Support for cluster unit selection of audio data as described in the Black and Taylor document "Automatically clustering similar units for unit selection in speech synthesis", published in Proceedings of the Eurospeech 1997, Rhodes, Greece.