Ling 441 - Advanced Phonetics - Speech Synthesis, part 1

Speech Synthesis, Phonetics

[afonsocarvalho]

Finally an useful video. Thank you so much

[tomerfeiner]

The main issue with the early attempts to make fem voices with KlattTalk was that the voice was quite heavy. They had to figure out a way to modify the waveform to reduce the vocal weight. The resonance was quite good even on the first try, _maybe_ they just had to modify it a bit. It was mostly a vocal weight thing

[jhoughjr]

brains are amazing. I could hear it when I knew what it was saying, but totally not beforehand.
It appeared to me to not be a good model of a real "wetware" vocal tract.
That might affect its intelligibility

[jhoughjr]

Ah DECtalk heard those a lot growing up.

[jhoughjr]

what did you say before that? tea or coffee. what did you do with it then? is what the formant synth said I think.

[pashtotechmaster]

Hi bro
I want to make a Text to Speech so if you guide me where to start and what to do.