Vocal synthesis -- automated workflows from .xml or text/midi

[claz]

Hello hello,

Does anyone have recommendations for vocal synthesis software that accepts a music xml file with notation and lyrics, and outputs the sung audio?

In case you don't know what I'm talking about, here's a website that works on that principle: https://www.sinsy.jp/

Realism isn't important, but I need an automated workflow for batch-rendering audio from notes/text files.

There's plenty of work-in-progress software on GitHub that does what I'm asking, but I haven't yet found anything more polished. Some of these accept .xml; some accept a combo of .mid/.txt (separate notes/text), which is more cumbersome, but still fine.

Ideally, I'd be able to access the software through a command line interface -- i.e., via code, rather than through a graphical interface.

I've used AlterEgo (VST link: https://www.plogue.com/products/alter-ego.html), with good results, but it's a bit painful to enter text. It's not designed for automation, but you can kind of hack it by creating presets with the text you want -- but even then, you need a way to render without going into the GUI.

I'm curious about Vocaloid, but there's no free trial or online manual -- an it's only useful to me if I can supply a whole music/text file for an automated workflow.

Anyway, just wondering if anyone else has already solved this problem!

Cheers,

[Tomás Mulcahy]

Vocaloid.

[claz]

Vocaloid accepts .xml? It does not seem so from their docs.