## Proposal
Full emotion with TTS is very difficult (and outside of the range of this project), but with SSML prosody tags we could create "preset" functions to provide some variation and suggest dif…
## TL;DR
Is it better to use the same speaker's voice data for learning?
## Question
I am a student studying speech synthesis at university in Japan.
I read paper about [Tacotron](https://arx…