[EAS] Test-To-Speech must be allowed - NOW!

[Sean Donelan]

On Fri, 2 Mar 2012, Eric Adler wrote:
> I suppose the 'right' answer here is to send the SOM and EOM with no 
> content as that's what's available.  Clearly this isn't the proper 
> answer to best serve the American public.

Yep. If you read the EAS-CAP Implementation Guide v1.0, it would have been 
acceptable and compliant for a CAP device to do so (pre-FCC R&O decision).

http://www.eas-cap.org/ECIG-CAP-to-EAS_Implementation_Guide-V1-0.pdf

An ECIG compliant CAP box could be sold without any text-to-speech 
capabilities, because TTS support is optional according ECIG.

If one Local Primary station bought an ECIG compliant CAP device without 
Text-to-Speech support and the other Local Primary station bought an ECIG 
compliant CAP device with Text-to-Speech support, what would be the result 
for downstream EAS stations for CAP-to-EAS messages?

Following the ECIG guide, sending an EAS code-only message would trigger 
duplicate detection in downstream EAS devices, and effectively block other
complete versions of the EAS message from a different source.  The 
different source could be either legacy EAS or another a CAP device at a 
different station which could retrieved either the audio file or had a 
text-to-speech capability to inject the complete EAS message locally.

Radio stations may still benefit from CAP using a first generation digital 
audio file, even one generated by text-to-speech software, via the CAP 
network instead of second or third generation telephone quality analog 
audio through the legacy EAS system. But ECIG also makes digital audio 
files with CAP messages optional.

When there is a CAP-to-EAS conversion problem, i.e. unable to retrieve 
digital audio file or missing text-to-speech support, is it better to 
generate an EAS code-only message? Or log a failure for the operator to 
fix and not generate an EAS message?

>From a emergency management system perspective, rather than a CAP device 
perspective, what is the best answer to serve the American public?

Option A:
All CAP networks must support distributing digital audio files/streams, 
including text-to-speech generated audio files? And CAP devices optionally 
support text-to-speech?

Option B:
All CAP devices must support text-to-speech generation.  And CAP networks 
optionally support distributing digital audio files/streams, including 
text-to-speech generated audio files?

Option C:
All CAP devices must support text-to-speech generation and all CAP 
networks must support distributing digital audio files/streams, including
text-to-speech generated audio files?

If some states choose one option, and other states choose the other 
option, will it cause problems with national or multi-state 
interoperability?

And of course, how would a "live" EAN message be distributed through the 
CAP system, end-to-end, in each option?  Would a CAP system tested only 
with text-to-speech messages be sufficient to assure the system would be
able to carry a "live" message end-to-end?