The state of the STL

[Carl Zwanzig]

I've been out of this for a while...

What are stations using these days for wireline STL? Direct T1? ISDN? Any
IP-based systems? (Would you trust "the network"?) What about backup links?
Anyone still have analog lines?

Thanks,

z!

[David Kaye]

"Carl Zwanzig" <zb...@radix.net> wrote

> What are stations using these days for wireline STL? Direct T1? ISDN? Any
> IP-based systems? (Would you trust "the network"?) What about backup
> links? Anyone still have analog lines?

Well, apparently in some parts of the country where Verizon (former General
Telephone) has wireline service, they want to eliminate ISDN. This is
causing engineers to scramble who are dependent on ISDN for things like STLs
and remote broadcasts.

The big issues is the latency and flakiness of IP. ISDN works fine for
remotes when you can set up one each way and the announcer at the remote can
hear the station's audio in realtime. But you can't do that with IP. The
packets are simply delayed too long on all but the most robust connections.

I have no idea what got into Verizon but I guess they figure they're not
making enough money on it to maintain its infrastructure.

[Eric Weaver]

It's a mix of Analog-FM and digital microwave on 950, T1 lines (though
those are less well maintained by the carriers as time goes on), and
some other stuff. ISDN is only used for backup facilities as far as i
know.

The posher places are using digital microwave AFAIK. KFJC is still on
analog FM. KKUP as well.

There is one local company that just set up a private broadband digital
microwave (TDM-plus-packet) from HQ to all their sites, with the hub on
a prominent mountain in the area. The carrier-based T1 lines will be
all gone when that goes live.

[John Higdon]

In article <kiu3pq$tsf$1...@dont-email.me>, Carl Zwanzig <zb...@radix.net>
wrote:

At my little station group, we will shortly be using our own digital
microwave network to carry programming to main transmitters and
boosters. No AT&T. No Verizon. No crappy Moseley toy UHF gear or lame
digital STLs. We're talking the real thing: 300Mbps, bi-directional
(symmetrical, of course), IP and native TDM, and it is all ours. No
monthly charges, on licensed bands that are far better coordinated than
any Part 74 nonsense.

I'm not at all sorry to kiss all those ancient means goodbye. The
equipment is already in place and the licenses granted. We're just
taking care of last minute details regarding our new LAN.

--
John Higdon
+1 775 253 3838

[Neil]

Two questions, purely for curiosity:

1. Is it fully redundant? Not just the gear, but also the pathing, such
that there's no single point of failure that can leave one or more
locations dead in the water?

2. What's the payback period, and when you hit it, what's the
anticipated savings?

[John Higdon]

In article <51606ce1$0$52777$742e...@news.sonic.net>,

Neil <weis...@sonic.net> wrote:

> 1. Is it fully redundant? Not just the gear, but also the pathing, such
> that there's no single point of failure that can leave one or more
> locations dead in the water?

Yes, we have redundancy but it is a fall-back just to keep things from
going silent. But that redundancy constitutes last resort backup. The
system itself is hardened to a great degree against weather, electrical,
and single radio failure. It's a pretty tough bird to kill.

But are you aware that not only is any failure in the system far less
likely than the failure of a telco T1 circuit, but the repair of any
failure can be accomplished far faster than it takes to get telco to
restore a downed T1. If we could live with telco, we will thrive with
this when it comes to reliability.

> 2. What's the payback period, and when you hit it, what's the
> anticipated savings?

The payback period is around two years. After that, the monthly cost
(for rent at one relay site) is down in the noise, so to speak.

[joewo]

I see the topic is STL and local delivery but what about long haul
real time content delivery?

If ISDN is going away what about those hosts who do their shows
remotely...often times from another state. How will they get their
content back to HQ if not by ISDN? Are there national hosts doing
their show via IP?

[John Higdon]

In article <nhh6m85ffbtm8jtph...@4ax.com>,

joewo <noe...@noemail.org> wrote:

> I see the topic is STL and local delivery but what about long haul
> real time content delivery?
>
> If ISDN is going away what about those hosts who do their shows
> remotely...often times from another state. How will they get their
> content back to HQ if not by ISDN? Are there national hosts doing
> their show via IP?

Long haul is routinely done via IP, and yes, there is content regularly
distributed via the Internet that ends up being broadcast.

--
John Higdon
+1 775 253-3838

[David Kaye]

"joewo" <noe...@noemail.org> wrote

> I see the topic is STL and local delivery but what about long haul
> real time content delivery?

Everything -- EVERYTHING is moving to IP. Think about it. Every piece of
equipment ever made and ever thought of that needs any kind of intelligence
has/will have its unique IPv6 addresss. Name it: printers, traffic signals,
intestinal probes, gas spectrometers, cell sites, cell phones, cameras,
sound and video recorders, automotive air pollution test stations --
everything.

In time the IP infrastructure will get to the point where packets move
faster and through more direct paths so that delays become less and less and
are wiped out enough that the ear won't hear them.

Buildings today are being built and remodeled for ethernet and/or fiber, not
for telco lines.

While a lot of broadcasting is still routed via ISDN, it's being replaced
daily. And as said here or elsewhere (maybe I said it?) Verizon (the wired
company, ex-GTE) is not selling anymore ISDN in some of its service areas
right now.

[spamtrap1888]

On Apr 8, 11:39 pm, "David Kaye" <sfdavidka...@yahoo.com> wrote:
> "joewo" <noem...@noemail.org> wrote

I don't recall "network congestion" affecting ISDN, but it does affect
IP.

[John Higdon]

In article <kk0a6i$tu0$1...@dont-email.me>,

"David Kaye" <sfdavi...@yahoo.com> wrote:

> While a lot of broadcasting is still routed via ISDN, it's being replaced
> daily. And as said here or elsewhere (maybe I said it?) Verizon (the wired
> company, ex-GTE) is not selling anymore ISDN in some of its service areas
> right now.

The date for that cutoff of new ISDN orders is in May (as was discussed
right here at the NAB Convention today). We're talking about dial-up BRI
service, not point-to-point PRI service, which many broadcasters still
use for STL duty. Ultimately, that service is going away, not because
telcos stop offering it, but because broadcasters ultimately will not
tolerate the cost and the increasingly poor reliability of these
circuits.

We will continue to employ PRI circuits (otherwise referred to as TDM)
for the foreseeable future. However the service will be provided by our
own equipment, owned and maintained by us.

Telco no longer cuts the mustard. By the way, this state of affairs is
something I publicly predicted more than five years ago. A lot of people
were surprised to find that no, Ma Bell will not be providing acceptable
communications service forever. A lot of srambling is now happening as a
result.

[David Kaye]

"spamtrap1888" <spamtr...@gmail.com> wrote

> I don't recall "network congestion" affecting ISDN, but it does
> affect IP.

Agreed, but ISDN is a real circuit and IP is a virtual circuit. Just about
everything you can think of is moving from real to virtual. Two-way radio
networks have moved from real channels to virtual channels (trunking).
Telco phone numbers haven't reflected real phone lines in several decades.
It's simply more cost-effective to switch data via trunking than to use
dedicated lines.

IP will get better because customers will demand it. Network congestion
will eventually be gone. Comcast has upped its throughput at least twice in
the 7 years I've been using them for internet. Used to be that a ping from
SF to bbc.co.uk took about 300ms. Now it takes about 159ms. That's almost
double the speed.

There are several STL units on the market using IP. There, of course, are
various streaming codecs designed for broadcast. They'll both fall flat
with today's IP because of congestion. But give it time.

[David Lesher]

John Higdon <hi...@kome.com> writes:


>The date for that cutoff of new ISDN orders is in May (as was discussed
>right here at the NAB Convention today).

Are there any cites for that?
--
A host is a host from coast to coast.................wb8foz@nrk.com
& no one will talk to a host that's close........[v].(301) 56-LINUX
Unless the host (that isn't close).........................pob 1433
is busy, hung or dead....................................20915-1433

[John Higdon]

In article <kk43jr$her$2...@reader1.panix.com>,

David Lesher <wb8...@panix.com> wrote:

> John Higdon <hi...@kome.com> writes:
>
>
> >The date for that cutoff of new ISDN orders is in May (as was discussed
> >right here at the NAB Convention today).
>
> Are there any cites for that?

Nope, just something heard in passing from people who knew. It isn't a
topic of enough interest to me to have compiled a bibliography.

:-)

[Neil]

On 4/10/13 9:42 AM, David Kaye wrote:
> ...Telco phone numbers haven't reflected real phone lines in several decades.

> It's simply more cost-effective to switch data via trunking than to use
> dedicated lines.
>

Well, not exactly. The telephone numbers that we dial are one level of
abstraction from the actual circuit numbers that are used to identify
the physical path, much like a URL gets translated into an IP address
through a DNS inquiry. Someone who has a hard-wired telco circuit all
the way to the CO could just as easily still have that circuit
identified by the tel#, but someone (like me) who's been transitioned
from copper to the CO over to a remote terminal (copper to the RT and
then fiber between the RT and the CO) can have the change made largely
in software. (Yeah, I do realize there are a couple of hard wiring
changes needed too.)

Further, this system facilitates number portability. If you want to drop
telco for a cell phone and port your number over to (say) VZW, this
makes it easy to do, compared with your phone number identifying one
specific hard circuit.

> IP will get better because customers will demand it. Network congestion
> will eventually be gone. Comcast has upped its throughput at least twice in
> the 7 years I've been using them for internet. Used to be that a ping from
> SF to bbc.co.uk took about 300ms. Now it takes about 159ms. That's almost
> double the speed.
>

But only a part of that is the local loop to Comcast's headend. Much
more of it is the infrastructure behind the scenes, the backbone
networks, the core routers, the Beeb's infrastructure of network and
servers, etc. (All facilitated by the Great Infrastructure Buildout of
the late 1990's, followed by the Dot Bomb bust.)

[David Kaye]

"Neil" <weis...@sonic.net> wrote

> Well, not exactly. The telephone numbers that we dial are one level of
> abstraction from the actual circuit numbers that are used to identify the
> physical path, much like a URL gets translated into an IP address through
> a DNS inquiry.

Well, not exactly. When I was in the call center business eons ago, we had
DID (direct inward dial) lines, which were trunks. We had 15 or so of these
lines but thousands of phone numbers. Dial a number in a block that was
assigned to us and it would come in on the next available trunk. I can't
remember the exact blocks, but I believe we had 415-995-2500 through 2699
and 2800 through 2999. And then 415-267-3000 through 3399, and I forget the
rest. So, though we had thousands of numbers we had only 15 actual talk
paths.

Our customers had either DCF (delayed call forwarding where the call would
forward from their number to ours after X number of rings, or they used
simple call forwarding into the number we assigned them.

This was over 20 years ago and it wasn't even new technology then.

[Dave Barnett]

On 4/10/2013 9:42 AM, David Lesher wrote:
>
> Are there any cites for that?
>

I received the info on a list server for broadcast engineers. Several
engineers on the East Coast say Verizon in New York and Maryland will
discontinue taking orders for ISDN this May. Apparently this is cited
in Talkers Magazine, although I do not have a copy and it doesn't appear
to be online anywhere.

So no solid documentation that I can provide, but it's being discussed
by credible individuals. We switched from ISDN last year and never
looked back:

http://www.thebdr.net/articles/audio/codecs/RI-ZIPOne.pdf

We get around the congestion problem by increasing the buffer size. I
realize that's not practical for some, but it works fine for our broadcasts.

Dave B.

[Dave Barnett]

On 4/10/2013 9:42 AM, David Lesher wrote:
>

> Are there any cites for that?
>

Aha. Found it.

http://www.talkers.com/2013/03/28/verizon-no-longer-taking-orders-for-isdn-service-in-northeast-starting-may-18/#more-17845

Dave B.

[Eric Weaver]

On 4/10/13 10:49 PM, Dave Barnett wrote:

> We get around the congestion problem by increasing the buffer size. I
> realize that's not practical for some, but it works fine for our
> broadcasts.

It gets to the point where you have to do the remotes as a "blind
broadcast" and go by the clock or countdowns (or talk on the cell
phone). Not as handy as full-interactive but it gets the job done.

[Dave Barnett]

On 4/11/2013 12:33 PM, Eric Weaver wrote:
>
> It gets to the point where you have to do the remotes as a "blind
> broadcast" and go by the clock or countdowns (or talk on the cell
> phone). Not as handy as full-interactive but it gets the job done.
>

Indeed. With profanity delay and/or HD radio the luxury of monitoring
live is gone anyway. Our programmers have become quite adept at
countdowns. The other method is to use bumper music from the studio and
let the remote people talk over it for the first couple of minutes, at
which point the board op just fades out the music and the show runs from
the remote site.

Dave B.

[Phil Kane]

On Thu, 11 Apr 2013 15:33:17 EDT, Eric Weaver <we...@sigma.net> wrote:

>It gets to the point where you have to do the remotes as a "blind
>broadcast" and go by the clock or countdowns (or talk on the cell
>phone). Not as handy as full-interactive but it gets the job done.
>

When we did remotes in the '50s - before Marti units - we had two
phone lines - one for the program and an order wire to the studio.

Phil Kane
Beaverton, OR

[Eric Weaver]

And indeed even well into the Zephyr-using days I have ordered up
dial-tone lines to go along with the ISDN, for order-wire doubling as
last-ditch backup (Winter Music Conference remote for Energy comes to
mind). Couldn't run a DJ live mix on that but could at least do
interviews.

[mike...@gmail.com]

John,

What gear are you using for this?

-mike