1
2
3
4
5
6
7
8
Seriously, though, Synthesizer V’s syllable onset and coda strength features are super useful, and so is scalable voicing over time for those syllables you just want to force to be unstressed.
(And ofc when that /z/ at the end just sounds weird.)

I think I tuned this 80% using the note settings and phonemes...

SynthV’s pretty good at automating the pitch transitions, and adjusting timing and scaling on those are just about enough to sound kinda natural for most purposes.

I did small tweaks on voicing, tension, and gender to improve the pronunciation, but I really just liked how well Eleanor’s natural voice went with a song like this so I didn’t try anything too crazy.