brailcom/speechd

What it solves

Speech Dispatcher provides a common, device-independent interface for accessing speech synthesis (Text-to-Speech). It allows blind and visually impaired users to interact with computers and the internet using free software by abstracting the underlying speech engine.

How it works

It acts as a high-level layer between client applications and various speech synthesis backends. The system consists of a central server that communicates with speech modules via a simple protocol over pipes. This architecture allows it to integrate a wide variety of syntheses, including C libraries, external commands, and HTTP services, regardless of their licensing.

Who it’s for

Developers creating accessibility software and blind or visually impaired users who need a stable, well-documented interface to manage multiple speech synthesis engines.

Highlights

• Broad Engine Support: Supports a vast array of TTS engines including Piper, Mary, eSpeak-NG, and Festival.
• Flexible Architecture: Uses a pipe-based protocol between the server and synthesis modules, making it easy to develop new backends.
• Multi-Language Bindings: Provides API libraries for C, Python, Common Lisp, and Guile, with Rust and Java bindings under development.
• Accessibility Focused: Part of the Free(b)soft project aimed at empowering visually impaired users.