MacSpeech Updates Dictation Software

MacSpeech today released MacSpeech Dictate v1.2—the first major update to its speech recognition solution made exclusively for Apple's Macintosh computers.

"Basically version 1.2 adds to the functionality that we had in the first version," says Andrew Taylor, founder and CEO of MacSpeech. "The first version gave us unbelievable functionality, and it worked just about everywhere. In version 1.2 there were a couple of things we wanted to add to it."

Among the updates to MacSpeech Dictate are:

• Spelling Mode, which allows users to dictate any specific word, name or acronym—regardless of its obscurity—by spelling it letter-by-letter.
• Phrase Training, which improves speech recognition accuracy over time by allowing users to train the application to recognize words or phrases that have been identified incorrectly.
• A "Move" voice command that allows users to relocate the cursor within a document for easier vocal editing.

According to Taylor, the updated MacSpeech Dictate version learns quickly and is more accurate because of the Phrase Training function. "As you’re dictating along, if the program makes a mistake with what it is you’ve said then you can immediately pick one of the alternatives," he says. "That will update your voice profile, it’ll update your document, and you can continue on with your dictation."

Taylor also notes that the Spelling Mode is further enhanced by providing users with the option of using the international radio alphabet—particularly for recognizing letters like C, B, D, and T. "By saying things like charlie, bravo, delta, and tango, it’s clearly something that’s very, very different. The program is able to pick up on that immediately, and so now you can spell just about anything," Taylor says.

MacSpeech Dictate 1.2  also offers interface enhancements and a maintenance release that fixes reported issues. The update is free  and now available for registered customers.

According to Jeff Ganyard, vice president of engineering at MacSpeech, the update had been in the works for a long time. "We’ve been working on it from the moment we released 1.0, if not actually before that," Ganyard says. "And actually, we were working on it before that. We have been working on these features for quite a while."

Ganyard says there were a number of issues that had to be worked through for the update, but notes that "The upside of that is that it gave us an opportunity to address everything we were finding out that users were encountering."

MacSpeech uses the same underlying speech recognition technology used in Nuance Communications' Dragon NaturallySpeaking dictation software.