds using a spectrograph view of audio tracks. The spectrogram displays time along the horizontal axis, frequency along the vertical axis, and amplitude as colors.

The AI can recognize common instruments, such as a guitar, drums, piano, or sax, and can be custom-trained on new instrument samples for automated track separation. Manual selection is also aided by AI, which can automatically select both the primary frequency of an instrument as well as its harmonic frequencies.

This potentially allows users to quickly identify sounds, such as the low rumble of an air conditioner or the sound of a cymbal, and remove them from recordings. Separated sounds and vocals can also be reused, edited, or processed to remove background noise. Features like the Voice Enhance module use AI to generate lost details in poor vocal recordings, which might help in the restoration of old movies and analog recordings, or save a poor live recording from being unusable.

Steinberg SpectraLayers 12 Pro with all the latest features has an MSRP of $349.99. SpectraLayers 12 Elements, a version with far fewer features, is available for $89.99. Both are available from the company website and authorized resellers.