What's the speech synthesizer the National Weather Service uses?

[Zachary-V]

If you have a weather radio, you can hear one of these broadcasts. (www.weatherusa.net/radio)
These broadcasts include official NWS forecasts and weather-hazard-related emergency alerts, which may be related to severe thunderstorm warnings, flash flood warnings, or even tornado warnings.

The NWS uses a specific speech synthesizer, and I'm wondering if it has a name. And if it does, if anyone's made it publicly accessible?

 

[MagicalMiku]

Interesting question! I guess it might be something related to Microsoft, since it's a company that has many government contracts, so maybe it's a modified speech synthesizer that you can find on Windows 10/11 too. If you go into settings of your pc, you can find many different voices characters for each language you've installed.
Searching a little, seems like that voice is called "Paul" (there are some online text to speech generators). Default voice in Windows 10/11 is David.

 

[Bookworm2]

Looking more into it because this is actually pretty interesting, it seems like they use an updated version of the voice Paul for English called NEW Paul, and a voice called NEW Violetta for Spanish. The new system is called Broadcast Message Handler, and it is not publicly available as of currently. You can view samples of each of their past voices from their official site: Voice on NWR. The old synthesis engine is called DECtalk, and is currently publicly available: DECtalk Archive.

Listening to old vs NEW Paul, the difference is pretty stark. Old Paul sounds far more robotic and unnatural, with extreme emphasis on the wrong syllables, sometimes incorrect pronunciation, and a lot of what we would consider engine noise. NEW Paul sounds much more natural, though the issue of syllable emphasis is still present.

 

[Haiyin-6]

Oh so that's what it is! I've seen people making covers using DECtalk Paul and was wondering where he was from