Fluency voices fail after first playback

Last reviewed: 3/26/2005

PRB Article ID: P030501

The information in this article applies to:

  • SpeechKit 4


Only the first StartPlayback with a Fluency SAPI 4 voice completes, then consumes CPU.


Like the Voxit Macaw voice, the Fluency SAPI 4 IAudio interface LevelGet and LevelSet functions fail causing an incomplete ChantTTS initialization.


Since Fluency and Voxit voices are based on RealSpeak technology, the original workaround for Voxit Macaw added in SpeechKit 4 build was removed and replaced with more extensive SAPI 4 fidelity checks to address all RealSpeak-based voices. Fixed in SpeechKit 4 build

Component Formats Impacted

All component formats.