Calgary, Alberta Ð Trillium Sound Research last month shipped TextToSpeech Kit for the NeXT, a synthetic-speech development kit and user application that uses the Digital Signal Processor to generate an acoustic resonance model of vocal chords.
A main dictionary that phonetically describes 100,000 English words is backed up by letter-to-sound rules for unknown words. Developers and users can create custom dictionaries, using phonetic symbols to describe their favorite words or pronunciations. A bundled pronunciation editor, PrEditor, uses phonetic-symbol sets derived from Merriam-Webster's dictionaries; the International Phonetic Association; and the work of David Hill, Trillium vice-president of research and professor of computer science at the University of Calgary. PrEditor manages speed, pitch, rhythm, and stereo placement. The user application, Big Mouth, can be accessed from the Services menu to read any selected text.
"The NeXT is particularly well suited for sound research," said Hill, a former researcher with the Phonetics and Linguistics Department at Edinburgh University, Scotland. "It fit perfectly with my research interests."
Possible uses envisioned by Hill include the ability to retrieve NeXTmail text messages over the phone or automated voice-reminder systems. Since a user will need only one installed copy of the program for any program to take advantage of text-to-speech capabilities, getting the product bundled is on the agenda. Said Hill, "We think every NeXT should have the capability of speaking."
Single-user development kits cost $290; the user license is $145.
Trillium Sound Research is at 1500, 112 Fourth Ave. S.W., Calgary, Alberta T2P 0H3, Canada and can be reached at 403/284-9278 or 403/282-6778 fax.