Hi all,
I’m looking for an easy to use playback for classical choir, which can sing lyrics.
I tried the trial of cantai on their website, but am not really convinced. What experiences do you have? Does the plugin work better than the trial?
Can you recommend other plugins, which I can integrate into Dorico (or Sibelius).
Thanks and all best!
I work extensively with Synthesizer V. For “legit” choir I’ve found using Dreamtonics’ Choir Voices - Collection 3 to be far superior to anything else available.
• use a slider to size a vocal section from 1 to 16 singers
• draw curves to style vocal tension, volume, etc.
• maintain a system-wide user dictionary of phoneme pronunciations
• export per-channel HQ audio bounces
I can’t rave enough about it. For the additional amount of editing required, the final product is amazing. And it works beautifully with Dorico. Fully usable manual and tutorial videos out there.
Here is an example: TTBB arrangement of Sinéad O’Connor’s This Is To Mother You
there’s nothing just now which is really pain-free. But, as @David_Maddux has just written, Synthesizer V is much the best option at the moment and i have used it to convert about 16 vocal and choral works – probably around 5 hours of music in all. Unlike with Cantai, it is not directly integrated into Dorico but requires a manual MusicXML import which means it’s easiest to have the final form of the project ready when getting to this stage. But it works very well in general with all the features described above.
For more classically orientated work, choir 2 is the most obvious choice, however I like some of the soloists in choir 3, particularly for fairytales so to me it’s worth getting the bundle unless your music goes in a very specific direction (choir 1 is more gospel/pop orientated)
incidentally the website trial of Cantai uses an older version than the current Dorico release. You can try the actual Dorico version by registering and then going into your profile where you’ll find a download option with a key for a 30 second limited rendering trial but even now, there are still a lot of issues – which isn’t to say Cantai’s not worth supporting because the plans are indeed ambitious.
There really is no singular best plug ‘n’ play choir VST. All of them require additional processes to create functional sound. Personally, I am vetted with Cantai because it is a game changer in the VST world and the developer is quite responsive to issues. With that being said, it has not yet matured enough to be usable except in the most basic of ways.
Others here have talked about another product that’s much more legitimate at the moment for high quality work. Luckily for you, entering the search for such things right now is very timely as we are about to experience some phenomenal approaches to this very need to have a VST or pluging that lets our vocal works sing the words, and sing them well.
Interestingly, nearly every alternative has gone wild with advertising since Cantai has launched their Dorico product. They are dropping their prices hoping that we will use their product and plastering that message all over social media. It’s a fun time to be watching this metamorphosis of non-human (aka AI developed) singing. If you’re willing, explore the various trial versions of different developers and watch the transformation occur over the next few months…
Hi all,
I just got Cantai, if you choose the monthly plan, you can cancel anytime. So it is not much $$ do try around with it.
Thanks for all input.
Yes, I’m also excited about better/easier playback possibilities coming up.
Cheers, Stefan
FYI:
I just discovered that Dreamtronics is offering all three choirs as a bundle for less than the cost of any two of the three.
So, since the slider can be set to 1, could I use this (with a single vocal line) to create a vocal solo?
Also, do you know if it works with Mac Intel chip? (Which Cantai does not.)
Thanks.
Not only can you use a Choral Collection database [i.e., vocal library] dialed down to a single voice, you can divide a section into chunks, assign a chunk to a single voice (as a pop-out solo) and return to the section again.
To answer your question:
Yes — Synthesizer V Studio 2 Pro runs on Intel Macs. Per Dreamtonics’ official specs:
macOS: 10.13 (High Sierra) or newer
CPU: Intel Core i5 (1st gen or later), or Apple Silicon M1+
RAM: 4 GB minimum
Disk: 1 GB
Formats: VST3, AU, AAX, standalone
One Logic-specific note: the AU ARA version is actually only compatible with Logic Pro running on Intel Macs (or under Rosetta on Apple Silicon). So Intel users have the edge there. The standard AU and VST3 plugins work across both architectures.
Realistically, AI voice rendering is CPU-intensive, so an older Intel Mac will be slower than an M-series machine — but it’s fully supported, not a workaround. If you’re on a reasonably recent Intel Mac with 8+ GB of RAM, you’ll be fine for most workflows.
At release when I bought it, it was exactly the same as the cost of two separately. But at release there was a sale on which has now ended – just the bundle costs much the same as before.
It’s a little pricey to get all one might need…I think I will wait for Richard and his team to improve Cantai. I guess I’m being a bit miserly at this point…
Certainly Cantai is more economical. I imagine the price will increase in due course when it works more reliably and when some of the planned new voices are out.
I have used several choirs from East West picked up in a sale. All work well for movie type choirs and the sound is not bad at all. I have especially used the phonetic language option which is rather flexible to get pronunciation of special languages like elfish or sanskrit as you want it. The Opus interface is a bit complicated and a a pain to setup in Vienna Ensemble but it does work.
I now also have Cantai and am still experimenting with it. When it works (to use VEP is not possible) it does surprisingly well with respect to pronunciation of the above mentioned languages they never sampled Replacing EW in an existing full score did not work till now. After a long rendering time you hear the acoustics of a hall with people moving but the singing never starts.
I find the EastWest solution to produced rather garbled pronunciation and tended to add Synth V voices to the EW part to make the pronunciation more present and distinct.
In fast sections with multiple syllabes per bar EW indeed gets lost also due to latency. Cantai is much better in those sections.
Apparently Cantai voices have to be the first in the VST rack to make them work which is of course an issue in a existing piece where they would be last after all VEP instances. already there. VST rack entries can unfortunately still not be shifted in Dorico.
That never really occurred to me. Do you have any examples? I used EastWest choirs for well over a decade and still like them for certain things but of course you can’t hear many of the words (although I occasionally hear renderings from folk who obviously gone to a lot of trouble to make the output less muddy than normal)
No examples that I can lay my hands on without a considerable search, but your point about not being able to hear the words is the point I was trying to make. As a (distant?) background chorus, EW, as I have experienced it, would be find, but as a musical theater chorus, the EW WordBuilder is not enough on its own.
For Duel of Fates from Star War EW did a good job. Most two syllable, I think sanskrit words of the piece are clear after a bit of respelling (mainly shifting consonants from first to second syllables of words and some playing around with duration of the syllables).
Unfortunately Opus did an update which for some reason added a lot of latency at the default settings and now the singing falls behind in fast sections and EW has become unusable for this piece.
Cantai works perfectly although the 6 voice choir is a bit thin. I have succeeded adding it to the existing full piece replacing EW and adding the voices in a second instance of VSL MIR 3D in the same hall as the instruments are placed through VEP. The reason it did not work was simply because the rendering of the whole piece takes 20 minutes or so.
Cantai does funny things when the rendering is not finalized and yiu start playback as John suggested in his live video. Some words are sung correctly and then it goes on singing another song in english in sections without words while it is rendering. Once rendering is finished all is correct however,