Here are seven audio files containing examples that demonstrate various elements of the discussion within the Cubase workshop article published in Sound On Sound August 2026.
www.soundonsound.com/techniques/cubase-15-rebalance-remix-and-resample-made-easy
These are all based around a verse/chorus extract from a basic electronic pop track featuring lead vocals, backing vocals, electronic drums, synth bass, and various synth-based keyboard sounds. With no great claims about the quality of the original song/track(!), it serves as a suitable vehicle to explore Cubase Pro 15’s new stem separation features
Cubase 15 Stem Separation Audio 01.wav
This example plays through the song extract a number of times, starting with the original stereo mix, before hearing an instrumental version (vocal stem removed) and then each of the stems — vocals, drums, bass and ‘other’ — in isolation.
Cubase 15 Stem Separation Audio 02.wav
or comparison, in this second example, you can hear exactly the same sequence of audio sections but resulting from Steinberg’s SpectraLayers Pro 12’s stem separation used at its highest quality.
Cubase 15 Stem Separation Audio 03.wav
For comparison, in this third example, you can hear exactly the same sequence of audio sections but resulting from iZotope’s RX 12 stem separation used at its highest quality.
Cubase 15 Stem Separation Audio 04.wav
This audio example demonstrates mix rebalancing based upon the Cubase Pro 15 generated stems. You hear the original stereo mix for reference and then two alternate ‘rebalanced’ versions. In the first of these, the vocal stem has been raised by 3dB, while the drum stem level has been automated to gradually increase (by a maximum of 4dB) over the length of the extract. In the second, the vocal level has been decreased by 2dB while the bass stem has been raised by 3dB. These level changes have been made simply to illustrate the potential, rather than with a view to improving the original mix.
Cubase 15 Stem Separation Audio 05.wav
This audio example demonstrates remixing based upon the Cubase Pro 15 generated stems. Two different remixes can be heard, both taking just the vocal stem from the original track and placing it in a different musical backing. In the first, the original 135 BPM tempo is retained, and new drum, bass and keyboard parts have been generated to create a different instrumental backing. In the second, the tempo has been reduced to 120 BPM (so the vocal stem has been subjected time stretching) and it has been layered with a solo acoustic piano backing. From the perspective of audio quality, first remix obviously works better, and in some use-cases, may provide an acceptable result. In the second, audio artefacts within the resulting vocal stem have been exaggerated by the time stretching, although the audio quality might be less critical if you are simply undertaking a remix experiment to see which direction you might wish to take the project in.
Cubase 15 Stem Separation Audio 06.wav
This example explores the possibility to repitched the vocal melody using VariAudio applied to the verse section of the vocal stem. For reference, you hear the original vocal stem in isolation. The same vocal phrase is then heard but raised in pitch by three semitones using VariAudio. Finally, the same vocal, back in its original key, can be heard with VariAudio used to adjust some individual notes (essentially re-writing the melody). In purely sonic terms, this is perhaps the least satisfactory experiment, and this is not helped by the fact the vocal stem contains multiple voices. However, used simply for proof of concept when considering changes to the key and/or melodic aspects of an existing project before undertaking new recordings, it might be useful.
Cubase 15 Stem Separation Audio 07.wav
This example explores the possibility of resampling sounds from a stem. In this case, I’ve targeted the bass stem, adding it to a Sampler Track, isolating one of the clearer individual bass notes within the stem, and then making suitable adjustments within the Sample Track to create a playable bass sound. You can hear the song extract twice; first with the original bass stem and, second, with a new bass part performance via the Sampler Track instrument and replacing the original bass stem. Clearly, this is a simple example but it demonstrates the principle and it could obviously be applied in more experimental or creative ways including drum or vocal resampling.