TMCnet News
Loquendo Launches New Text-To-Speech Voice EngineBy MAE KOWALKE
TMCnet Associate Editor for Channels Loquendo, global provider of speech-interaction solutions, today announced the latest release of its Text-To-Speech (TTS) voice engine.
The new engine incorporates improvements in three key areas: fluency, naturalness, and accuracy. The technology provides TTS pronunciation capabilities in 17 languages, using 39 different voices. Loquendo used Concatenation Smoothing to improve the TTS program’s fluency, resolving acoustic discontinuities in speech synthesis.
Phrasing and prosody enhance the program’s naturalness, allowing for pauses and vocal intonations that imitate the characteristics of a human voice.
The program also uses a Morphological Analyzer, which examines word structure to resolve pronunciation ambiguities.
In addition to the improvements to fluency, naturalness, and accuracy, Loquendo’s TTS software now allows users to create their own customized voices by adjusting pitch, speaking rate, timbre, and other settings.
Other modifications include a feature called LexEditor—which allows users to insert their own phonetic transcriptions of specific names or words—and expanded audio controls (such as reverb and stereo) that can be applied to any text.
About Loquendo
Loquendo is a global leader in the speech technologies market. The company develops solutions for text-to-speech and automatic speech recognition. Headquartered in Turin, Italy, the company is a leading technological partner for integrators in the speech solutions field throughout Europe, the U.S., and Latin America. For more information, visit http://www.loquendo.com
-----
Mae Kowalke previously wrote for Cleveland Magazine in Ohio and The Burlington Free Press in Vermont. To see more of her articles, please visit Mae Kowalke’s columnist page.
|