These three audio examples demonstrate the Omnivocal techniques described in the Cubase workshop article that appears in SOS February 2026.
www.soundonsound.com/techniques/cubase-15-exploring-omnivocal
Cubase 15 Exploring Omnivocal Audio Example 01.wav
This example demonstrates using Omnivocal to create simple vocal ‘hooks’ that might be used as an ear-candy element within an electronic/pop production. The example plays through a short section of a simple electronic pop backing track and, in each case, Omnivocal has been used to create a hook-style vocal part. The four variations are as follows:-
1) In this case you simply here the raw Omivocal performance with no other processing applied. The female voice style has been used. In the first four bars, the voice is simply using ‘do/da’ vocalisations while in the second set of four bars, simple lyrics have been added.
2) In this variation, you hear exactly the same vocal track but with some suitably catchy vocal processing applied.
3) In this variation, the same Omnvocal vocal has been passed through a vocoder/vocalsynth processor in a typical OTT electronic music fashion.
4) In this variation, two further Omnivocal parts have been added, one using a further female voice and the other with a male voice. These are panned within the stereo image to give the overall vocal element greater impact.
Cubase 15 Exploring Omnivocal Audio Example 02.wav
This example demonstrates using Omnivocal to create backing vocal and harmony vocal parts to sit alongside an existing, human sung lead vocal. The example plays through an eight-bar section of an electronic/pop track song demo. In each instance, you hear a different combination of the backing track, lead vocal and Omnivocal supplied backing/harmony vocals as follows:-
1) The first playback features just the basic backing track (drums, synth bass, and two synth parts) and the (human) lead vocal (dry, without any processing applied).
2) In the second playback, you hear the same lead vocal alongside two instances of Omnivocal (one male, one female) that are supplying simple backing vocals to double the vocal phase in selected bars. The Omnivocal parts are panned to give a sense of width. In this case the backing track has been muted to make it easier to hear the details of the three vocal parts. All the vocals are dry.
3) The third playback is based upon the same parts as the previous example, but the backing track has been added and all three vocal parts (one human, two Omnivoice) now have some basic processing added (a combination of compression, reverb and delay).
4) This follows the pattern from the second playback but, in this case, two additional instances of Omnivocal have been added (again, one male and one female) providing simple 3rd above harmonies. These were easily created by simply adjusting the MIDI notes for the respective parts.
5) As above but with the backing track added and compression, reverb and delay processing applied to all the Omnivoice vocal instances.
6) Following the same pattern, two further Omnivoice instances have again been added but this time the MIDI data has been adjusted to provide an element of counter-melody to the lead vocal and other backing vocal lines. Again, the backing track is muted so you can hear the vocal lines more clearly.
7) As above, but with the backing track added and suitable processing on all the vocal lines. Things are perhaps starting to get a bit busy, but this is ‘pop’ and, for the purposes of demonstration, I’ll allow myself some artistic licence!
Cubase 15 Exploring Omnivocal Audio Example 03.wav
This example demonstrates using Omnivocal to create lead vocal lines within the context of a songwriting process. A 16-bar section of a rock/pop song demo (just basic drum, bass and guitar parts at this stage) provides the context. This is divided into two 8-bar sections with contrasting intensities. This backing track can be heard four times and, in each case, a single lead vocal part can be heard (no doubles or backing vocals added) but the lead vocal does have some basic processing – compression, reverb and delay – added to sweeten it a little within the rough mix. The four different lead vocal versions are as follows:-
1) Omnivocal’s female voice character is used in this first example. Initially, I recorded a MIDI-based melodic line. This was then used as a basis to add the lyrics. This proved to be something of an iterative process to try and get the best balance between pronunciation and a smooth delivery. Even so, there are sections where I’m sure I could have achieved a better result with a little more experience with the instrument. For example, there are certain words where the pronunciation is not quite correct. Equally, there were some words that I wished to stretch out over multiple pitches, and I couldn’t always manage that successfully. In all these cases, I simply. adjusted the note pitches, lengths and lyrical content to achieve the best result that I could. I also applied some automation to try and make the character of the vocal reflect the low intensity/high intensity contrast of the of the two song sections. While the final vocal is serviceable as a ‘scratch vocal’ to capture the basic idea, its synthetic nature is very obvious, particularly in the quieter, less intense, first 8 bars of the song idea. Delivery in the more intense section seems better to my ear and this might be because there are fewer sustained words within this section of the performance and that expose the synthetic nature of the vocal.
2) For the second pass, I simply copied the MIDI from my first Omnivocal instance onto a track with a second Omnivocal instance but, this time, using the male voice character. I made some minor changes to accommodate the different nature of this male voice (for example, adjusting some note lengths) but then added some further minor changes to the melody and lyrics in a few locations simply to demonstrate how easily this can be achieved (although not necessarily because they improved the composition!). The ability to re-write your vocal line while the song writing processes is still ongoing is one of the main attractions of the Omnivocal concept. I found it somewhat easier to achieve a smoother delivery with the male voice model compared to the female model.
3) For the third (and fourth) play through, I copied the MIDI from my first Omnivocal track and imported it into an instance of Dreamtonics Synth V. I selected a female voice bank (Natalie – based upon US session singer Natalie Nicole Gilbert) and then recreated the lyrical content within the Synth V editor. This is done in a similar way to Omnivocal but the user has much more control over the pronunciation, timing and emphasis placed on particular phonemes. I then added a few additional melodic embellishments to exploit the additional pitch control Synth V provides and, as with Omnivocal, applied different vocal ‘characters’ to provide performance contrast between the two sections of the demo. The result is obviously more polished than Omnivocal is currently capable of but then Synth V is not included as a freebie within Cubase 15!
4) For the fourth play through I simply added a further Synth V instance and copied the MIDI/lyric data over from my ‘Natalie’ track. I then selected the Hxvoc (based upon US rock singer Seann Nicols) voice bank and went through a fairly rapid round of minor editing and adjusting the vocal character parameters to adjust the style of delivery. Again, the end result is more polished (and realistic) than was currently possible with Omnivocal and, for this particular song demo, the style of delivery was more what I was imagining in my head when I started the idea. That said, I’m impressed with the initial Omnivocal beta release and especially that it is available for all Cubase 15 users, Elements and upwards. It will be interesting to see if future Omnivocal updates can build on the initial potential by adding the kinds of additional performance and voice character options software such as Synth V and ACE Studio are currently capable of.