@DavidGoldfield@allovertheplace.ca
2026-07-16 18:15 UTC
Replies (2)
-
@prism@infosec.exchange 2026-07-16 18:25
@DavidGoldfield@allovertheplace.ca @NightDrake@dragonscave.space @EJGilbert@dragonscave.space @bmoore123@tweesecake.social @mcourcel@allovertheplace.ca @mewproj@fwoof.space The thing I can't abide with "new" voices (E.G sample-based TTS) is how it handles punctuation. Periods, commas, semicolons, dashes and exclamations all have the same inflection. espeak/eloq/dectalk varry the pitch or take just a little bit longer or shorter for each which helps me a lot to determine the intent of something. It also puts different stresses sometimes depending on the capitalization. I, for instance, sounds different from i, for espeak. But not with, say, vocalizer. This is also why I avoid piper voices.
-
@NightDrake@dragonscave.space 2026-07-16 19:00
@DavidGoldfield@allovertheplace.ca I have noticed that is a thing with consonants and Eloquence. But you are also right about the fact that it does do really well with super high speeds. I do use it sometimes. But if I'm going to read for pleasure or something that's long I might go with one of the higher quality voices. I like Tom as well. @EJGilbert@dragonscave.space @bmoore123@tweesecake.social @mcourcel@allovertheplace.ca @mewproj@fwoof.space