Lans Rothfusz, MIC

For over two years the National Weather Service has been using a computer-generated voice for its NOAA Weather Radio broadcasts. I'll be the first to admit it was not the most pleasant or popular change we have ever made in our services. The clarity of "the voice" and its occasionally-slaughtered pronunciations have been the source of some dissatisfaction by our cherished listeners. In fact, when asked what they would like to hear on the Weather Radio, 45% of visitors to our recent Open House asked for a human voice. Let me assure you, however, we have been listening (to you and to "the voice"), and we are taking steps to bring its quality to near-human standards.

Before I go further, perhaps it is good to reflect on why the voice is used in the first place. There are several advantages to using it, not the least of which is speed! Let me explain.

In the "old" days of humans reading the messages on the Weather Radio, it could take several minutes for a message to be written and then read. Even when it was read, there was no guarantee of clarity, coherence, or correctness, as people sometimes misread words, stumbled over them, or raced through the text making it unintelligible. With different readers, there was little consistency in our overall broadcast. In the case of severe weather warnings, such time delays, mistakes, and inconsistencies are unacceptable.

With our new system, the voice reads exactly what is written, with solid consistency from one product to the next. Once the kinks were worked out (and that took a little while), the use of the voice gave us the "four Cs" we need in a broadcast: Clarity, coherence, correctness and consistency. Moreover, instead of taking minutes to broadcast a warning, it now takes seconds. In fact, with new software we use to generate the warning, the time between a meteorologist's decision to warn and the Weather Radio alarm sounding in your home can be as short as 45 seconds. That speed would not be possible without "the voice."

Despite all these advantages, the voice has some rough edges we would like to correct. Therefore, we are evaluating new voices which you can hear at http://www.nws.noaa.gov/nwr/voicesamples.htm We would appreciate your comments on which voice you prefer. This is your chance to help us decide which voice we will use in the future. In the meantime, keep listening to "that voice!" It's your best connection to the source of forecasts warnings in Georgia! It's the voice of the National Weather Service!

Previous MIC Messages
"Allow Me to Introduce Myself - Again" "Georgia SKYWARN HAMs Upgrade Equipment"

Lans P. Rothfusz
Meteorologist in Charge
(KD5EJN)

lans.rothfusz@noaa.gov