At first I thought Omnivocal was fairly decent or at least usable to convey a song, but I haven’t listened after a couple of updates and I don’t know if that affected but for example, the first word was Lost, over 3 beats and all I hear is about a 16th not La. Then Lone, almost 4 beats is just La (choke), like a 16th, and lots more cutouts. Other things sound almost like hiccups.
I haven’t seen any such issues. The only times I’ve seen sounds getting dropped were on very short notes, in which case making them a little longer would make a difference. (I think the details related to the specific ending and starting sounds for the short word and the word that followed it, respectively.)
If I’m understanding your symptoms correction – i.e. start of a long syllable (i.e. in turns of duration of its note) coming through but then not the rest of that same syllable – it makes me wonder if maybe you have a gate on the track, where the initial part of the word is loud enough to cross the threshold to open the gate, but then it quickly goes below the threshold, and the gate closes down again, possibly with a fast release time (and no hold) that ends up around a 16th note in duration.
It’s nearly unusable. Completely subpar compared to other similar software. Just another gimmick to to try and sell a lackluster upgrade, as we all have come to find out. Steve will probably delete my comment/opinion again, but it’s the truth.
My only hands-on point of comparison on the “similar software” front was trying Yamaha’s Vocaloid back in 2004 in the course of writing a review for the now-long-defunct CakewalkNet ezine. Omnivocal is a major improvement, in both quality of results and workflow, over that early version of Vocaloid.
While I’ve read a bit about some of the more modern alternatives (including the latest Vocaloid among more popular current offerings like Dreamtonics Synthesizer V), I hadn’t really been in the market for anything of that sort. Thus, when Steinberg included Omnivocal in Cubase 15, it was more of a curiosity, than something I really expected to use. However, I was wrong on that front.
I’ve used Omnivocal on every single song I’ve recorded since I first tried it. With one exception, where I used it to double my background vocals (which I’m unlikely to attempt again due to the major pain in cleaning up MIDI extracted from my vocals’ audio sufficiently well to be able to use that in practice), Omnivocal parts weren’t used in the final recordings. Rather, they were used for placeholder lead vocals (and sometimes background vocals) while I was working on the instrumental arrangements and tracking. (In one case, I was also using it for that same purpose while still writing lyrics to the song, making it way more efficient to just edit lyrics in Key Editor than having to set up a mic and record real vocals to let my cowriter hear how my lyrics were fitting with his music.)
Using Omnivocal this way was a big improvement over how I’d filled this function in the past – basically picking some sort of vocal pad or other vocal-like sound doing oohs or ahs. Not only is it helpful having the lyrics, but the Omnivocal voices come a lot closer to filling the same sort of spectral space that my own vocals would eventually fill once I get the arrangement far enough along to track my parts.
As for the doubling my own background vocals with other voices function, I’ve found that to be way more convenient, and with better results, using IK’s ReSing. There’s no need for extracting MIDI from audio there since it just coverts my own vocal to another voice. But Omnivocal has been valuable as a workflow aid in the arrangement stage of my recent recordings. And, in that context, I really don’t care much about most of the quirks that require some diving into the phonemes.
Hello,
Just a couple of thoughts…
What is the Buffer size you are working with? Have you tried increasing the Buffer size?
Also, each voice on its own output channel? Are there overlaps of the notes?
Sorry, but that is all I can think of at this time.
Regards,
KKSky
Buffer size is 1024. There was some compressor on, but I removed and also routed to a master with no effects. It sounds like a lot of hiccups. Simpler than many other projects, only 24 active measures with one voice. A more involved project that I checked didn’t have this happening. And while comparing an earlier save, that had the same problem, Cubase just closed, 3rd time today.
That sounds like a potential clue – maybe some project corruption and/or some plugin that is causing Cubase to crash??? (If you run Windows, and there is a DMP file generated, that might potentially provide some clues as to what is causing the crash. Similar thing for Mac, but it would be some other file type.)
Do you have an earlier save that did not have the problem? If so, once you verify that, maybe you could remove the track from the current project and import it from the earlier save without the problem.