Have you ever wanted to listen to non-voiced quests in Genshin Impact or Honkai: Star Rail? Well, now it's possible with the use of AI.
On lower-end PCs, your game can take a SIGNIFICANT performance hit.
Also the use of AI voices is completely optional. So if you don't want to use them, you can skip some parts of the installation process.
- Git
- Python 3.10 or Anaconda
- CUDA (If you want to use GPU)
- FFmpeg (Can be skipped if you don't want to use AI)
- RVC-TTS-WEBUI (Can be skipped if you don't want to use AI)
- Install Python 3.10 or Anaconda (Anaconda is useful if you want to use different Python versions and/or you don't want to mess with your main Python installation.)
- Install CUDA version 11.8
- Install FFmpeg and make sure that it's added to your PATH environment variable.
- Install Git
- Set up and install RVC-TTS-WEBUI. If you are using Anaconda environments, you can skip making a venv while following the install guide for RVC-TTS-WEBUI.
- Clone this repo.
git clone https://github.com/nitolar/No-Voice-Overed-Quests-TTS
- If you created a venv or Anaconda environment for RVC-TTS-WEBUI, you can reuse it or create a new one.
- Install Python requirements.
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
pip install -r requirements.txt
git pull
pip install -r requirements.txt --upgrade
- Configure
settings.yamlto your liking. - Add some voices. An example of how to add one is in
voices/example.yaml. (This step can be skipped if you don't want to use AI voices.) - Start RVC-TTS-WEBUI. (Only if you plan on using AI voices.)
python app.py
If you want to force the RVC-TTS-WEBUI to use your CPU instead of your GPU, you can edit the config.py file like this:
# AROUND LINE 60
# USE GPU IF AVAILABLE (ORIGINAL CODE):
if torch.cuda.is_available():
# FORCE USAGE OF CPU:
if torch.cuda.is_available() == False:
- Start the script.
python tts.py
- You are ready to go!
- Improved the process of cleaning up text passed to be read by pyttsx3 or AI TTS.
- Improved finding model file
- Added some new commonly misinterpreted characters.
- Increased the possible area for a name to appear in genshin
- Improved the process of cleaning up text passed to be read by pyttsx3 or AI TTS.
- Added more debugging text, making it easier to detect errors when creating the
Text Final. - Added colors to improve the readability of the output.
- Added some new commonly misinterpreted characters.
- Fixed the alt-tab feature not alt-tabbing to the game if no text was found.
- Refined the text extraction accuracy; it should now be more precise.
- Added a new auto alt-tab feature; more info is available in
settings.yaml. - Added some new commonly misinterpreted characters.
- Fixed a lot of spelling mistakes 😑 and rephrased a lot of the text.
- Fixed the script crashing if no text was found.
- Fixed the script not replacing commonly misinterpreted characters.
- Added some new commonly misinterpreted characters.
- Initial release.
Like what you see? Give a star it motivates.
Found any bugs? Report them here: https://github.com/nitolar/No-Voice-Overed-Quests-TTS/issues
- MrQuariti For testing
- litagin02 For creating RVC-TTS-WEBUI