EdgeTTS.DotNet

EdgeTTS.DotNet is a C# (.NET) library that allows you to use Microsoft Edge's online text-to-speech service. It is a feature-complete migration of the popular Python edge-tts library, designed for performance, cross-platform compatibility, and ease of use.

Features

• High-Quality Speech: Access Microsoft Edge's neural TTS voices for natural-sounding speech.
• Multilingual Support: Supports over 400 voices across numerous languages and regions.
• Subtitles: Generate SRT formatted subtitles from WordBoundary or SentenceBoundary events.
• Customizable Prosody: Adjust speech rate, volume, and pitch to suit your needs.
• Cross-Platform: Built with .NET 10.0, fully compatible with Windows, macOS, Linux, and mobile platforms like .NET MAUI.
• Robustness: Includes built-in clock skew correction (DRM), automatic 403 retry, WebSocket compression, and comprehensive error handling.
• Voice Metadata: Access voice tags including ContentCategories and VoicePersonalities.

Installation

dotnet add package EdgeTTS.DotNet

Quick Start

Basic Usage

Save text to an MP3 file:

using EdgeTTS.DotNet;

var request = new Communicate("Hello, world!", voice: "en-US-EmmaMultilingualNeural");
await request.SaveAsync("hello.mp3");

Generating Subtitles

Use the SubMaker class to create subtitles:

using EdgeTTS.DotNet;
using EdgeTTS.DotNet.Models;

var communicate = new Communicate("Hello world!", rate: "+10%", boundaryType: "WordBoundary");
var subMaker = new SubMaker();

await foreach (var chunk in communicate.StreamAsync())
{
    if (chunk is AudioChunk audio)
    {
        // Handle audio data
    }
    else if (chunk is MetadataChunk metadata)
    {
        subMaker.Feed(metadata);
    }
}

string srt = subMaker.GetSrt();
File.WriteAllText("hello.srt", srt);

Listing Available Voices

using EdgeTTS.DotNet;

var voices = await Voices.ListVoicesAsync();
foreach (var voice in voices)
{
    Console.WriteLine($"{voice.ShortName} ({voice.Gender}) - {voice.Locale}");
    Console.WriteLine($"  Categories: {string.Join(", ", voice.VoiceTag.ContentCategories)}");
    Console.WriteLine($"  Personalities: {string.Join(", ", voice.VoiceTag.VoicePersonalities)}");
}

Advanced Options

Option	Description	Format
voice	The short name of the voice (e.g., en-US-EmmaMultilingualNeural)	String
rate	The speed of the speech	+0%, -50%, etc.
volume	The volume of the speech	+0%, -25%, etc.
pitch	The pitch of the speech	+0Hz, -5Hz, etc.
boundaryType	Type of metadata events to receive	SentenceBoundary (default) or WordBoundary
proxy	Proxy server for network requests	http://proxy:port

Error Handling

The library provides specific exception types for robust error handling:

Exception	Description
NoAudioReceivedException	No audio was received from the service
UnexpectedResponseException	Unexpected response format from the server
UnknownResponseException	Unknown response type received
WebSocketException	WebSocket connection or communication error
SkewAdjustmentException	Clock skew correction failed (missing server date)

Testing

The project includes a comprehensive test suite using xUnit:

dotnet test

License

This project is licensed under the MIT License. See the Python project edge-tts for more details on the original implementation.