Beta tester here 
Windows 11 i9 lot of ram. Omnivocal beta produces nasty crackles when midi-part is muted (but instr track isn’t). Also, when you do have a un-muted midi part after a muted part, the Omnivocal Beta breath-in sounds like Darth Vader btw.
Same crack when you put negative track delay (ms) to the Omnivocal midi part.
I get static as soon as I click on the first note in editor and the whole track is very staticy when I play it back with text and female voice
Hello,
I upgraded to Cubase 15 Pro mainly because of the new Omni Vocal feature, and I have to say I’m very disappointed with its current performance.
In my experience, Omni Vocal is noticeably noisy and introduces artifacts, especially on exposed vocal tracks. The processing sounds grainy and unnatural, even when used very lightly. More importantly, it does not properly distinguish between male and female voices or adapt well to different vocal ranges, which was one of the main promises of this tool.
I’ve tested it on clean, well-recorded vocals with proper gain staging, and the results are still far from professional or production-ready. Instead of enhancing clarity, it often degrades the vocal quality. At the moment, it cannot replace existing professional vocal tools, and I find myself avoiding it entirely.
As a professional musician and producer, this is frustrating, especially since Omni Vocal was a key reason for upgrading to Cubase 15 Pro. I really hope this feature is taken seriously and improved through updates, because in its current state it does not meet professional expectations.
I would appreciate knowing if fixes or improvements are planned, as this feature clearly still needs work.
Thank you.
Louis Toteda
Well, I guess you’ve learned a valuable lesson — at a reasonable price, I’d say. Personally, I didn’t upgrade because of OmniVocal. That Instrument is just a bonus I hadn’t counted on. But you’re right: its practical applications are still somewhat limited.
I’ve had some good results using the female voice to thicken or create additional parts in background vocals. I’ve also used OmniVocal to experiment with vocal timing. It works well as a guide track when envisioning vocal arrangements or scores — my own way of arranging/composing accompaniment changed, since it already gives you a rough idea of how its sounds with a vocals in there. The OmniVocal male voice is problematic so far, but less problematic than my own singing voice
I think OmniVocal is just a starting point, and Yamaha might be exactly the right company to develop something like this. I remember: the FM synths (my old DX patches still load in FM Lab), the physical‑modelling synths the TG series, the early digital pianos, their DSP innovations. What am I forgetting? Real instruments: the superb trumpets, the clarity and quality of their acoustic pianos and grands, which I often preferred over the famous ‘German’ brands depending on the application. The wood work of their guitars. Still more? Silent orchestra, silent brass, the digital saxes that saved many people from arguments with their neighbours, and so on.
At least the crackles are gone with the new Omivocal beta 1.1.3 (that I had to download manually.) Key editor solo track behavior is still non compliant.
Anyways here’s an little omni vocal timing experiment, some old medieval street song, some other guy wrote an arrangement, and I let AI come up with a new English text loosely based on the old German : https://www.hamicast.nl/hans/ytresources/bachomni.mp4