Thursday, September 06, 2001


Microsoft's Vocal Chord

SEATTLE


To view the entire article got to http://www.washtech.com/news/media/11867-1.html

Aiyaiyaiyaiyaiy.


Such is the approximate sound Shelley Reynolds emits when she locks the tip of her tongue behind her teeth and extends it, like a back flip, as far as it will protrude. It is not a pretty sight. That is followed by a series of well-calculated grins, which don't look funny. Then she kneads her jaw.


Those are oral calisthenics. Reynolds is a professional. Her job: to use her trained voice to greet millions of people every day across the United States.


You may already recognize the voice of El Edwards, the man behind AOL's famous "You've got mail" line, which reaches the ears of more than 30 million subscribers when they log on to the nation's largest online service. Microsoft Corp. has taken a page from AOL's playbook, giving its No. 2-ranked MSN service a feminine touch to help humanize the silicon-and-metals-and-plastics body of the computer.


Lately, the two tech giants have been clashing over a variety of competitive issues, from Internet access to how to design software. But in the sound of a human voice, they have arrived at a common industry pursuit: to make technology friendly.


For Microsoft, Reynolds is the soul in the machine when its subscribers log on and hear her cheerful voice greet them with "Good morning," "Good afternoon" or "Good evening." (She also bids people "goodbye.")


"I honestly believe I have enough positive energy — this is totally an Aquarius-type deal — if I record these things thinking very positively, I kind of believe metaphysically it cheers people up. They'll get a positive hit," Reynolds says.


Her voice — a gravelly yet soft contralto — is apparently doing some good, if only to reassure consumers that in technology there is nothing to be afraid of.


"There's a comfort level that comes with a voice," says Mike Uretsky, director of the Center for Advanced Technology at New York University.


Fear of technology goes back more than 200 years, he says, when the invention of the power loom caused people to wonder whether machines can think. Even now, the mystery of technology inspires fear.


"The large majority of people don't understand it," Uretsky says. "The thing that gets press is the bad things. The thing that gets into the press is the [computer] viruses, the invasions of privacy. This [voice] becomes an important step, saying, 'Hey, I'm almost a person.' "


The power of voice goes even deeper: "As it turns out, humans are genetically designed to understand verbal language," says Rich Gold, manager of research in experimental documents at Xerox Corp.'s Palo Alto Research Center. "It's the primary means of transfer of knowledge and information."


Voice conveys different meanings, depending on the age and sex of the implied speaker. People respond to what is familiar and to the frequency and pitch of a voice, which in the case of women tends to be higher and thus easier to hear, experts say. Through voice — the lilt in expression, the shades of tone — people connect in a way that words on a computer monitor cannot, a distinction becoming increasingly obvious as people spend more time alone hunched over a keyboard, seeking virtual contact with other people through Internet chat rooms, bulletin boards and messaging services.


"The humanizing of the machine is more than just about voice, it's about language, it's about linguistic interaction," says David Nahamoo, head of human-language technologies at International Business Machines Corp.'s Thomas J. Watson Research Center in New York. "We want to simplify our interaction with the world around us," he says. "The world around us is getting more and more complex."


Nowadays, everything and everybody seem to be talking. Computer don't just talk to people — people talk back, using voice-recognition technology. Automated teller machines remind users to take our receipt. Subway trains warn to stay clear of the doors. Even cars are getting into this verbal thing, barking out map directions.


Fast-growing MSN, which has nearly 7 million subscribers, recognizes the trend and credits the intangible benefits that come with Reynolds's voice. (Her services cost substantially more than the $100 that AOL reportedly paid Edwards in 1989 but well short of six figures.)


"It's worth it for the emotional connection," says Hillel Cooperman, group manager of MSN's user-experience team.


Increasingly, companies are using human voices to personalize Internet sites and services, not only because it makes good marketing sense but also because of technological advances in the past year. Voice — transmitted as data on a computer network — can now be compressed into smaller file sizes.


"It's just starting," says Adam Goodman, president of Voicehunter.com, a New York-based Internet company that specializes in providing voice-over talent.


But Edwards, the voice of AOL, says that Microsoft is coming late to the table with Reynolds's online greetings. "I'm rather surprised it has taken them this long to do it," he says. "I don't think they can improve on what AOL started — what — 12 years ago."


When Microsoft took the leap, introducing Reynolds's voice last year, the company chose someone who is like many of the consumers it's trying to reach: a stranger to the technology. Reynolds, a 34-year-old Seattle actress, has spent all of about two or three hours online in her life, and she doesn't plan to use the Internet "until I can go up to a computer and say, 'Fix me a turkey sandwich.' "


"I'm a total Luddite," she says.


Reynolds had done a number of commercial voice-overs before, including a Microsoft webcast of a U2 concert. She heard about the potential gig at MSN through her talent agency.


When she came in to record her voice, she told Cooperman, the MSN executive, that she didn't know anything about the Internet, and he responded, "Perfect."


"He did not want a formal telephone operator sound," she says. "He wanted it to be warm and personable. He said he wanted it to have soul, and when he said that, I started to understand."


Cooperman and a few other Microsoft engineers and technology wonks selected Reynolds from a pool of about 25 candidates, drawn from talent agencies and the company's own employees. There was nothing scientific about it, he says. Microsoft, the world's largest software maker, made a gut call in choosing Reynolds.


"She sounded sincere, smart, friendly, just someone you'd be psyched to know," Cooperman says. "In the end, a lot of this is subjective. . . . When we heard the voice, we said, 'That's her. Shelley's it.' "


The first thing you notice about her is the black, wing-tipped shoes, minus socks. There's the black pedal pushers that begin at mid-calf and work up to a black tank top and faded seersucker button-down. Then there's her fire-red lipstick and an unruly copper bob.


At least four days a week, Reynolds says she is apt to be found at the Lobo Saloon & Cafe, her neighborhood blue-collar bar, wielding a mug of beer, or perhaps a Jack Daniel's on the rocks, and an American Spirit smoke.


"I'm not really sophisticated," she says. "I'd rather run with people who are construction workers and bricklayers rather than CEOs."


Her crowd also includes casino croupiers, such as her 26-year-old boyfriend, Matthew "Hawk" Ramsay, who stops by the bar in his Calvin Klein pajamas to see his lady. In these confines, she's a celebrity whose photograph graces the arts page in the local yellow pages.


"She's got that southern whiskey voice," says bar buddy Derek Reese, the 34-year-old chef at a nearby restaurant.


Reynolds, a New Orleans product, says her parents "sound way more country" and her two brothers "sound more Cajun." With years of acting training, she sounds more like a cross between Cybill Shepherd and Kathleen Turner.


Like her well-known counterparts, Reynolds tends to play brassy roles. In her latest incarnations on stage, she's been playing "white trash with a heart of gold" parts, she says. In her last local production, "Killer Joe," she was required to come on stage in nothing more than a T-shirt.


For the rest of the world, Reynolds's identity is supposed to remain a mystery. The idea is, she is whomever the MSN subscriber wants her to be. To further connect Reynolds to consumers, her greetings are often accompanied by the subscriber's name.


Reynolds will limber up her voice — what she calls her articulator — with a variety of techniques, which include repeating this phrase several times: "The lips, the teeth, the tip of the tongue, the back of the gums, the alveolar ridge."


To date, Reynolds has recorded about 10,000 names in 14 sessions over 30 hours. In some cases, she will record a name — such as Tamara — in different ways to cover every possible pronunciation.


Sitting on a stool, clutching a sheaf of paper listing names, Reynolds will recite each one as though she knows the person. In some cases, it will be the name of a boy she had a crush on in school years ago. Other times, she will come across a name such as Bruce and imagine she is saying hello to Bruce Willis, the movie star. In her voice is an implied message, she says, a tone that conveys a thought like "Did you enjoy your bagel?" When she gets tired, she will step out of the studio, sit in her 1981 Chevy Luv pickup and listen to country music.


The studio work is paying off. Some friends who signed up for MSN have called her and said, "Dude, that's you!" One friend and neighbor, Jay Martell, became a Microsoft subscriber last year and instantly recognized Reynolds's voice.


"She can say anything to you, and her voice vibrates," says the 29-year-old glass artist. "It picks up a certain audio decibel, and it goes right into your brain, and you can never forget it."







Copyright (c) 2001 by Post-Newsweek Tech Media Group, Inc. All Rights Reserved.

No comments:

Post a Comment