Welsh Neural Voices for NVDA - v2025.11.1 Beta

Welsh Neural Voices for NVDA - v2025.11.1 Beta

⚠️ BETA SOFTWARE - This is a beta release for testing. Please report issues on our GitHub issues page.

About This Release

Welsh-only neural text-to-speech for NVDA using Piper models. This addon provides high-quality Welsh (Cymraeg) voices for NVDA screen reader users.

What's New in v2025.11.1

Bug Fixes

• Minor bug fixes and stability improvements

Previous Release (v2025.11.0)

Features

• Welsh Neural Voice: Multi-speaker neural voice (3 speakers, medium quality)
• Automatic Download: First-run setup with user confirmation
• Bilingual: Full Welsh and English documentation
• Accessible: Progress announcements and clear user feedback

Accessibility Improvements

• User confirmation dialog before downloading (77 MB)
• Progress announcements at 25%, 50%, and 75% completion
• Completion dialog before NVDA restart
• No surprise restarts - user has control

Installation

1. Download techiaith_tts-2025.11.1.nvda-addon from the assets below
2. Open with NVDA (or use NVDA menu → Tools → Manage add-ons → Install)
3. Restart NVDA when prompted
4. A dialog will ask if you want to download Welsh voices (77 MB)
5. Click Yes to download, then OK when complete to restart

System Requirements

• Operating System: Windows 10/11 (x86 or x64 only)
  • ⚠️ ARM64 Windows NOT supported
• NVDA Version: 2025.1 or later
• Internet Connection: Required for initial voice download (~77 MB)

Known Limitations (Beta)

• Only one Welsh voice included (3 speaker variants)
• x86/x64 Windows only - ARM64 not supported
• NVDA 2025.1 or later required
• Beta software - limited testing performed

Technical Details

• Version: 2025.11.1 (year.month.revision format)
• Author: Stephen Russell, Uned Technolegau Iaith / Language Technologies Unit, Bangor University
• Based on: Sonata-NVDA by Musharraf Omer
• License: GPL v2
• Voice Engine: Piper TTS
• Runtime: Sonata by Musharraf Omer

Acknowledgments

• Musharraf Omer - for Sonata and Sonata-NVDA
• Piper TTS project and Rhasspy community - for neural voice models

Support

• Issues: https://github.com/techiaith/nvda-addon/issues
• Website: https://techiaith.cymru/cynnyrch/nvda/

SHA256 Checksum

4ed6dc65ae4f41045169b6b7fabf5d80e95d3f8cfc1b0734aac099812b1e6f8b