A library to make your pure-data patches speak, using a database of wav files.
Speaking allow building a speech program by connecting elementary say blocks. Each say plays one of the user's soundfiles (a "phrase"), then triggers the next block. The soundfiles have to be organized in a precise tree structure, in the same way as the proposed example:
myspeech-wavs/
├── jack
│ ├── goodbye
│ │ ├── 1.wav
│ │ ├── 1.wav
│ │ └── ...
│ ├── hello
│ │ ├── 0.wav
│ │ └── ...
│ └── thanks
│ ├── 0.wav
│ └── ...
├── joe
│ ├── goodbye
│ │ ├── 0.wav
│ │ └── ...
│ ├── hello
│ │ ├── 0.wav
│ │ └── ...
│ └── thanks
│ ├── 0.wav
│ └── ...
├── phrases.txt (must contain the list of the phrases to use, one by semicolon terminated line)
└── voices.txt (must contain the list of the voices to use, one by semicolon terminated line)
Open speaking-help.pd for example of use.
PureData (version 0.48 minimum) + externals :
- iemlib
- hcs
- ggee
Antoine Rousseau antoine@metalu.net 2018
license : GNU GPL (see LICENSE.txt )