Acapela Releases Version 9.4 of Its TTS SDK

Acapela Group yesterday released version 9.4 of its text-to-speech software development kit, with new voices, audio boost, enhanced spelling mode, and linguistic improvements.

Available for Windows, Mac OSX, iOS, Android, Linux Embedded, Win CE, Windows, and Linux Server, Acapela TTS 9.4 includes new French, Norwegian, and Swedish children voices. In addition, Anais, a new French voice with a telephone operator tone, has joined the portfolio, bringing to nine the number of voices for this language. And 'Claudia Smile' adds a cheerful voice to Acapela's German offer.

Other language enhancements include Belgian versions of all French voices; an updated part-of-speech tagger, new breathing model, and enhanced phonetizer and stress restoration systems for Norwegian, French, and Dutch voices; the alignment of phrasing/chunker on the French voices' natural phrasing; and improved morphological analysis for German voices, leading to more accurate pronunciations/phrasing on compound words.

Regular maintenance and bug correction have been performed on all other languages in Acapela's portfolio.

The addition of audio boost provides extra loudness, for further volume expansion when it is activated to the maximum level.

Acaqpela also enhanced the spelling mode, optimized memory usage, and enhanced the synchronization on words, especially for groups of words being highlighted, as single words that are now highlighted word by word.

Developers using the SDK can also now add a voice mode load option. The registry configuration has been replaced by file configuration, support for SAPI4 has been removed, and handling for special HTML/XML meta has been replaced replaced by a single character. Fix all SSML "says-as" tags have been replaced by single-space characters, the spelling tag has been replaced by an Acatts RMS tag, and the URL file in SSML tags is now supported. New tags to support Acatts-based exceptions/abbreviations (dynamic exceptions/abbreviations and files) have also been added, allowing users to load 10 files at a time.

SpeechTek Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues