voicesynth

Voice Synthesizing Library

Note

The project uses Silero TTS

How to use:

First, install the package

pip install voicesynth

Then import it into your code and go ahead and use it:

from voicesynth import Model, Synthesizer, show_available_models

show_available_models() # Showcase of all tts models available

# initializing the model
# setting show_download to False does not show model downloading progress
model = Model("v3_en", model_path="model.pt", show_download=False)  
model.set_speaker("en_73")

# creating a synthesizer instance
synthesizer = Synthesizer(model)
synthesizer.say("This is a good way to spend my day!")

Instead of using .say() method, you can synthesize an audio and then play it whenever you need:

import time

synthesizer.synthesize("Whats'up!", path="audio.wav")
time.sleep(3)
synthesizer.audio.play_audio("audio.wav")

The library automatically handles cross-platform audio playback natively. You don't need to specify different modules for different operating systems:

model = Model("v4_ru", "model_ru.pt")
synthesizer = Synthesizer(model)

# Automatically plays using the most reliable native method 
# (e.g., playsound on Windows, paplay/PipeWire/ALSA on Linux, afplay on macOS)
synthesizer.say("Как дела?") 
synthesizer.say("Хорошо, а твои как?")

All models support simple ssml tags:

synthesizer.say("В н+едрах т+ундры в+ыдры п+ели п+есни", prosody_rate=90)   
# I added prosody as a parameter, so that people who are not familiar with ssml tags
# could change speaking speed without knowing how to manually do it

By default, logging is enabled. If it bothers you, you can disable it

from voicesynth import disable_logging

disable_logging()