Cubase 16 adds six-part stem separation and AI vocals

Steinberg has released Cubase 16, expanding its digital audio workstation with new tools aimed at producers, remixers and musicians who would rather spend time making music than navigating endless menus. One of the biggest additions is improved stem separation, developed with audio technology company zplane. 

Cubase can now split a mixed recording into six components: vocals, drums, bass, piano, guitar and other instruments. The previous version supported four, making the two additional instrument categories particularly useful for remixing and rearranging existing material.

The update also introduces Instrument Stacks, allowing up to eight virtual instruments to be combined and played from a single track. New MIDI tools can make programmed chords sound more like natural strumming, while redesigned tempo detection and searchable settings aim to speed up everyday production tasks.

Yamaha's Omnivocal Basic vocal synthesizer is another notable addition. Using the new Omnivocal NX engine, it generates singing from MIDI notes and lyrics entered directly in Cubase's Key Editor. Basic includes two voice banks, while the separately priced $149 Omnivocal Pro expands the selection to ten and adds vocal stacking for layered performances.

Cubase 16 is available in three editions: Pro ($599.99 / €599), Artist ($299.99 / €299), and Elements ($99.99 / €99), through the Steinberg online store. Omnivocal Basic is included for free, while Omnivocal Pro costs $149.99 separately. Customers who activated a Cubase license on or after October 7, 2026, qualify for a free update to the corresponding Cubase 16 edition.

Stem separation and vocal synthesis are hardly new concepts, but bringing them closer to the actual production workflow is the interesting part. Whether the results sound convincing is another question, particularly when the AI is asked to sing something a human probably shouldn't.