Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
If you enjoy Fully
Modulated, please like, follow,
share and leave a reviewwherever you get your podcast.
Your support helps others findthe program.
I'm Tyler Woodward, a seasonedsenior broadcast engineer for a
(00:47):
network of public media stations.
I've been doing this since2014,.
And I'm currently certifiedfrom the Society of Broadcast
Engineers as a CBNT.
This is Fully Modulated wheresignal meets podcast.
It's a show for the curious,whether you work in broadcasting
or you're just fascinated bywhat happens when you hit play
(01:07):
and that voice comes out of yourspeaker.
We explore the how and the whybehind the broadcast world,
especially from the engineeringside.
Today we're going into theemergency alert system again,
but this time we're looking athow the people and the processes
behind it make it all work.
(01:28):
We're talking automation, humanoversight, alert relays, state
plans and the engineers andstations that keep the system
running even in the most chaoticmoments.
Because the truth is, when analert comes through through,
there's a whole chain of command, a carefully choreographed
(01:49):
system of stations, filters,priorities and decisions before
it ever reaches you, and in themiddle of all of that there's an
engineer who's checking logs,listening to the audio and
hoping that nothing breaks.
(02:23):
The emergency alert system isdesigned like a carefully tiered
pyramid.
Every layer has a job.
When FEMA issues a nationalalert, say, in a genuine
national security situation, itdoesn't go out to every single
broadcaster at once.
It goes first to a handful ofhighly protected, carefully
(02:46):
chosen stations known as thePrimary Entry Point Stations or
PEPs.
These are mostly powerful AMstations that have backup power,
hardened transmission sites andsecurity clearance.
They're built to keep thingsrunning even when the grid goes
down.
From there, the message doesn'tgo straight to your local NPR
(03:11):
affiliate or your favoriteclassic rock station.
It passes through a state relaystation or SRs, which act kind
of like a distribution hubinside of each state.
These are stations thatrebroadcast alerts from the PEPs
and pass them along to localbroadcasters.
Myself, I currently engineer afew SR sites myself and they're
(03:36):
a critical part of the systemPart automation, part watchdog.
Next we reach the local primarylevel LP1 and LP2 stations.
Lp1s are the go-to source formost stations in a given area.
They're monitoring the SRs, thePEPs and often a National
(03:58):
Weather Service source.
Lp2s are designed to be thebackup.
If an LP1 goes offline or isunreachable for whatever reason,
the LP2 can step in.
I've worked at both LP1 and LP2stations and there's a constant
awareness that you're carryingthe responsibility not just to
(04:19):
your own listeners but to theentire regional broadcast
ecosystem.
This whole relay system isenforced by the FCC and
coordinated through state EASplans.
Each station is told exactlywho to monitor your EAS decoder.
A device like a Sage Indec or aDASDAC is set up to listen to
(04:42):
at least two sources and it'salways listening 24-7, waiting
for a valid alert.
Of course there's one morepiece in play, and that is CAP,
or the Common Alerting Protocol.
Cap messages come over theinternet from sources like the
National Weather Service, femaor state officials.
(05:05):
These messages include morethan just audio.
They can have text, graphicsand even location maps, and most
modern EAS gear blendsover-the-air alerts with CAP
messages automatically.
Redundancy is the name of thegame.
If one path fails, say a relaystation's off the air, cap can
(05:29):
still get a message through, orvice versa.
The goal is that, no matter whathappens, the alert gets to air.
So that's the listening side ofthe system.
(05:53):
But what happens when an alertis actually received?
How do stations decide what toair and when, let's say, a
tornado warning is issued forDane County, wisconsin, your
(06:18):
local LP1 station receives itfrom a nearby weather radio
station.
Now what the EAS decoderimmediately kicks in.
First it's going to check theevent code In this case it's TOR
, which stands for tornadowarning.
Then it checks the geocodeusing what's called FIPS.
(06:39):
Is this alert meant for one ofthe counties we're configured to
serve?
If both match and the event isconsidered high priority, it
triggers the automatic relay andgoes to air.
For most stations the process isentirely automated.
Alerts get aired instantly,overriding whatever is on the
(07:03):
air your favorite song, yourfavorite show, it all stops.
The alert tones play, followedby the warning message.
But not every station iscompletely hands-off.
Some choose to hold certainalerts for manual review,
especially things like amberalerts, civil emergencies or
anything from a non-weathersource.
(07:26):
In those cases a live operatoror engineer reviews the alert,
decides if it's relevant andthen manually pushes it to air.
That adds a layer of scrutinythat can prevent misfires.
Engineers also have to thinkcarefully about their filters.
(07:46):
Every EAS decoder has a filtertable, a set of rules that say
what kind of alert could passthrough in for which locations
and under what conditions.
For example, a flood advisory150 miles away probably
shouldn't interrupt yourclassical music programming in
Milwaukee, but a tornado in yourimmediate county?
(08:09):
Well, that needs to be on theair right now.
And then there's the test.
Every station is required tosend a required weekly test, rwt
, and participate in a requiredmonthly test, rmt.
These might just be shortbursts of tone and text, or they
(08:32):
might include audio, dependingon the configuration.
Tone and text or they mightinclude audio, depending on the
configuration.
Tests are vital, not just forthe system's health but for FCC
compliance.
If you're off the air duringyour assigned RMT window, you'd
better have that in your log andyou better have a good reason
why.
Behind all of this are theengineers updating software,
(08:56):
tweaking filters,double-checking coverage maps
and making sure the audioquality stays clean and
intelligible.
We balance automation with humanoversight, because we've all
seen what happens when an alertgoes out with garbled audio, bad
text-to-speech or even thewrong location.
(09:17):
And when something does gowrong, it's not just a technical
issue, it's a trust issue.
Listeners rely on alerts tomean something.
If we interrupt programming forthe wrong thing or too often
for irrelevant warnings, peoplestart tuning them out.
It's called alert fatigue andit's a real thing.
(09:40):
That's the worst-case scenario.
So our job is to make surealerts are timely, accurate and
relevant.
So if you find this podcasthelpful, please take a moment to
(10:18):
like, follow, share and leave arating or review.
Your feedback helps othersdiscover, fully Modulated and
supports our growth.
Now let's dig into what EASequipment is really doing behind
(10:40):
the scenes.
Every message it receives isparsed into four key categories
the event code.
The nature of the alert, suchas TOR, cem, ean and well others
.
The Originator Code, who sentit, like the National Weather
Service, civil authorities orthe president himself.
(11:03):
Fips Codes defining thecounties or areas affected.
The duration how long the alertshould be considered active.
Modern EAS units are smartenough to decide what to air and
what to reject based on thisdata.
They can also log every actionthey take.
(11:24):
That logging is vital.
It's your paper trail.
If the feds come to inspectyour station, one of the first
things they look for is your EASlog.
Did you receive the weekly andmonthly test?
Did you relay them?
If not, why?
Most index and DAS decks letengineers fine-tune these
(11:45):
responses.
You can prioritize nationalalerts over local ones.
You can block certain alertsfrom auto-forwarding.
You can even set up differentpolicies by day part.
Maybe alerts are auto-forwardedduring the overnight hours but
held for review during themorning show.
(12:08):
The deeper challenge is gettingall the tech to play nicely
together.
Eas isn't just one box doingone job.
It has to work with yourstation's automation system,
maybe your satellite feeds andsometimes even remote control
gear.
As engineers, we also have towatch for quirks.
(12:29):
For example, a poorly formattedcap alert can cause a decoder
to glitch or reject a validmessage, or a relay station
might error a message twice bymistake, causing double playback
.
Things like long silence, gaps,badly encoded audio or
(12:50):
misaligned headers can all tripup the chain.
Cap integration, while powerful,brings its own risk.
As one engineer told RadioWorld, the more moving parts you add,
the more you have to test itregularly.
Cap helps reach more devices,but it also adds complexity.
That's a trade-off weconstantly manage.
(13:16):
Ultimately, it's all aboutreliability.
Eas is one of the few systemswe have that can reach nearly
every American in real time.
Our job is to just make surethat it all works.
(13:46):
Today, we explored how EASactually works behind the scenes
.
Eas actually works behind thescenes From the national PEP
stations to state relays, tolocal LP1s and LP2s.
We looked at how alerts arereceived, filtered, relayed and
sometimes reviewed by humansbefore hitting the air, and we
(14:09):
dove into how engineersconfigure and maintain the gear
that makes it all happen balanceand trust, compliance and
clarity every step of the way.
In the next episode, we'll lookat eas in action real world
examples of when alerts workperfectly and sometimes when
(14:29):
they didn't.
What happens when the stakesare high and the clock is
ticking.
If you've got a cool story aboutEAS or maybe a question that
you want answered, text it tothe link in the episode
description.
Whether you're an engineer ormaybe just a curious listener,
I'd love to hear from you.
(14:49):
Thanks for listening to fullymodulated.
(15:19):
If you haven't already, pleaselike, follow, share and rate the
show.
It really helps other peoplefind us.
You can also become a modulatorby supporting the show for as
little as three bucks a month.
All you got to do is visitfullymodulatedcom to sign up and
help keep the podcast goingstrong.
Thank you,