all that AI, and your TTS engine still doesn't understand how to transition between words realistically (in speech, sound of next syllable changes based on initial mouth shape, afaict?)
High-end TTS will try to actually alter phonemes (or at least mangle stuff together a little more), but basic and midrange TTS still doesn't handle that kinda stuff, yeah. :\
@sydneyfalk even high end doesnt really do the between-words stuff
I've heard what seemed like good emulation of the inbetween, but I suppose I've got a crappy ear for it, too. Maybe it's just my bad aural processing. O_o Sorries. <3
@sydneyfalk also i wish literally all of them just read out abbreviations instead of expanding them so i dont get every 'st' expanded to 'street' and "SOC' as "social"