[EAS] 2019 EAS National Periodic Test -- The Good. The Bad. The Ugly.

[Dale Lamm]

Sean (and the list),

RE: the attention tone removal... Our operational area went through this a few years ago. The solution was a Sage firmware update to version 89.2 or better, and careful adjustment of the attention threshold setting inside the configuration file. We reached a point where a 3 or 4 hop off-air relay was successful in stripping the attention tones at each point. A "fresh" tone was applied by each relaying station (and properly stripped off by the next station in line). Note that if you don't strip incoming tones, 8 seconds of the message buffer is lost to non-spoken information. Over multiple hops, these 8 second tones can stack up and constrain the amount of spoken emergency information that is relayed. Here is a snippet of an old email from our state ECC, which does a fine job of helping participants use EAS successfully.

[ from ECC ]

I send this info to you as an active link in the EAS LP relay path from C-bus to Cleveland.     We know of the cascading two-tone attention signal problem as it propagates from here to Dave.    In speaking with support at Sage, and verified in testing between EMA and WNCI, we've figured out how to prevent this.    The 'ATTENTION THRESHOLD' setting is defaulted at 23.   This is too large a value (meaning it is LESS sensitive) to discern two tone and spoken word.  At default, it allows the two-tone to get embedded with the audio.   We chain this from EMA to NCI to VNO to HBC to TAM........and you know the end result.  A recent software update from Sage (if you are on 89.2, you have it) allows an easy fix for this.
 
Suggested setting is now '5'  (the virtual setting is relative).   The LOWER this number (setting) the MORE sensitive the box is to determine what is two-tone (or the NOAA/NWS 1050Hz alert tone - which is a separate threshold setting) versus what is real voice/audio.   With the setting at '5', it will 'restart' the recording in RAM between the gap of the two-tone and the message, thereby 'erasing' (deleting) the two-tone. 

On Thu, Aug 22, 2019 at 1:42 PM Sean Donelan <sean at donelan.com> wrote:
>On Thu, 8 Aug 2019, Sean Donelan wrote:
>> Switching and cut-off audio.  Some stations don't program sufficient delays
>> when switching EAS audio, and sometimes cut-off part of the EAS data bursts
>> or audio.
>>
>> WASH-FM: EAS AFSK header decode error (part of header cut-off)
>> WJFK-FM: EAS AFSK header decode errro (part of header cut-off)

>Good news.  WASH (FM) and WJFK (FM) fixed their digital switch timing
>issues during this month's RMT (Wednesday morning in the DC area). Both
>had a clean decode of the AFSK data burst and this month's test audio.

>Bad news.  SIRIUS XM broke the timing on their digital switching and
>cut-off the triple EOM data bursts early at the end of the RMT audio this
>month. I still have no idea what the appropriate audio levels for the
>SIRIUS XM barker channel is. It varies widely.

>A question for EAS box vendors -- Can the EAS Attention Tone removal
>process be improved? Or better guidance for configuring the audio chain
>and transmitter's processor for the attention tone allowing clean
>detection of the attention tone? Obviously, every station will configure
>their audio processor for regular programming and paying advertisors,
>not EAS.

>__________________________________________________________
>The EAS Forum Discussion List is hosted by the BWWG (Broadcast Warning Working Group). The Core members of the BWWG are Adrienne Abbott, Clay Freinwald, Suzanne Goucher, Barry Mishkind, David Ostmo, Darryl Parker, Richard Rudman, Gary Timm, and Sharon Tinsley. http://eas.radiolists.net
>Please invite your friends to join our Forum! The sign up is at: https://lists.radiolists.net/mailman/listinfo/eas
>___________________________________________________________

--
Dale Lamm
Director of Engineering
WHBC AM-FM
Alpha Media
330-471-1583 office
330-575-3873 mobile