Skip to content

gtts.md

Latest commit

 

History

History
45 lines (33 loc) · 1.27 KB

gtts.md

File metadata and controls

45 lines (33 loc) · 1.27 KB
Raw

gTTS Engine

GTTSEngine uses the Google Translate Text-to-Speech package. It is a simple network-backed engine with no API key, but it is not an offline or commercial service contract.

Install

pip install "realtimetts[gtts]"

The source uses gtts, pydub, and a temporary WAV file before queueing PCM. Make sure your environment can decode MP3 through the audio stack used by pydub.

Minimal Use

from RealtimeTTS import TextToAudioStream, GTTSEngine, GTTSVoice


if __name__ == "__main__":
    voice = GTTSVoice(language="en", tld="com", speed=1.0)
    stream = TextToAudioStream(GTTSEngine(voice))
    stream.feed("Hello from gTTS.")
    stream.play()

Source Notes

  • GTTSVoice(language="en", tld="com", chunk_length=100, crossfade_length=10, speed=1.0) controls language, regional TLD, and optional speedup.
  • Passing a string to set_voice() creates GTTSVoice(language=that_string, tld="com").
  • get_voices() builds combinations from gtts.lang.tts_langs() and common TLDs.
  • Output is queued as mono 16-bit PCM at 22050 Hz after MP3-to-WAV conversion.

Troubleshooting

  • Network failures or service changes surface as synthesis errors.
  • If speedup sounds choppy, keep speed=1.0 or tune chunk_length and crossfade_length.