Cantai: singing your Dorico files. Some results

ONE 2006 MacPro w/ multiple 1TB SSDs & HD’s at Snow Leopard / Metric Halo FIREWIRE

ONE 2009 MacPro w/ multiple 1TB SSDs & HD’s at … MULTIPLE BOOT Leopard/Snow Leopard/Mountain Lion/Mavericks … it CAN run Yosemite but I won’t because of the dreaded iCloud Sign in LOOP! / Metric Halo FIREWIRE

THREE 2007 MacBook Pro computers w/ 1TB SSDs at Snow Leopard / Metric Halo FIREWIRE

ONE 2009 MacMini at Yosemite to run ONE APP that ONLY runs on that Mac and OS!!!

FOUR 2012 Mac Minis w/ 2TB SSDs (bought before SSDs prices got insane) DUAL Boot into Sierra & Mojave

2013 MacPro … 2TB SSDs, 128GB RAM … DUAL Boot into Mojave & Monterey

TWO 2014 Mac Minis … one w/ 250GB SSD at Monterey, other w/ 1TB SSDs … DUAL Boot into Mojave & Monterey

Many of the above I accepted in lieu of getting paid in my music playing/teaching/arranging work AND when I used to REPAIR Macs (when Macs could be repaired)… Yup, I KNOW! I’ve chosen some GREAT lines of work, haven’t I? Any body wanna buy some Beanie Babies! They’ll be worth a FORTUNE one day, you just watch! LOL

I think I have a box somewhere in the basement with about 20 Beanie Babies that I bought for investment purposes. Alas, they’re probably only worth letting my cats have as toys.

The cats will think they’re priceless… for about five minutes!:grinning_cat_with_smiling_eyes:

Where do I find a keyboard that can do this? I realize with special software Steven Hawking could put together very complex sentences in real time nudging a single lever, but I ain’t no Steven Hawking.

These days software exists that can listen to us speak into a mic and type it all out for us. Would that be the way would work? Say in exactly what you want, and some software produces output in this IPA?

There are a number of online keyboards. This is one (ipa.typeit).

Another Lexilogos nice and visual

Cantai Ensemble Palette:

https://youtu.be/If_5Gm3KQ3k?si=b20UOW3TfYKI2Nkh

So, which version has the “chamber choir.” I find no such thing in v.3.0.2 for Dorico/Sibelius.

I do not think there is a category called Chamber Choir. Look at the stave names (in that video) to see the voices chosen.

Well, the way it’s presented, is odd. It mixes Ottetto and Choral Voices. Choral voices are much less refined in v.3.0.2 in my experience with almost every aspect so Richard has some capability that we do not have.

Hello all! September 14, 2026 today … I’m really enjoying v3.1 for Silicon Macs … (M1 MacBook Air 1TB/16GB … M3 MacBook Air 2TB/24GB … M4 MacMini 1TB/24TB … M1 MacStudio 2TB/32GB ) …

M1 MBA is struggling … M3 & M4 are better … kind of disconcerting to know that the 5 year mark is what we’re gonna have to expect from Macs now???

Any in case, v3.1 is great; much faster and no crashes!

I am sssssssstill having the melisma issue;

…my wanting the word “Hosanna” to be sung as Ho -(Whole Note)-san-(Whole Note)-na(Whole Note), Cantai will sing Ho-(EighthNote & quarter & 3 beats of sssssssssss).

Also, slurred-notes aren’t working… for instance: one word on 2 notes (quarter tied to a half) on the word “Rex” and Cantai will sing “Rex sssssssssssss” out of time (the “exssss” happens after an eighth note) …

(QuickTime letting me output the AUDIO of my Mac (the mic is the only option for sound … SoundFlower doesn’t work on Silicon Macs so I have no Audio in the video to export)…

EDIT: I was able to do a QUICKTIME while connecting my Roland GO… here’s a video … ok … can’t upload a video … here’s a link …

Cantai most of the time can NOT do these simple things. Kind of like many singers, so I guess it’s actually realistic … sadly, not joking, but not what’s needed.

Also, trying to getting words like “Gloria” in English, Spanish AND Latin to be sung as a diphthong (Glo-ria) rather than “Glo - ree - yah”.

The word “to” in English being sung as “tuh” … SOLUTION (Duplicate the score and label it a “CANTAI ONLY” and RE-LYRIC entire portions as “Glory two God in the High yest” … “Hoe saw nah” … “Ah men” … Spanish elisions work better, and sometimes the elisions work in Latin, but the melisma and tied/slurred note issues are still there.

it might be interesting for those of us with a PC and have not installed QuickTime to also be able to see this – or just export the audio. Regarding melismas, they’re by no means as bad as they used to be from what I’ve heard so far, though there is certainly still room for improvement. In general 3.1 seems to me the most stable and best version to date.

Cantai still has a lot of pronunciation issues. It has a problem with “x” in all Latin words like “Rex”, “lux”, etc. The issues that Cantai has seems to increase the longer a project is open that uses Cantai. I frequently start working on a project in the early morning, leave it active, do some outdoor tasks or run some errands, then return to it in the late afternoon or early evening. By the time I return, notes are dropping out and sometimes even entire voices and pronunciation is screwy, etc… I’ve reported this multiple times with nothing more than a “thanks” in return. Apparently, they do not yet have a complete grasp on their product and how to fix what isn’t working. So, as I have said before, Cantai remains an toy at the moment and hopefully it will become a professional tool in the near future.

Same here @kasky1… We’re sending Cantai/Dorico exports to our singers to get a preview so they can prep recordings, and they’re getting a great laugh at a lot of the things I’ve needed to do in the score to make em sing right.

Make NO MISTAKE though … Cantai is THE BEST out there in hearing your scores sing. I can only imagine the work that it takes to get everything done, and the good folks at Cantai are doing a stellar job!

I myself am working on recording samples of my Suzuki Hand Chimes (which we use at a couple of the parishes and schools I work with) as there are none to purchase, not VSL, not EastWest, … unless anyone knows of where I can purchase the (digital) samples? In any case, getting all notes recorded are their own challenge, and I only have the time and patience to get 1) long notes and 2) short notes each recorded with a Stereo Ribbon (Royers) & a Sphere DLX <<< THIS alone is a TON of work, so I can only imagine getting pronunciations, dynamics, words, … grrrrr!

We’re lucky and blessed at how Cantai (and Dorico) has (have) made our way to present our music and arrangements to people such an amazing experience (and get more work!) … having that extra task on top of what we have to normally do has been an adjustment, but, the pay off is worth it :folded_hands:t4:

This is why I always have an engraved score and a “playback score”.

Through Score for MIDI/DAW export … Condensed Score for Publishing … Latin version … English version … I hear ya!

I do that, too. I duplicate the flow and make the funny lyric adjustments in the duplicate for playback to get a more correct rendition.

v3.1.1 just posted:

  1. Smoother phrasing: fixed the clipped words and unexpected breaks in the middle of phrases that appeared in 3.0.2.
  2. More pronunciation fixes: “temp’rate”, “temperate” and archaic verb forms such as “grow’st”, “mak’st”, “dost” and “doth” now sing correctly.
  3. Fixed a hang in Dorico when closing a project (“waiting for VSTAudioEngine”).
  4. Fixed stuck synth notes playing over rendered audio in scores with several Cantai instances.

It still has problems with hard consonants at the end of melismas of almost any length. It is improving but not quite there just yet…

I tested the new version on the choir song, I posted here some time ago. I think this is the first version, that the altos sing the word disciples (the second i as ‘ai’) correctly. The overall impression is a continuous improvement over the several versions that passed by. There is a greater sense of calm and the messa di voce gives it a warmth that I like very much.