After spending many tiring hours chatting with AIs, I finally got my Cantai free trial to work on a SATB choir score. I’m not sure what the problem was, but it might have been a lower sound resolution in the configuration of both Dorico and my system (I increased both to 48000) or a conflict with VLC. Or both. Now, there are two problems with the rendering: (1) despite having picked the ‘otetto’ instruments for each staff, I got solo voices; and (2) the voices sounded very airy, as if rendered with low resolution.
Has anyone come across the issues above? When I tried the MuseScore version of Cantai, sound quality was a lot better and I did get choir voices (though only a few, like a very small chamber choir), but it wouldn’t play tuplets correctly and the remote rendering takes an eternity…
All experiences so far have been disappointing and very time-consuming. I’m assuming there is a learning curve, but the online instructions seemed to me insufficient. Help will be much appreciated. Thank you in advance.
I also am using the new Cantai for Mac Intel. Have not tried any choir playback yet, but I agree that the rendering seems to take a long time, even for solos. But my primary problem isn’t the rendering. It’s the vocal quality of the female voices. The baritone called Rigoleto is the only voice I’ve found that has a rich warmth and no sign of strain on the high notes. I’ve found nothing comparable to him in the females.
I also agree that the online instructions are very inadequate.
I’m sure other users with more expertise will chime in, but one thing that occurs to me in regard to your solo sounds on choir parts is that you may have selected single ‘instruments’ when you created the parts in Doricao. I think I remember hearing that you have to select the ‘group’ instruments when creating choir staves. If you haven’t done so already, you might check on that.
Are you able to share some bars of your song, (delete the rest of it), upload in Dorico format. Create an audio file of it and upload, or audio of each singer as well.
Someone here then can render the same from your Dorico file and post the audio so we can compare.
My guess is that because the rendering is done offline in some cloud, the audio returned is probably (at least sounds to me) extremely compressed - honestly sounds like the lowest bit rate of a compressed mp3. Below is a simple export, dry, no effects at all with Ottetta: Kim.
The fact that it is even possible to hear a Dorico score sing is enough for me to forgive the artifacts and low quality - for now ( but I do need a new Mac )
(not assuming that this was your choice - just illustrating the first consideration); and using one of Dorico’s reverb plugins in Play mode. Cantai’s reverb is not terribly useful in my opinion. Just the tiniest bit of reverb gives a more wholesome quality to the voices in my experience.
The fact that you’re using an Intel Mac shouldn’t really cause the issues you’re describing but I would certainly interact with the Cantai chat function to share your issue and open a ticket with Cantai should that be the case. And, rendering locally on an Intel Mac will probably always be rather slow because Cantai uses the CPU for rendering on these machines rather than the GPU on Apple Silicon (and Windows???).
I think your diagnosis may well be correct, as I didn’t select the otteto voices from the choral option. I assumed that the choral options would be too generic, like basically just female voices and male voices. From what you say, I was probably wrong and will try that. Will also try what L3B suggested above – using Dorico’s Section Players, rather than Single Players.
Now, if selecting choral voices instead of soprano, alto, tenor, bass happens to solve the problem of solo rendering (rather than the desired choir voices), then why is there the otteto option inside each voice? I was assuming that (for example) by selecting soprano → otteto I would get the specific timbre of a soprano section of a choir, which sounds different from the timbre of an alto section of a choir – considering ‘otteto’ means octet, that is, eight sopranos.
You might be overthinking Cantai just a bit. Ottetto “eight” is essentially 2 sopranos, 2 altos, 2 tenors and 2 bases available as individual parts; as male and female voices (4 voices each); and, as all voices (all 8 voices) en masse. I’ve not tried the latter because it seems a little odd to me as what is really going on with that that is not achieved with separate male and female voices.
Individual voices listed under, for example, Soprano->Ottetto:[name of singer] just selects that particular singer from the overall Ottetto group. Oh, sometimes, you will have timbre and sometimes you will not - the product is still working on that.
Cantai, though lightyears ahead of other options, is not perfect and expecting some of the subtleties that you describe may be a bit premature. The VST has a LOT of “growing up” to do but it’s going the right direction with decent momentum. I hope this sheds some light.
So, I’ve avoided the Cantai playback template from the beginning because I don’t use groups vs. singular singers: I simply use the voices independently. (I use my own project templates based initially off of Dorico templates.) This means that I have to setup Cantai independently for each voice that I want it to sing using Play Mode. I’ve found this less annoying that Cantai just choosing whomever it wants for the voice parts. Once it selected a soprano voice for each part of an SATB work - not what I would have expected when I used the playback template.
Now, if you’re not comfortable setting up VSTs manually, then you probably should use the playback template and the Group vs single voice setups. You can still change the voices as desired.
What do you mean by using the voices independently? I don’t know what a Cantai template is. When I created the score on Dorico (in fact, it’s an xml import from Finale), Dorico automatically assigned S, A, T and B (single instruments) to my score and automatically loaded Note Performer (which is my default). Then I went to Play Mode and created four Cantai instances, one for each staff, and selected these Cantai instances for playback, rather than Note Performer…
Hi, thank you for that. Interestingly, when I tried the reverb effects above, the basses got too soft and blurred whereas the tenors got very loud and a lot less reverb even as I applied the same amount of REVerence reverb to both. But I did it using the generic male and female choir instruments. Can you tell me the details of how you achieved the above?
I assigned Ottetto:Tenors to Tenor voice and Ottetto:Basses to Bass voice. The REVerence reverb (the default reverb for Dorico) and loaded REVerence’s NYC Chapel setting. For the RoomworksSE reverb, I added it to each voice using the mixer function and changed the reverb tail time to 2.0 sec.
If you need details on how to use these plugins and the mixer, I suggest that you take the time to read through the online manual. It’s a little too detailed to delineate here. Once you start “playing” with these settings, you can find what rings true to your own ears in terms of how your work sounds. A lot is trial and error but the manual gives the foundation to experiment. The Dorico support page also has a document regarding the plug-ins for the mixer.
)
