Here’s something nobody thinks about.
Notation software was never designed to sing. It was designed to trigger samples. You write a note, you hit play, it fires a recording of somebody playing that note. That’s the model. It’s been the model for thirty years.
So what happens when you want the software to actually sing your lyrics? Shape vowels, land consonants, breathe in the right places, carry a phrase the way a real singer would?
It turns out you can’t just bolt that on. The scoring application itself has to meet you halfway.
The Wall
When I started building Cantai, I assumed the hard part would be the synthesizer. The voice modeling, the phoneme engine, the expression system. And that was hard. But I hit a wall I didn’t expect: the notation software had no way to pass lyrics to an instrument. Think about that — you’ve got lyrics sitting right there in the score, and the playback engine has no idea they exist. It never needed to. An “ooh” and “aah” sample doesn’t care what the words are.
So I started reaching out.
Three Companies, One Problem
Over the past year and a half, I’ve been working directly with the engineers at three major notation platforms — MuseScore Studio, Dorico, and Sibelius. The people who actually build these applications. Every single one of them had to create something new to make Cantai work.
I can’t go into the technical details — that’s their work and their IP. But I can tell you the scope of it. These were architectural decisions that required real engineering time from teams that have plenty of other things to build. And they did it. All three of them.
Three separate companies, each with their own codebase, their own priorities, their own roadmap — all independently decided that singing synthesis was worth investing engineering resources in. They saw what Cantai could do and decided their users should have it.
Dorico’s Turn
The MuseScore Studio team went deep — they built the most extensive integration of the three, and Cantai has been live there since December. Now it’s Dorico’s turn. Daniel Spreadbury and his team at Steinberg have been actively involved in making this integration as native and as thorough as possible. If you know Daniel’s work, you know that’s exactly what you’d expect from him.
Three major software companies restructured their architectures to support Cantai — and every vocal plugin that follows will walk through that same door. This is a permanent change in what notation software can do.
Free Beta — April 15
The Dorico beta goes live on April 15 — the first time Dorico users can write lyrics into a score and hear them sung back. The beta is free.
If you want to lock in a lifetime license, they’re half price — $150 instead of $299 — through the April 30 launch. One license covers MuseScore Studio, Dorico, and Sibelius when each goes live.
Pre-order here: Canti VST for Dorico & Sibelius - Pre-Order Now | Cantai
Happy to chat about any of this in the thread!
Best,
Richard deCosta
Composer, Founder
Cantai Intelligent Singing Synthesizer / Turing Opera Workshop