any sentence, sung by a rule from about 1026.
reading:
Your browser blocked sound for this page. Check that the tab isn't muted, then press sing again.
Around 1026, a monk named Guido wrote down a way to turn any sentence into a tune: give each vowel a note, and sing what comes out. Type something and hear it.
Every sound is synthesised in code. Nothing is AI.
what the rule decides
Which note each syllable gets. In strict mode that's all of them, so nobody chooses anything. With
wander on, the rule offers a menu and a dice roll picks from it, one roll per note.
what i decided
Guido's rule gives pitches and nothing else. The speed, the voice, the room and the equal note lengths are mine. The tuning comes from the monochord he describes in chapter 3. Setting the last note to G is my reading of what he did at the end of his hymn.
where mine differs
His rule was built for Latin, and English spelling doesn't behave. So in English it takes the first written vowel of each spoken syllable, which means "through" sings o. Words it splits badly can be fixed with a hyphen. It sings vowels only, because that's all the rule reads.