WEBVTT

00:00:01.170 --> 00:00:05.030
Before anything else, I gotta give you the usual

00:00:05.030 --> 00:00:10.410
spiel. This is my own opinion, my own point of

00:00:10.410 --> 00:00:13.710
view, and none of it speaks for anybody I work

00:00:13.710 --> 00:00:17.269
for currently or have ever worked for. It's just

00:00:17.269 --> 00:00:22.789
me. So, a while back, I'm standing in a transmitter

00:00:22.789 --> 00:00:26.010
building, and everything in that room appears

00:00:26.010 --> 00:00:30.089
to be perfect. Listening to the sounds from the

00:00:30.089 --> 00:00:36.450
magic tile transmitter it's happy power is good

00:00:36.450 --> 00:00:40.729
the antenna is fine every light that should be

00:00:40.729 --> 00:00:46.649
green is green and yet the station was off the

00:00:46.649 --> 00:00:53.969
air not because anything in the room failed because

00:00:53.969 --> 00:00:59.270
well the audio stopped showing up the thing that

00:00:59.420 --> 00:01:04.560
carries the sound from the studio to the building

00:01:04.560 --> 00:01:10.560
had quietly died somewhere out there and the

00:01:10.560 --> 00:01:14.159
transmitter was sitting there doing its job perfectly

00:01:14.159 --> 00:01:19.140
it was just transmitting nothing but the carrier

00:01:19.140 --> 00:01:27.120
that link it sort of has a boring name it's called

00:01:27.120 --> 00:01:32.689
the STL And that stands for the studio to transmitter

00:01:32.689 --> 00:01:37.510
link. And it is at most stations in this country,

00:01:37.849 --> 00:01:41.810
the single most important thing. Nobody has a

00:01:41.810 --> 00:01:47.430
backup for. This is the Tyler Woodward project

00:01:47.430 --> 00:01:50.409
where people like tell me how wrong I am on everything

00:01:50.409 --> 00:01:53.909
in the comments section. And well, today I'm

00:01:53.909 --> 00:01:58.469
talking about that one cable. That one radio

00:01:58.780 --> 00:02:03.739
RF path or that one circuit that can take your

00:02:03.739 --> 00:02:07.379
entire station off the air while every other

00:02:07.379 --> 00:02:11.099
piece of expensive gear is Sitting there working

00:02:11.099 --> 00:02:22.020
fine All right, let me set the scene your studio

00:02:22.020 --> 00:02:26.280
it's in one place Your transmitter is usually

00:02:26.280 --> 00:02:30.139
somewhere else out there on a hill, on a tower,

00:02:30.520 --> 00:02:34.080
out on the edge of town, where the coverage usually

00:02:34.080 --> 00:02:39.719
works out a little better. Sometimes that's a

00:02:39.719 --> 00:02:43.860
couple of miles. Sometimes it's a lot more. So

00:02:43.860 --> 00:02:47.580
you need a way to get the audio from here to

00:02:47.580 --> 00:02:51.819
there, from point A to point B, all day, every

00:02:51.819 --> 00:02:55.780
single day, without ever taking a break. That's

00:02:55.780 --> 00:03:00.889
the STL. That's its entire job in the problem

00:03:00.889 --> 00:03:07.590
with the STL is structural it Sits in the middle

00:03:07.590 --> 00:03:10.949
of the chain. It is single. It is the single

00:03:10.949 --> 00:03:15.090
most important piece everything downstream of

00:03:15.090 --> 00:03:20.810
it can be absolutely flawless and It really doesn't

00:03:20.810 --> 00:03:23.370
matter because if the audio doesn't make it across

00:03:23.370 --> 00:03:30.819
that link your broadcasting silence. There's

00:03:30.819 --> 00:03:34.300
a guy named Doug Irwin who, you know, writes

00:03:34.300 --> 00:03:37.360
about this stuff in radio world. If you haven't

00:03:37.360 --> 00:03:39.639
seen his articles on radio world, go check it

00:03:39.639 --> 00:03:41.379
out. He's got a couple of good ones on here.

00:03:41.599 --> 00:03:43.539
I shouldn't say a couple. He's got way more than

00:03:43.539 --> 00:03:47.939
a couple. But, you know, he's a serious engineer.

00:03:48.379 --> 00:03:51.159
This guy knows his stuff and he put it about

00:03:51.159 --> 00:03:54.960
as plainly as you can put this. An STL system.

00:03:55.759 --> 00:03:59.319
I mean it's just as crucial in a radio station's

00:03:59.319 --> 00:04:04.740
air chain as the transmitter or the antenna.

00:04:06.120 --> 00:04:12.520
If there's no audio to transmit, even the best

00:04:12.520 --> 00:04:18.079
transmitter and antennas are of little use. It's

00:04:18.079 --> 00:04:21.259
kind of useless. Nobody tunes in to hear dead

00:04:21.259 --> 00:04:23.939
air. Boom right out the gate right there at the

00:04:23.939 --> 00:04:28.459
top of this article from back in 2018 That's

00:04:28.459 --> 00:04:32.500
the whole thing right there Nobody tunes in to

00:04:32.500 --> 00:04:38.319
hear dead air So why does almost everybody run

00:04:38.319 --> 00:04:45.519
just one of these STL systems? Money mostly a

00:04:45.519 --> 00:04:48.819
real backup basically doubles what you spend

00:04:48.819 --> 00:04:56.540
and there's Sort of this inertia this it's you

00:04:56.540 --> 00:05:01.120
know, it's worked for 15 years. So Why would

00:05:01.120 --> 00:05:05.019
I touch it? If it ain't broke don't fix it. How

00:05:05.019 --> 00:05:09.300
many times have I heard that saying? and honestly

00:05:09.300 --> 00:05:13.660
Fewer and fewer stations. I mean have an engineer

00:05:13.660 --> 00:05:18.420
around To even build out the backup because those

00:05:18.420 --> 00:05:23.730
people get consolidated out years ago. And the

00:05:23.730 --> 00:05:27.230
gear is old. There's a guy out there who does

00:05:27.230 --> 00:05:30.069
this stuff for a living, runs a company called

00:05:30.069 --> 00:05:33.750
Max Connect, and he walked into a station recently

00:05:33.750 --> 00:05:38.170
that was still running an STL transmitter built

00:05:38.170 --> 00:05:44.230
in 1988. That radio is older than, you know,

00:05:44.310 --> 00:05:47.889
a chunk of my audience. It's still on the air.

00:05:48.919 --> 00:05:53.860
Until the day it isn't So let's talk about you

00:05:53.860 --> 00:05:56.519
know the actual ways you can get audio across

00:05:56.519 --> 00:05:59.720
because there's more than one option nowadays

00:05:59.720 --> 00:06:03.899
and This is where it's going to get very very

00:06:03.899 --> 00:06:09.480
interesting an old Classic is a licensed radio

00:06:09.480 --> 00:06:13.800
link in the 950 megahertz band You point a dish

00:06:13.800 --> 00:06:16.959
from the studio at a dish on the tower. It's

00:06:16.959 --> 00:06:22.189
all line of sight and you basically beam the

00:06:22.189 --> 00:06:24.930
audio across that link. It's point to point.

00:06:25.790 --> 00:06:30.769
When the path is clear, it's great. Low latency,

00:06:31.790 --> 00:06:35.370
it's your own license frequency, no monthly bills.

00:06:36.550 --> 00:06:40.730
The catch is there's barely any room left in

00:06:40.730 --> 00:06:43.490
the band, especially if you're in a busy market.

00:06:44.649 --> 00:06:51.040
The gear is also very expensive. And if a building

00:06:51.040 --> 00:06:55.680
goes up in your path in your way or you get interference

00:06:55.680 --> 00:07:04.160
You're kind of stuck Doug Irwin counted it out

00:07:04.160 --> 00:07:11.199
once and figured there's something like 16 usable

00:07:11.199 --> 00:07:14.899
channels in that whole band and you know in a

00:07:14.899 --> 00:07:19.439
big city where everybody's antennas you know

00:07:19.439 --> 00:07:22.139
converge on the same hill you're gonna run out

00:07:22.139 --> 00:07:27.300
fast then there's IP audio over the internet

00:07:27.300 --> 00:07:31.920
basically it's sort of over oversimplifying it

00:07:31.920 --> 00:07:35.680
but this is where most new STL installs are going

00:07:35.680 --> 00:07:39.500
you take the audio you packetize it you send

00:07:39.500 --> 00:07:44.079
it over a circuit could be fiber could be a regular

00:07:44.079 --> 00:07:46.899
internet connection could be your own private

00:07:46.899 --> 00:07:53.939
microwave link and this is the part i actually

00:07:53.939 --> 00:07:57.620
you know really want to push on because not all

00:07:57.620 --> 00:08:01.199
internet is the same not all internet is the

00:08:01.199 --> 00:08:06.160
same internet if you can get fiber get fiber

00:08:06.160 --> 00:08:10.060
and if you can get two fiber circuits that don't

00:08:10.060 --> 00:08:13.120
ride on the same backhaul that don't funnel back

00:08:13.120 --> 00:08:17.000
through the same path somewhere upstream. That's

00:08:17.000 --> 00:08:21.240
the best thing going. That's the dream. Because

00:08:21.240 --> 00:08:24.839
the whole point of a backup is that the two paths

00:08:24.839 --> 00:08:31.480
don't die together. That's the part I think some

00:08:31.480 --> 00:08:36.100
people miss. You can have two circuits that look

00:08:36.100 --> 00:08:40.000
separate on paper and both of them run back through

00:08:40.000 --> 00:08:43.049
the same junction. you know, five miles away.

00:08:44.129 --> 00:08:49.149
In the day, a backhoe finds that junction. Both

00:08:49.149 --> 00:08:53.210
of your circuits are going down at the same time.

00:08:53.950 --> 00:08:56.789
Backhoe fade. Most radio engineers know that

00:08:56.789 --> 00:08:59.830
term. It's a real term because it's a real problem.

00:09:01.549 --> 00:09:07.269
Always, always have a plan for the backhaul dying.

00:09:08.289 --> 00:09:13.620
Not just the last mile to your site. And please,

00:09:13.960 --> 00:09:16.879
for the love of God, don't build your station's

00:09:16.879 --> 00:09:20.399
lifeline on a residential internet plan. Business

00:09:20.399 --> 00:09:25.580
class, at the absolute minimum. Enterprise, if

00:09:25.580 --> 00:09:28.100
you can get it, if you can afford it, that's

00:09:28.100 --> 00:09:31.379
the good stuff. That's where you get an actual

00:09:31.379 --> 00:09:34.659
service level agreement. And somebody who picks

00:09:34.659 --> 00:09:38.419
up the phone, I mean, seriously, the difference,

00:09:38.600 --> 00:09:41.419
between residential and enterprise is the difference

00:09:41.419 --> 00:09:45.379
between we'll get to it when we get to it and

00:09:45.379 --> 00:09:51.100
we're already on it. DSL and cable, I'd stay

00:09:51.100 --> 00:09:54.620
away from both of those. DSL because it runs

00:09:54.620 --> 00:09:56.879
on copper and we know the phone companies are

00:09:56.879 --> 00:10:02.539
actively trying to walk away from copper. Let's

00:10:02.539 --> 00:10:06.919
consider that a dead technology at this point.

00:10:07.690 --> 00:10:13.889
and cable because, well, of latency. Coax just

00:10:13.889 --> 00:10:18.149
isn't built to be consistent the way you need

00:10:18.149 --> 00:10:21.830
a station feed to be consistent. I mean, it'll

00:10:21.830 --> 00:10:25.690
mostly work right up until the moment it doesn't.

00:10:26.649 --> 00:10:31.850
And then there's the third option. This one has

00:10:31.850 --> 00:10:35.830
gotten really good over the last few years. Cellier.

00:10:36.870 --> 00:10:42.549
4G and 5G. You can run a modem as a backup that

00:10:42.549 --> 00:10:45.269
kicks in within seconds when your main path drops.

00:10:46.309 --> 00:10:48.970
The nice ones, I mean they can bond multiple

00:10:48.970 --> 00:10:52.490
carriers together so you're not betting everything

00:10:52.490 --> 00:10:56.470
on one cell network having a good day. Peplink.

00:10:58.070 --> 00:11:01.289
They sell many different routers where you can

00:11:01.289 --> 00:11:05.110
do Bonding, you can bond multiple cell networks

00:11:05.110 --> 00:11:08.090
together. You can bond other networks that aren't

00:11:08.090 --> 00:11:11.309
cellular, your fiber, your cable, your whatever

00:11:11.309 --> 00:11:12.990
other circuit you've got out there. If you've

00:11:12.990 --> 00:11:15.210
got a WISP, a wireless internet service provider,

00:11:15.289 --> 00:11:18.309
you can run that through a Peplink and bond multiple

00:11:18.309 --> 00:11:21.610
connections together either used in their cloud

00:11:21.610 --> 00:11:24.649
service or a piece of gear that sits in your

00:11:24.649 --> 00:11:30.870
own rack. A simple 5G backup behind a Good fiber

00:11:30.870 --> 00:11:33.929
primary circuit is a I mean, that's a totally

00:11:33.929 --> 00:11:38.049
legitimate setup now. I Mean, it wasn't a few

00:11:38.049 --> 00:11:42.990
years ago, but it is now You know and the other

00:11:42.990 --> 00:11:48.210
thing is when people hear audio over a cell connection

00:11:48.210 --> 00:11:53.490
and they think that eats into a fortune of data

00:11:53.490 --> 00:11:56.860
and yeah it used to especially if you were doing

00:11:56.860 --> 00:12:00.299
raw PCM audio at like a meg and a half just for

00:12:00.299 --> 00:12:05.220
one audio path that's one way but there's this

00:12:05.220 --> 00:12:09.259
awesome thing now called micro mpx it lets you

00:12:09.259 --> 00:12:12.940
send the whole composite signal the full FM stereo

00:12:12.940 --> 00:12:17.480
picture the RDS everything in about 300 to 400

00:12:17.480 --> 00:12:22.879
kilobits a second that's super tiny i mean that's

00:12:22.879 --> 00:12:26.100
that's literally nothing You can run that over

00:12:26.100 --> 00:12:28.659
the most modest connection and it sounds like

00:12:28.659 --> 00:12:35.120
the real thing. And for the gear itself, the

00:12:35.120 --> 00:12:38.299
range is wild. At the bottom of the pile, you've

00:12:38.299 --> 00:12:41.399
got those little barracks boxes. An encoder on

00:12:41.399 --> 00:12:44.100
one end, a decoder on the other end. I mean,

00:12:44.179 --> 00:12:48.000
a pair might run you five, six hundred bucks.

00:12:48.639 --> 00:12:52.419
They're not fancy, but they work. Just put them

00:12:52.419 --> 00:12:56.009
behind a VPN. Don't put them on the open internet

00:12:56.009 --> 00:12:58.789
and if you don't know what I'm talking about

00:12:58.789 --> 00:13:01.750
go Google it I'm sure you'll find plenty of radio

00:13:01.750 --> 00:13:05.169
insight radio world other articles about barracks

00:13:05.169 --> 00:13:08.110
boxes being on the open web, you know and looking

00:13:08.110 --> 00:13:12.289
at Broadcast general store not sponsored by the

00:13:12.289 --> 00:13:16.129
way Just saying, you know, it looks like well,

00:13:16.129 --> 00:13:20.350
I guess the the extreamer 500 is about 900 bucks.

00:13:20.350 --> 00:13:23.440
So I mean that's still that's not too terribly

00:13:23.440 --> 00:13:29.980
bad in the middle is the workhorse something

00:13:29.980 --> 00:13:37.120
like a comrex you know brick link three around

00:13:37.120 --> 00:13:41.139
i think these go for about i don't know 2200

00:13:41.139 --> 00:13:47.600
bucks it's basically built to be an stl it does

00:13:47.600 --> 00:13:51.429
the automatic failover for you it'll jump to

00:13:51.429 --> 00:13:54.690
a cell modem on its own. The guys at Vermont

00:13:54.690 --> 00:13:59.409
Public run something like 30 of them. And the

00:13:59.409 --> 00:14:02.509
word on these? I mean, they just plain work.

00:14:02.850 --> 00:14:07.210
And you forget they're there. That's a high compliment.

00:14:08.769 --> 00:14:11.070
That's the highest compliment you can pay a piece

00:14:11.070 --> 00:14:14.549
of broadcast gear, especially nowadays. If you

00:14:14.549 --> 00:14:17.169
can forget it's there, that's what you want.

00:14:18.269 --> 00:14:22.379
Then you've got options like Tie Line. They make

00:14:22.379 --> 00:14:24.440
great little stuff in the same range. They've

00:14:24.440 --> 00:14:29.659
got the mpx1 and the mpx2 and At the top at the

00:14:29.659 --> 00:14:32.879
very tippy top. You've got the full licensed

00:14:32.879 --> 00:14:37.139
radio links You've got the the old standard the

00:14:37.139 --> 00:14:42.899
moseley starlink sl -900 3q then you've got things

00:14:42.899 --> 00:14:48.799
like the the gates air Intraplex We run these

00:14:48.799 --> 00:14:52.159
at work as our primary STL path to get audio

00:14:52.159 --> 00:14:57.539
to our sites. Think of this, by the time you've

00:14:57.539 --> 00:15:00.279
bought the radios and the dishes and you paid

00:15:00.279 --> 00:15:06.980
a crew to climb the tower, you're easily into

00:15:06.980 --> 00:15:11.799
tens of thousands. In a real, a real backup,

00:15:12.519 --> 00:15:18.289
like the man said, roughly doubles that. Here's

00:15:18.289 --> 00:15:22.309
where I actually land on all of this The fix

00:15:22.309 --> 00:15:26.690
it's obviously not a mystery Nobody has to invent

00:15:26.690 --> 00:15:32.210
anything. We know exactly how to do this to pass

00:15:32.210 --> 00:15:37.809
that don't share a single failure point things

00:15:37.809 --> 00:15:41.350
that make the switch over automatic make sure

00:15:41.350 --> 00:15:45.690
also that something Pages a human when it happens

00:15:45.690 --> 00:15:50.330
That's it That's the whole playbook man, and

00:15:50.330 --> 00:15:55.450
it's been published over and over and most stations

00:15:55.450 --> 00:16:00.509
still They don't do it Not because it's hard

00:16:00.509 --> 00:16:05.370
Because there's no money allocated to do it and

00:16:05.370 --> 00:16:07.870
no engineer left in the building to fight for

00:16:07.870 --> 00:16:13.950
it the same consolidation that Pulled the people

00:16:13.950 --> 00:16:17.309
out is the same reason there's nobody around

00:16:17.309 --> 00:16:23.590
to notice the second that audio drops. That's

00:16:23.590 --> 00:16:28.429
the part that actually really bothers me. It

00:16:28.429 --> 00:16:33.049
used to be when the audio stopped, somebody working

00:16:33.049 --> 00:16:36.970
in the building noticed it. Somebody there heard

00:16:36.970 --> 00:16:40.750
it. And within a minute or two, they were out

00:16:40.750 --> 00:16:47.500
there fixing it. Now, You've got, you know, one

00:16:47.500 --> 00:16:52.500
engineer covering a dozen or so stations across

00:16:52.500 --> 00:16:58.299
three or four different states. And the silence

00:16:58.299 --> 00:17:01.980
detector, when it kicks in, it'll start probably

00:17:01.980 --> 00:17:07.059
looping back some backup audio. In this station,

00:17:07.380 --> 00:17:12.579
technically, I mean, it's alive, but it's really

00:17:12.579 --> 00:17:15.849
dead. And it can sit like that for a while before

00:17:15.849 --> 00:17:20.150
anybody notices. I think back to the transmitter

00:17:20.150 --> 00:17:23.369
building that I started this story with. Everything

00:17:23.369 --> 00:17:26.769
green, everything looking... I mean, perfect.

00:17:28.170 --> 00:17:32.210
But, still, silent. Because the whole thing,

00:17:32.930 --> 00:17:35.529
you know, the one thing nobody backed up was

00:17:35.529 --> 00:17:39.390
the one thing that really mattered that day.

00:17:41.029 --> 00:17:44.059
It's the cheapest... Insurance in the whole building.

00:17:44.319 --> 00:17:48.500
It's it's the thing that gets Skipped time and

00:17:48.500 --> 00:17:51.319
time again a few hundred bucks in an afternoon

00:17:51.319 --> 00:17:56.019
would probably cover most stations and They rolled

00:17:56.019 --> 00:18:00.460
the dice instead Because it's worked so far and

00:18:00.460 --> 00:18:04.359
so far is doing a lot of work in that sentence

00:18:04.359 --> 00:18:09.500
Get a second path make it real make it diverse

00:18:09.500 --> 00:18:15.259
That's the whole episode Subscribe on Apple Podcasts

00:18:15.259 --> 00:18:19.099
Spotify or I heart radio Rate and review the

00:18:19.099 --> 00:18:21.819
show follow me on Facebook and YouTube at the

00:18:21.819 --> 00:18:24.119
Tyler Woodward project Instagram threads and

00:18:24.119 --> 00:18:26.940
blue sky at Tyler Woodward me reach out say hello

00:18:26.940 --> 00:18:32.200
subscribe to the sub stack I do I Guess full

00:18:32.200 --> 00:18:34.180
-on blogs over there a little more in -depth

00:18:34.180 --> 00:18:37.680
than I do here on the show That link is sub stack

00:18:37.680 --> 00:18:40.359
dot Tyler Woodward dot me. I'm independent. This

00:18:40.359 --> 00:18:42.480
is my own thing. I don't have any sponsors, so

00:18:42.720 --> 00:18:46.079
If you like the show, show some love. Just share

00:18:46.079 --> 00:18:49.240
it. That helps. I'll catch you guys next week.