System Engine

SystemEngine wraps the operating system TTS voices through pyttsx3. It is the best first smoke test because it does not need cloud credentials or model downloads.

Install

pip install "realtimetts[system]"

This installs pyttsx3 plus the RealtimeTTS playback dependencies. Available voices depend entirely on the host OS.

Minimal Use

from RealtimeTTS import TextToAudioStream, SystemEngine


if __name__ == "__main__":
    stream = TextToAudioStream(SystemEngine(voice="Zira"))
    stream.feed("Hello from the local system engine.")
    stream.play()

Source Notes

• SystemVoice(name, id) stores the OS voice name and ID.
• SystemEngine(voice="Zira", print_installed_voices=False) selects the first installed voice whose name contains the requested string.
• set_voice_parameters(**kwargs) forwards values directly to pyttsx3 setProperty, so rate, volume, and other support are platform-dependent.
• Output is queued as mono 16-bit PCM at 22050 Hz.

Troubleshooting

• If voice selection appears to do nothing, call SystemEngine(print_installed_voices=True) and use a substring from an installed voice name.
• On Linux, install PortAudio headers before installing PyAudio. Some systems also need an OS speech backend configured for pyttsx3.