Voice Cloning Script

This project is a Python script designed for generating text-to-speech (TTS) audio using voice cloning technology. It employs pre-trained models to synthesize speech from text input, allowing for voice cloning from a given reference audio file. The script includes options to output audio in both .wav and .mp3 formats and to extend the audio duration by adding silence at the end.

Table of Contents

• Introduction
• Installation
• Usage
• Features
• Dependencies
• Configuration
• Examples
• License

Introduction

The Voice Cloning Script leverages the TTS Python library and other audio processing tools to generate high-quality synthetic speech. This script allows you to clone a voice using a reference audio file and generate speech from a given text. Additionally, the script supports outputting the audio in .wav and .mp3 formats and provides an option to extend the audio length by adding silence.

Installation

To use this script, ensure you have the required Python packages installed. You can install them using pip:

pip install TTS soundfile numpy

Note: The script references a local .wav file for voice cloning. Ensure this file exists at the specified path or modify the path accordingly.

Usage

To generate cloned TTS audio, run the script with the desired text input. You can customize the output file paths if needed. To make the program faster, modify gpu to True.(Disclaimer: you need to have a gpu with cuda 11.0+ support and cuda installed on the device).

python voice_cloning.py

Example Command

python voice_cloning.py "Your text here" "audi reference path here" 

This command will generate a TTS audio file with the cloned voice based on the reference audio specified in the script.

Features

• Voice Cloning: Clone a voice using a reference .wav file.
• Text-to-Speech Generation: Convert text input to synthetic speech.
• Audio Output: Save the generated audio in .wav format.
• Audio Prolongation: Add silence at the end of the audio to extend its duration.

Dependencies

The following Python libraries are required to run the script:

• TTS: A Python library for text-to-speech synthesis.
• soundfile: For reading and writing sound files.
• numpy: For numerical operations, including handling silence in audio.

Configuration

• Text Input: Modify the text variable in the script to change the input text for TTS generation.
• Reference Audio: Update the path to the reference .wav file for voice cloning in the generate_cloned_tts function.

Examples

Below is an example of generating speech with a cloned voice:

1. Set the text variable to your desired input:

   text = "It took me a long time to develop a voice, and now that I have it I'm not going to be silent."

2. Run the script:

   python voice_cloning.py