Skip to content

Windows Integration and UI Refinement

Choose a tag to compare

View all tags
@senigami senigami released this 03 Apr 04:28
· 9 commits to main since this release
v1.8.4
42d7bc4
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Audiobook Studio 1.8.4: Windows Support & Predictive Progress

This release marks our first fully verified milestone for Windows installations, alongside a major stabilization pass for the version 1.8.4 lifecycle. We have transitioned the "Engine-per-voice" workflows into a hardened, production-ready state with a focus on cross-platform reliability and UI fluidity.

🪟 Official Windows Support

  • Native Installer Logic: Updated the Windows installation scripts (run.ps1) to be more robust, including better dependency checking for ffmpeg and python paths.
  • Compatibility Fixes: Resolved issues with FFmpeg DLL handling on Windows, ensuring that XTTS voice conditioning and sample previews work out-of-the-box without manual configuration.

✨ New in 1.8.4

  • Predictive Progress Tracking: Progress bars now use interpolation logic to animate smoothly between websocket updates. This eliminates visual "stutter" and provides a real-time representation of the rendering process.
  • Persistent Voice Profiles: You can now set and save default voices at both the Project and Chapter levels. These selections persist across sessions, ensuring regenerations respect your chosen speaker.
  • Descriptive Queue Labels: Replaced generic "System Task" messages with clear, engine-specific titles (e.g., "Voxtral Synthesis" or "XTTS Preview Build") for better background visibility.
  • Bifurcated CI/CD: Optimized the development pipeline with a "Quick" validation path for Pull Requests, reducing feedback wait times by 60%.

🛠️ Fixed in 1.8.4

  • Voxtral Regeneration: Fixed a bug where "Regenerate" actions failed to clear old buffers; it now triggers a true audio rebuild.
  • Duration Metadata: Hardened detection logic to prevent audio metadata from collapsing to 0.0 during complex probing tasks.
  • Path Sandboxing: Implemented isolated database path resolution for safer automated testing.
  • Linting: Resolved pre-push issues where deleted files would trigger ESLint errors.

💡 Why this matters

With full Windows support and predictive UI, Audiobook Studio is now more accessible and professional than ever. By removing the friction of installation and smoothing out the synthesis feedback loop, this release lets you focus on what matters: your content.

🧪 Technical Note: Massive Test Coverage

This release includes over 5,000 lines of new testing code. We’ve added comprehensive integration tests for engine handlers, API endpoints, and queue recovery to ensure future stability as we continue to scale.

Assets 2
Loading