[espeak-ng:master] new issue: How does this program convert phonemes to audio? (English) #github
email@example.com Integration <espeak-ng@...>
[espeak-ng:master] New Issue Created by Yam76:
#663 How does this program convert phonemes to audio? (English)
I am trying to make a very, very, very simple TTS into Rustlang (for English).
If I understand correctly, this project uses additive speech synthesis, and thus does not actually have any speech samples. Rather, there is a lot of computing that goes into the background to generate what basically amounts to sine waves mashed into coherent speech.
Then, what files "map" each phoneme onto a certain sine wave combination? More importantly, where can I read more about this "mapping"?