"VoiSona Talk" version 1.3.7 is now available.
Please download it from the DOWNLOAD page.

Major Improvements

• Enhanced the REST API (version 0.9.1)
  • Added an endpoint to retrieve WAV data via the REST API.
  • Added support for specifying the duration of each phoneme during speech synthesis.
  • Added an API that returns the name of the current audio device.
• Added controls (shown as filled circles) to the accent editing screen for splitting and joining accent phrases and for adding and deleting pauses.

Main changes

• Changed the default settings to 1 track, 50 blank sentences, and 1 audio track.
• Changed the default voice library for new tracks to the voice library most recently selected on the voice selection UI.

Main issue fixes

• Fixed an issue where the REST API returned 405 instead of the correct 404 for non-GET requests to non-existent endpoints.
• Fixed an issue where track numbers were sometimes not assigned sequentially when adding tracks.
• Fixed an issue in the accent editor where the duration of a mora could not be set shorter than the duration of its first phoneme.
• Fixed an issue where range selection in the sentence list did not work correctly after selecting a sentence in the track list.