This investigative project is mostly the background to the voice behind the legendary voice mail system, that has been branded AUDIX (the acronym known as Audio Information Exchange), Intuity, Modular Messaging and smaller systems like Partner and Merlin Messaging. Technical information or specific dates or years is not part of the narrative because she doesn’t have that information. Regardless, the early days of the enterprise voicemail system has some interesting history in itself.
Despite her claim to fame, she was not the first voice of Audix. According to her, a woman with a Texan drawl (the person’s name is unknown) had done the prompts for at least Release 1. The Bell Labs team wanted the voice to sound more New York, however they didn’t know where to go. Hey I wouldn’t blame them too. In the world of business, if you had a Texan (or heck someone from the West Coast) giving you prompts, would you go asleep or a loose a prospective customer? Especially when a product of AT&T was about to evolve into the competitive marketplace during the time Divestiture?
A man who had once worked on a Bell Labs project of a system with an A/V interface that could bridge such equipment in various rooms or classrooms through a telephony system; was tasked to find the voice. The said project is believed to never gone to market. This manager called a film producer in the Yellow Pages and asked he knew any voice over talent. The film producer had recommended a radio talent to the Bell Labs manager. They spotted a radio news reporter in the Denver market who worked at KADE in Boulder, then KADX going by the name “Lauren Hendricks.”
Despite the illusion of multiple personalities (read below), the woman they found would be Lorraine Nelson.
(On a sidenote: I guess name spoofing wasn’t just isolated to the world of Shadow Traffic or Metro Networks reporters! I never understood concept of a same voice, but different names on different radio stations – thought it was always a slap in the face to the listening audience’s intelligence.)
Editorial aside, this was probably the best move. A native (and now a resident again) of Connecticut, as she told me where she “grew up to speak properly!” who also studied at the University of Colorado with a Communications major – not the telecommunications, but in the radio, TV scope. She met their crieteria – but could she pull it off?
After the discovery and making the decision, she would arrive to a frugal Bell Labs factory, with low end technology with no quiet place to record since this was a manufacturing plant. Not only that, apparently AT&T could’ve paid her a little better for such an enormous task.
How come? What they had was a reel to reel tape deck in a cubicle, and apparently according to her they wanted to mimic (in her words) a “telephoney” sound. Because of this low tech practice in a dark time in the 1980s; people didn’t like the voice, and it felt too quick or abrupt. They gave her another chance to re-record the fragments this time they didn’t over direct her. By this time she interjected her own personality (and from seeing that other video – this would explain the “nicest secretary” vision.) This seemed to help according to Nelson and kept it for Audix 1 and 2 (again released in mid to late 1980s most likely.)
Despite publicized peer reviewed reports on the System 75 in January 1985, with developers touting the design of the human in mind, the Audix team apparently didn’t have the interests of the users in the beginning according to Nelson. She would record nearly a thousand prompts (known formally as “fragments and menus”) into the phone at the factory after the work day ended there and had to go through each one and dial it in to be able to record, enter the fragment number, press a command to playback, and if it didn’t sound well to hit a command to rerecord. She was annoyed at how she would record it without any problems, but the system would cut off part of her speech.
Essentially what she did was no different than a customer getting root or Administrator access to the system and basically change the voice prompts, because in modern voice mail systems if you dislike the voice over you could in theory rewrite their voice. (If only I could get those 100 prompts to rewrite my Asterisk box it would be so awesome!) In the early days, there was no studios, no MP3, WAVE or AIFF PC/Mac based files; this was a simple rewriting over the voice of that Texan woman by logging into a telephone and press buttons to do the overwrites.
Because the AUDIX history (at least in the mid 80s) is hard to find and hardware probably been vanished from Earth (and thank you Avaya for destroying your historical collection!) I could possibly speculate how they would reproduce the new voice on newly produced systems. I can imagine that the new AUDIX became test machine for the new voice in the shop; perhaps take a backup of the new voice and just insert them into the new systems and do it over and over – since afterall she was recorded this first version on the shop floor. (And phone systems, mind you, don’t get reproduced often like PCs, like making thousands a day. A lot of times, these specialized systems would be made specifically for their customers near the time of purchase.)
Reflecting this primitive procedures, one would consider the production or development of this system as a glorified answering machine despite the very high rich, expensive nature of the equipment. First thing I compared this to today’s standards was, if say a friend with a good voice you wanted on your answering machine, and he recorded it on an MP3 send it you and play it off a BlackBerry. (I did this before we moved to a new house and use the service provider’s voicemail.) Or even 25 years ago if you wanted friend to record a quasi professional recording on a micro cassette player for you on those other tape based answering machine.
She would come in for changes over time including additional work for Lucent (by this point) recorded with enhancements to AUDIX, a faster pace, voicing over for two commands per prompts (oh I mean “fragments” and “menus”) and provided the voice for the Partner Messaging and Merlin Messaging (mid to late 90s) as well and even the IP Office in the last decade. And by this time it was more professional and was recorded digitally as well.
She is still strong and active, and still heard by millions (including your humble currator, when his mother occasionally misses a call made by me to her office set.)
Note: Many thanks to Lorraine for her cooperation, answering and the prompt response through email. If only there would be more people that could send meaningful email within a few minutes, it’s a rare exception.