Tired of typing? Make yourself heard

Scott Simmonds admits he's a poor speller - and a lousy typist, on top of that.

As an insurance and risk-management consultant who sends lots of e-mails and writes articles, white papers, and books, Simmonds's productivity was being hampered by the keyboard. So, he turned to voice recognition software.

"Now I can focus on my thinking and put stuff down on paper without worrying about the writing," says Simmonds, who works out of his home office in Saco, Maine. "My fingers no longer get ahead or behind of my thoughts."

Voice-recognition software is evolving into a $1 billion market. It has long been used for automated customer service attendants, simple data entry, and dictation.

Now, language response systems are embedded in cellphones (say "Call Mary" instead of dialing) and global positioning systems, as well as toys (a diary that opens when you say the password.) And voice controlled lamps, clocks, and remote controls have been introduced in the home electronics arena.

With only a few brands on the market - IBM, Microsoft, and the leader, Nuance-Dragon -voice-recognition software has made huge leaps, with better algorithms, advances in processing power, and improved microphones.

Simmonds, for example, tried an earlier version of his current software but abandoned it when it often made mistakes like interpreting the word "cat" as "hat." When he purchased a new version four months ago, he found that the accuracy had amazingly improved.

Still, even the best speech-recognition software has limits. For instance, the use of voice recognition at home - whether it's to dictate commands to your PC or type papers - is still in its infancy.

And the software must be trained to recognize a user's voice - your accent, inflection, pronunciation, and other idiosyncrasies - meaning that it can't be easily used for conducting interviews with unfamiliar people the software doesn't "recognize." But the software learns as you use it, improving accuracy.

"People have Star War images of being able to sit down and immediately start talking to their computer, but anyone with that expectation is bound to get frustrated and give up immediately," says Dana Abramson of Achieve Results, an upstate New York company that specializes in training clients in effective use of speech recognition software and has offices in Mansfield.

Abramson warns that speech recognition software is just another input tool. "You should still use the mouse, keyboard, and microphone when it makes sense," she says. "Sometimes, for example, instead of trying to control the menu by using your voice, it makes more sense to just click the mouse."

Although it is possible to virtually control the entire computer by voice - moving up and down paragraphs, opening menus, and maneuvering around the screen - this takes time, says Abramson. "Don't try to learn too much too fast. Master the essentials, because there are thousands of voice commands."

But Abramson says there is an added benefit to using voice recognition software. As long as you speak fluently and enunciate, the software understands you, says Abramson, but if you start speaking with blurred, jumbled speech, it can't interpret what you're saying.

The result? "Your speech and communication skills, get very, very keen," Abramson says. "It's boosted my public speaking incredibly."