Version 1.0.5
This release is primarily a refactoring release - the system we use for compiling native dependencies has been totally rebuilt in preparation for adding a new and improved voice detector in a future release.
Integrations
No integrations are available for this version of DissonanceNew Features
- Added an inspector for the SamplePlaybackComponent which shows statistics on the playback pipeline
Bug Fixes
- Fixed a multithreaded torn variable read on 32 bit systems
- Fixed some serialization errors in the Forge demo scene
- Included LLAPI demo scene in package
- Fixed the `WaveFormat` property of the SpeechSession being wrong value
- Fixed the playback system consuming samples at the wrong rate in some circumstances (causing serious desyncs)
- Fixed the microphone capture pipeline losing samples when the pipeline is overloaded
- Automatically resizing the buffers in the capture pipeline to make it harder to overload
Known Issues
- Lacking platform support for:
- Windows Universal Platform
- VAD does not always work well (particularly on mobile devices)