well, for now, NO voice is available.. I just get that “loading voices…” message in the “Voice” drop down. it’s been 15 minutes now. I had time to take the puppy out for her walk.
So I’ll have to say that so far, Cantai is a bit out of my league as a musician who uses a computer (rather than a computer person who makes music.)
Well, it is in beta. If you save, close and reopen your project, the voices should load. Might have something to do with the Vst engine? I don’t know. Closing and reopening a project helped me to get the sounds to load-although in my last example, it wasn’t worth the wait 
Maybe they’ll hire you to write the music for the next BlumHouse ghost project?
Ha! I really do admire what Turing Opera is attempting to do here, so I’m willing to give them time. I can’t imagine how difficult it must be to create a workable vocal Vst that will meet the needs of such a diverse group of composers. At least this implementation is better than the one in Musescore.
well, for now, NO voice is available
I have the same issue. Clicking on the settings (ellipses button in the Cantai UI) shows the Cantai Rendering Engine status as: Not connected. I did not install to the default location, not sure if that has anything to do with it.
I found the BBCSO library was very slow loading its instruments, but this is a bit on the ridiculous side.
I also have the Cantai Rendering Engine status as Not Connected.
I sent the log files to their chat bot as requested in the guide, but it just keeps asking if I want to report an issue, even after saying “yes”. Hopefully it got through. I’m on Windows, btw.
Cantai applied to a lead sheet. In Play Mode VST and MIDI, added Cantai to the list, and in Track Inspector, click Edit Instrument icon to be able to choose the singer you want (I chose Marina). Then make sure the vocal track is using Cantai. I used Halion Sonic Etude piano for the chord track. Well, not bad for a mock up. This could be the start of something amazing! BTW, this song is in public domain.
Oh Papa Blues (1st page) (Cantai).pdf (47.5 KB)
I have an issue of selecting the vocal part and the voice to use, having ot show up in the Cantai Vst, but not rendering at all. Clearing the cahe didn’t help. Seems the rendering engine is a bit ‘hit or miss.’
What I noticed during my short test with a few bars:
Cantai refuses to render as soon as a non-English word is entered into the lyrics. The voice then remains completely silent. Only after deleting the obviously unrecognized (in my case, German) syllables could I hear the voice again.
Or should I have waited longer (several minutes for 16 bars)? I’m also missing a rendering progress indicator; is there one somewhere?
The Vst pop up window shows ‘Rendering’ in the lower left hand corner. I have an SATB homophonic song that has been rendering for a long time now-and isn not yet playable. Time to give up and go to bed now.
@Michel_Edward If you’re using an Intel-based Mac, that will be why Cantai isn’t working for you: at the present time, Cantai only supports Apple Silicon (M-series) Macs.
Nope, on a Windows PC.
However, according to one comment on the Cantai discord, it might have to do with my not having installed the library in its default location…
If this is the case, then I will not be able to use Cantai either ever, or until they allow users to install to other locations/drives on their computers. (I install all my VSTs on a partition specifically for that use.)
My work is in Spanish and it was also very hit or miss, and mostly missed. Especially when dealing with sixteenths or eighth-note triplets.
Rendering sometimes completed (without making things noticeably better) or not. I’m also unclear which voices are better than others or which are more appropriate for various genres etc. That sort of information would be helpful. Same with a progress bar for rendering. I’m on a M2 MacBook Air so it shouldn’t be an issue in terms of performance.
Having a beta that expires on a specific date for everyone, a date that is two weeks away, seems less preferable than having a beta that could be started by the user at any date that then expires after 2-4 weeks (which is how many other beta programs work). Especially when being told things will be better by the date the beta expires. I’d rather they had waited to make those improvements before releasing this beta or extending the end date into May. It clearly has promise but it does seem to have a ways to go.
It only starts working if you enter the serial number. After I entered the code (and turned Dorico off and back on), it started working perfectly. You just have to wait a few moments to have the various voices available and select them. I found this video very helpful: https://www.youtube.com/shorts/DNezKiUk3is
unlike some others, I had no problems getting this thing to work and indeed it’s a bit of a miracle that Cantai can indeed just read the score as promised and in general (there was the occasional glitch) prepare the rendering in no time. So congratulations on that!
For the most part, that’s where the good news ends as things stand. The lack of current support for anything other than English severely limits things – when faced with a German text it remained silent. At least Latin should be available before official release because my experiences with Latin in the web renderer were not good and there’s no way of knowing if it will work properly. Dynamics support seems non-existent unless I’m missing something- at any rate when I change the dynamics in Dorico – even CC7 general volume-- nothing happens. Cantai chooses it’s own dynamic which can mean entire sections being submerged or less commonly too loud. I know that the idea of this software is to be plug-and play NotePerfomer style and it does this very well but a complete lack of any sort of configurability (such as the aforementioned vibrato) as far as I can determine severely handicaps this if it’s to be used for mock-ups.
As does the generally rather primitive vocal models. One or two like Hidenon, Hanneke or Violeta are just about usable and at times (especially Hanneke) even quite expressive but so many of the choices (and the choirs) are frankly far too crude for the sort of music I thought Cantai was aiming at – ie more classically orientated.
Agreed. My only use for it would be a classical choral sound. Until that materializes, my pre-order is just funding other people’s stuff (and I was happy to support it… just waiting).
the serial number is entered and recognized.
I’ve exited Dorico multiple times (I installed it last night so it actually got a good night’s sleep as well.)
as I said, there is the distinct possibility that this is caused by not having installed Cantai where it wanted to be installed.
actually in terms of a choral sound, probably EastWest get closer. Main problem being it’s hard to make out the words. But at least there are words, unlike the conventional sampled alternatives.
