This script uses VOICEVOX to generate tex-to-speach voices for renpy.
Important
It is recommended to use this generator only for placeholders. The audios generated by VOICEVOX are not of good quality.
Tip
The use of human voices is recommended, the voices generated with VOICEVOX are not perfect and human talent must always be supported.
python -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt
once the packages are installed, go to renpy and generate dialogue.tab in "Extract dialogue".
Important
it is recommended to check "Remove text tags from dialog".
Move the dialogue.tab file to the root folder of the script (next to main.py) and run main.py with python.
It will start downloading VOICEVOX and you will have to run main.py again when it finishes downloading and you will have to wait for VOICEVOX to generate the voices, the voices are generated in /audio/.
in the speakers.json file you will find the list of voices you can use, the important thing is styles[0].id since that is the voice to include in the config.py file.
{
"supported_features": {
"permitted_synthesis_morphing": "SELF_ONLY"
},
"name": "もち子さん",
"speaker_uuid": "9f3ee141-26ad-437e-97bd-d22298d02ad2",
"styles": [
{
"name": "ノーマル",
"id": 20
}
],
"version": "0.14.3"
}Tip
You can listen to an example of the voices in: https://voicevox.hiroshiba.jp/#characters
character="h" # renpy character define
def_lang="ES" # Default renpy dialogues language
speaker=20 # speaker id
Important
This script can only help you generate voices for a character you specify.