[EAS] What is the solution for the TTS problem?

[Ed Czarnecki]

Alex --> Agreed.  In addition to our included standard speech synthesis tool
on both the DASDEC and OneNet, we have also been using an advanced product
for years on our EAS systems.  One major TV broadcast group contracted our
advanced speech synthesis solution for all their stations.  Notably, we have
many units in campus environments, which almost exclusively rely on text to
speech conversion.  We have had good experiences using these advanced
solutions.  And, yes, one benefit is that the pronunciation lexicon can be
relatively easily customized.  That can be handled either by the end user,
or we do that for them.

Ed

Edward Czarnecki, Ph.D.
Senior Director - Strategy, Development & Regulatory Affairs
Monroe Electronics, Inc. / Digital Alert Systems
ed.czarnecki at monroe-electronics.com

www.monroe-electronics.com
www.digitalalertsystems. com

-----Original Message-----
From: eas-bounces at radiolists.net [mailto:eas-bounces at radiolists.net] On
Behalf Of Alex Hartman

Cepstral is as close to a perfect paul as you can get. It's cheap, and it
even knows how to properly pronounce Eau Claire amazingly.

There are several TTS engines out there in the market, PerfectPaul was of
course commissioned by NOAA, but also not available to purchase.

Most of the TTS engines on the market today are pretty good, probably around
90-95% accurate. It's not my fault Monticello is pronounced "Monty-Sell-O"
here, and "Monty-CHELL-o" somewhere else.

With many of them though, you can update the lexicon by hand for certain
stuff, but that would absolutely be left up to the end user which is where
the problems come in.

--
Alex Hartman

On Thu, Jan 19, 2012 at 1:02 PM, Barry Mishkind <barrym at oldradio.com> wrote:
> At 11:22 AM 1/19/2012, Tom Taggart wrote:
>>The big boxes (Sage, Dasdec, Tri-lithic) are all computers.
>>I am assuming if heads were beat together to use a standardized 
>>text-to-voice program it could be made to work on all three & just 
>>downloaded at a later date.
>
>        Well, that was my unstated concept.
>        Are we afraid of some idiotic conspiracy,
>        or collusion?  Or can we accept that
>        Perfect Paul (or some other file) be
>        made the "standard?"
>
>        Does the government own PP, or
>        would that create (oh my, lawyers
>        salivate) royalty issues?
>
>        Maybe the EAS manufacturers could
>        combine efforts in an "open software"
>        effort to make a standard?
>
> _______________________________________________
> This is the EAS Forum Discussion List
>
> Please invite your friends to join our Forum!
> http://lists.radiolists.net/mailman/listinfo/eas
>
> And, remember the main page: http://eas.radiolists.net
>

_______________________________________________
This is the EAS Forum Discussion List

Please invite your friends to join our Forum!
http://lists.radiolists.net/mailman/listinfo/eas

And, remember the main page: http://eas.radiolists.net