ShapeVoice is not a voice AI or speech synthesis system.
It generates synthetic audio using waveform-based methods.
- Text → frequency mapping system
- Waveform-based audio generation
- File-based input system
- WAV audio export
- Experimental speech-like rhythm (not actual speech synthesis)
git clone https://github.com/ThatOneUntitledProgrammer/shapevoice
cd shapevoice
pip install -r requirements.txt
python main.py
- Python 3.x
- numpy
- scipy
You can install dependencies using either method:
pip install numpy scipyor:
pip install -r requirements.txtIt doesn’t matter which way, both install the same dependencies.
- Reads text from input file
- Converts letters into frequency values
- Generates waveforms (triangle / square / noise)
- Builds audio signal
- Exports WAV file
- Experimental audio synthesis
- Text-to-sound engine
- Shape-based waveform generator
- Windows
- Linux
- macOS
As long as Python and dependencies are installed, ShapeVoice will run on all platforms.
Each run overwrites output/result.wav by default.
To prevent overwriting, change the output filename:
export_audio(wave, "output/yourfilename.wav")