I can’t find an Omnivocal phonetic for the sound of the letter U as heard in the words Cup or Cut. The manual suggests to use ah for this but that sounds wrong - making Cup sound like Carp at Cut sound like Cart. Has anyone found a way of generating the right sound for this? Thanks.
Maybe try something like this (for “cut”):
The “!” in front of the t makes it clear to pronounce the t seperately, so that it can be heard but it is probably not necessary.
Thank-you Johnny, that does sound better, I didn’t use the exclamation mark.
On a separate issue, it seems the velocity controls in the piano roll don’t work with Omnivocal. Unless I’m missing something?…
1 Like
I came to the same conclusion. It seems we have to automate the volume or the expression parameters of the plugin.
That’s a shame, note by note volume control would be a good feature emulating something more akin to how a human sings. Of course Omnivocal is still in Beta, maybe Yamaha could include it when the program is released proper? I’d vote for that.
Unfortunately, Omnivocal does not implement the full International Phonetic Alphabet, only the phonemes used in the Japanese language, which can do some level of English words, but not all of them.
It would be, but if not, I keep all note on velocities at 64, to be aligned with note off velocities at 64, by default.
This happens, with instruments like organ or harpsichord however is the Expression conroller, CC11, implemented or the equivalent NRPN value?
I was doing some look around and I have suspicions Omnivocal is a feature limited version of Yamaha’s VX-Beta, which is a product that’s free for any Vocaloid 6 owner and integrates with Cubase exactly like Omnivocal does.
No that IS interesting. For me the $225 purchase price for Vocaloid with the VX free bundle would be a worthwhile investment once both programs are out of Beta. With Cubase and Vocaloid being under the Yamaha umbrella no doubt integration between the two would be seamless. Thanks for posting the link Bruno, interesting and appreciated.
So what is/are the difference(s) then?
I use Emvoice One, and it is great at what it does however, but it does create system load, simply being enabled due to network activity so if there is a way to only need the DAW for vocalisation, then I will look at that even though there are only 2 voices currently available.
Unlike Omnivocal, VX-Beta has multiple voice banks and more settings to tweak.
Omnivocal does not seem to be connected to any cloud service and it’s quite fast, taking only a few seconds to do changes. It’s about as fast as Vocaloid 6.
I have tried Synthesizer V, and I am hesitant with using multi-lingual voices, for English singing, since I don’t like the idea of using a non-English voice, for example, to create English words, as sung.
Emvoice is all in English, is easy to use, and has enough voices for my purposes.
I know in Synthesizer V, there is a selection for language, but the native tongues are mostly Spanish I believe so are any voices in Vocaloid, or the beta that are actually English native singers?
I guess it is how far the programmers wish to take it. The British English singing voice for instance has a different vernacular than an American English voice, more pure in my opinion. Unless the programmers start getting into the realms of emulating a specific artist’s voice (I certainly hope they don’t) I think it is more fun to work with a limited range of voice banks, so if artificially generated voices have something of an accent I’m OK with that.
The thing is the “virtual vocalist” idea pretty much comes from the Japanese when they made the first Vocaloid, which was not Hatsune Miku like a lot of people think. Then you had VirSyn’s Cantor, which came 4 months later even though the first Vocaloid engine was developed by VirSyn. Hatsune Miku only happened 3 years later with Vocaloid 2. Then you had things like Voiceroid, UTAU, CeVIO, Alter Ego and finally Synth V, which was the first of the bunch to use AI, then we got the AI centric ones like EmVoice. SynthV gained traction and wide adoption in the western market because of the complications involving the purchase and licensing on Vocaloid and its voice banks which were only sold in Japan. One could argue the idea of a virtual voice synthesis engine specifically tailored for the English language is still new.
I think this is a good analysis, of the situation.
Emvoice is great as a song writing tool, which is what it is advertised as and while it can replace (a) real vocalist(s), the sounds that are produced are sometimes difficult to work with so I am hoping for even more evolution, as time goes on.
I am aware that it all started independently of Western technology and I am very thankful there is anything at all because I cannot imagine being able to work with that many (5) voices at once, just to get a harmonious blend of vocal characters, without people being paid, and why waste peoples’ time, in any case when an AI can do what singers would normally do, and that is add style, if an even greater reward can be reaped down the track, once melodies, rhythm and chord progressions, are well established.
Anyone know how I can distinguish WIND (as in the blowing stuff) from WIND (as a clock)? Omnivocal wants to do the clock wind version but I want the other. The manual is not clear on this (or I’m a dummy!).
Hi Spyder
Type wind in the text box. You will get [w ay n d] phonetics. Change the ay only to i and you will get wind as in run like the wind.
Got it! Many thanks!