<img height="1" width="1" alt="" style="display:none" src="https://www.facebook.com/tr?id=311512539006924&amp;ev=NoScript">
  • Telos Alliance
  • Telos
  • Omnia
  • Axia
  • Linear Acoustics
  • twntyfive-seven
  • Minnetonka Audio

RSS header - this is hidden

Get the Internet You Deserve

Posted by Kirk Harnack [TWiRT] on May 19, 2014 9:37:00 AM

Find me on:

TWiRT 212Do you know your D-SLAM from your DOCSIS? ISP and Network Engineer Bob Holowenko explains some little-known but common-sense ways to get the best Internet and WAN service from your ISP choices. Plus Telephony Guru Joe Talbot joins us discussing how to get reliable VoIP service across your Internet connection. And IP Solutionist Chris Tobin tells how he got a big Internet mess solved - sort of - in his New York office building. It’s practical Internet connection help and guidance on this week’s TWiRT.

 

 

Watch the Video!

Read the Transcript!

Kirk: This Week in Radio Tech, episode 212, is brought to you by the Axia Radius IP Audio Console. Radius is the console that proves you can have your cake and eat it, too with easy connections and quick setup. See Radius at axiaaudio.com/radius.

Do you know your DSLAM from your DOCSYS? ISP and network engineer Bob Holowenko explains some little-known, but common sense ways to get the best Internet and WAN service from your ISP choices. Plus, telephony guru Joe Talbot joins us discussing how to get reliable VoIP service across your Internet connection.

Hey, welcome in to This Week in Radio Tech. It's time for the show that you've been waiting for. We do it every week - about every week - right here on the GFQ Network. This is the show where we talk about broadcast engineering, and that's everything from the microphone to the beacon at the top of the tower, and now it extends out into the interwebs and all the digital media that we try to get into. But there are still things that you have to handle at the radio station, whether it's a virtual studio or a real studio; whether you're working with a program director in the next office, or one that's four states away who's programming your station from there. I'll tell you, it's become quite a variety of how radio stations operate nowadays, for good or for bad. I think there's definitely movement in both directions here.

So this is the show where we talk about these days, and on today's show, we're bringing back a guest we've had on before, Bob Holowenko. And Bob is going to . . . well, Bob is kind of an expert at some of the newer features of the IP, the IT, the Telco industry, the things that engineers rely on outside their studios, outside their transmitter plants, often connecting those things together to get audio and other information from here to there. And it's a whole new world, both for the telcos and alternative suppliers of such services. I mean Internet, private WANs, this kind of thing.

So it's new for them, it's new for us, and we're trying to figure all that out. We've got several experts on the show. You know, a lot of services are brought in from telcos that are still telephony-related. Also, Joe Talbot is on the show with us to talk with us and with Bob about how new Telco services and the things that we have depended upon for years, telephony services, can mix and mingle together.

And of course Chris Tobin is with us here. He's the guy who's a Kodak and IP expert, IP audio expert, so he'll be in the mix, too. So I'm probably going to shut up for this episode and we're going to introduce people one-by-one here. Let's first run over to Manhattan. I guess that's where Chris is, the best-dressed engineer in radio. Hey, Chris Tobin, how're you doing?

Chris: Hello, Kirk. Hello, Bob and Joe. I'm doing well, despite the weather conditions that are rolling into the Big Apple, but doing very well.

Kirk: Good deal. Glad you're here. Are you in your office or are you in some secret location again?

Chris: No, today it's the office. I'm in the office, today.

Kirk: Good deal.

Chris: No secret locations today, but next week, maybe.

Kirk: Now the man who does have a secret location, right just within spitting difference of Area 51, it's my buddy Joe Talbot. He's a fellow Telogian. Hey, Joe, welcome in.

Joe: Hi, Kirk. How're you doing?

Kirk: I'm terrific and I'm so glad to have you back. It's good to talk to you. You are the product manager for the Telos line at the Telos Alliance, and so I appreciate you taking an hour to be with us, but you're also a telephony guru, having started in telephony . . . you and AG Bell were kind of working together, weren't you?

Joe: We're tight, yeah. We're very tight.

Kirk: Okay. I think you wrote some of the Bell System manuals, didn't you?

Joe: Practices. Bell System practices.

Kirk: Yeah, practices, there you go.

Joe: No.

Kirk: And now the youngest guy in the group, and the guy who has . . . he's got the information us old fogies need, and that's Bob Holowenko from Canada. Bob, hey, welcome in.

Bob: Oh, thanks for having me back, Kirk.

Kirk: You're very welcome, and we're glad to have you. Well, tell me about the people who you work with on a regular basis, or you work for on a regular basis. What gives you your expertise here?

Bob: I work for a telecommunications company that's also an ISP/WAN service provider. They pretty much do everything end-to-end. As well, I do a little bit of my own work on the side, working for small businesses. So my entire focus day-to-day is large to medium, sometimes even small enterprise, the kinds of needs they have both for Internet, both for WAN services, telephony services, particularly IPTEL. I don't really do a lot with the TDM technologies, although I deal with those terms day-to-day as sometimes issues can cross over.

So a pretty good mix of stuff. I also do a little bit of broadcast engineering, which is how I ended up finding this podcast. So, yeah, a pretty big mix of things.

Kirk: Good deal. Well, we're glad to have you on the show. So I wanted to theme this show around some practical knowledge of what broadcast engineers need to be knowing about or learning about or at least know where to turn for services that they need, and so I'm really not sure where to start this conversation because there's so many areas we could go into.

Why don't we jump right into something that both Joe Talbot and actually all of us, or all of you guys, are expert at - I'm still trying to figure it out - and that is the common practice, or becoming more common, of bringing your telephony services in through a third-party provider. So you've got an Internet provider or provider of WAN or other connectivity, IP connectivity services, and that provider is not providing you with the ultimate, the dial tone. Maybe you're using Vitelity or some other company, Windstream or whomever to bring the so-called dial tone in to you. The SIP provider and the ISP, let's say they're different people. Let's talk about how you try to make sure that that works really well.

And so I'll tell you what, first of all I want to go to Bob. And Bob, kind of give us your view, your take, on that situation, because I feel that's a common situation for small and medium-market broadcasters. The large-market broadcasters will often be getting these services from the same company. They'll demand that, and if it costs extra, they'll pay extra for it.

But stations like mine, we're going to want to get dial tone from a place like Vitelity, and we're going to want to use our regular Internet provider, although we want to get the best Internet we can from that provider. So Bob, kind of charge headlong into this and tell us what we ought to be concerned about in trying to get telephony services into our station not through POTS, not through ISDN, but through the Internet.

Bob: So yeah, using the Internet as opposed to using something like a private WAN service, because a lot of these companies are not carriers. If you're going through your consumer or small business or whatever you want to call it DSL, cable, maybe like a FiOS type solution, you're going to want to do your homework ahead of time to make sure your link between you and whomever you're using is as clean as possible.

It used to be that the local loop, from the provider to the presence, or sorry, to the premise, used to be kind of the more hostile connection. Now the hostility or the danger in the network is going to be at the peering points, so where you have company A meeting company B, there's going to be a lot more saturation. And this is due to the fact that you've got a lot of peer-to-peer type applications that are going to be saturating it.

Things like this conversation we're having right now between my location and I guess Andrew's location is going to be going through a couple peer points. You know, we're not on the same ISP. We might not meet at a common datacenter, so there might be a company like Level3, XO, Tata, somebody in between us. And any time you go company-to-company, because of the volume of traffic, the diversity of the types of traffic, there's zero guarantee. And Internet as itself is a zero CIR service. There is no - there's another one of those terms. CIR means Committed Information Rate.

Kirk: Thank you.

Bob: It's a term from . . . yeah, it's a term that came from back in the frame relay days or even earlier which said you get this many bits guaranteed, no matter what, dedicated. In the case of Internet, an Internet Service Provider can't do that. A lot of people look at the Internet as being this big, important resource, and sure, it is. But it's not a tariffed service like a terrestrial phone service would be, like a landline. You're dealing with something that is always going to be best effort.

The provider is going to sell you let's say it's a 15-meg DSL service or a 50-meg cable service. That's access. They're selling you 15 megs of access to their network. What happens upstream, downstream, wherever - or I guess it can ever be upstream - is not within their control. Once it leaves their network, they can't guarantee anything. So if you're going to be . . .

Kirk: I want to explore that concept for just a second, because that's kind of a hot button in this whole age of discussing net neutrality and are ISPs, are these carriers - should they be classified as common carriers?

Public utilities? I'll just toss out an idea. You brought this up, and I want to spend just a moment on it. And you mentioned you're buying access. I've got 50 megs down and 12 megs up here at my office. That is access to Comcast network. Now that's all I'm really buying, and Comcast does more or less guarantee that 50 megs into their network, or from their network, but they don't guarantee anything beyond that.

 

Personally, I think that's where the problem is. That's where if Comcast would guarantee at least statistically - I don't necessarily need my own dedicated bandwidth between me and every peering point they have, but if statistically I would get essentially that bandwidth all the way into and through their peering points, if I needed 50 megabits from another backbone and I could . . . if Comcast would guaranteed that I would get that whenever I asked for it at the peering point, there wouldn't be a need for talk about net neutrality.

In other words, okay, if I buy a membership in a gym, right, and I show up to the gym and it's oversold, I'm going to have an issue with that. If I buy a plane ticket and I've got a confirmed reservation seat and I show up and they kick me off the plane, which they occasionally do, I'm going to have a problem with that and they'd better reimburse me for that, and they do.

But in the world of Internet, I buy 50 megs and all I've got is 50 megs of access to their network. I don't have 50 megs to other places, at least not guaranteed. Now oftentimes I do get that, but it's not guaranteed. Again, I just think all this net neutrality and talk about public utilities and re-regulation and fast lane and slow lane and all that, all that would go away if we could get the bandwidth that we're buying all the way to the peer points.

Now I realize that if whoever's server I'm connecting to on somebody else's network, if they're not paying for 50 megabits or if they're overloaded, that's not my fault and that's not Comcast's fault; that's whoever I'm connecting to's fault, and there ought to be tools to let you know where the problem lies.

But anyway, that's my diatribe. If we were guaranteed that when we show up at the gym, at Gold's Gym, the machines we want to use are available to use, then nobody has a problem with that if you sell the machines to other people while I'm not using them. But unfortunately at the moment, that's not the way it works. The systems, especially at the peering points as you point out Bob, are oversold. And so when people start watching Netflix that causes a problem because Netflix takes up so much bandwidth. Okay, I'm done with my diatribe. Maybe you can continue on with what you were saying. Go ahead.

Bob: You're absolutely right. What they call that is statistical multiplexing. It's not multiplexing at all; it's predictive usage. So if you've got a peering connection, Netflix is a good example because people are starting to peer directly with Netflix, like their caching system. So that's taking some load off these company-to-company peering locations. So I'm talking peering with XO, peering with smaller guys between maybe a Time Warner and a Qwest. Actually, they're not called Qwest anymore. I'm not so up on my American telcos at this point.

Kirk: CenturyLink.

Bob: CenturyLink, that's right. I should know that, because down in Seattle, just below us, they've got a field. So now that people are moving, big companies that are heavy bandwidth users, the Googles, the Facebooks, Google being as well YouTube, and now Netflix, they're starting to have direct peering relationships. So what that's doing is that's taking the load off company-to-company peers.

So XO, L3, all these big guys whose only job is to carry traffic across the continent or to other continents. So that's good. That takes some of the stress off the local cores. The problem is these cores are still getting hit now with additional traffic. So instead of it coming from company-to-company peer points, it's now coming really fast and rapid from local, more regionalized connection points.

So you still end up with saturation, and that's why all these companies that are trying to do the triple play with the TV services and everything, they're trying to get more and more bandwidth closer and closer to the customer to get around those sorts of things. It's kind of funny that you brought up the net neutrality, because I actually have a note sitting here in my notes that says, "Do not go down the net neutrality rabbit hole."

Kirk: [Laughs]

Bob: But the problem is it's so relevant in this case, because net neutrality could - for certain uses, it could be beneficial to a broadcaster. The way I look at the net neutrality thing is that ISPs need to do something. They either need to do a usage-based billing, or they need to bill higher. As somebody who uses a lot of bandwidth, probably an unfair amount of bandwidth, I should be paying more than grandma who only checks her Facebook twice a day and likes all my posts and checks her email for ten minutes and she's done with the Internet for the day.

Kirk: Yeah.

Bob: So as much as I don't want to pay more, maybe that's a more fair model. Who knows what way they're going to go?

Kirk: I'm going to bring Joe in in just a second, but I've got one more question for you. You mentioned saturation at peering points. If I'm getting my telephony service, my dial tone - if I pick up my phone here, which here at my office, I've got a phone. Let's see if it'll pull over here, right here. Okay, this phone is an IP phone and it's registered with a SIP server at Telos. This is my business phone at Telos, so it's connecting through the public Internet, Comcast on my end, I don't know what on their end.

Joe: Time Warner.

Kirk: Time Warner? Okay, so somewhere they're peering. Whenever I pick up my phone and make a phone call, those bits are going through a peering point, at least one, then to Time Warner and into Telos's SIP Asterisk server there. And you mentioned, Bob, you mentioned saturation at the peering point. When that peering point gets saturated, my bits never were guaranteed but there's probably going to be loss if there's real saturation going on there. Am I correct? And if I am, that's why that peering point saturation affects broadcasters.

Bob: Exactly. So that is an example of all packets are treated equally. Your packets are going to be dropped just like somebody else's torrent traffic or somebody else's, I don't know, some other peer-to-peer protocol. There's Skype, for instance. As far as what you can do to make sure that you're not going to have packets dropped as much or as probable, you know what I'm trying to say, by a peering point, would be to find out what your other provider is using.

So in my case, I'm using . . . I've got a VoIP system at home where I get a DID from a company. It's a Vancouver number because that's where I'm based out of, but they're actually based out of Seattle. So I'm going to pick on Voip.ms just because I know a little bit about their network. They connect at what's called The Seattle Internet Exchange with a bunch of other companies.

But based on trace routes that I've done from various points, I can see that their Seattle cluster actually talks to multiple peers. I don't know if their hosting company that they've got there is doing that, or if that's their own, but regardless they've got multiple different connections in and out.

When my company that I've got my Internet service providing with, or my ISP, met them at what's called a meet-me room in Seattle, the Weston Building there, which is a big data center for anybody who's in the Seattle area, they'll be familiar with it, it's a big peering point.

My ISP connected there not so long ago, and I noticed that my audio quality, or not necessarily audio quality, but my jitter and all of the settings that I can get out of my IP phone showed that my latency was way down, my jitter was way down, and my MOS score, Mean Opinion Score, was actually really good, really high. I had never really had a problem with it before, but the numbers were getting to a point where it was kind of scary. Mostly, it was in the downstream direction which is actually kind of backwards from what I would expect. Usually the upstream direction is where it gets a little scarier.

Kirk: Joe, Joe Talbot, come on in here. Joe, you've been involved with so many installations in the last two or three years of voice-over-IP systems, and even before that before you came on board with Telos. Joe, have you seen - you know, the kind of VoIP I'm talking about where your provider and your ISP are not the same company? The ISP is just providing a pipe between you and your VoIP provider. Are you seeing things generally get better or worse or stay about the same in the last few years?

Joe: Excuse me. I'd say yes. Even here, I'm kind of close to Area 51, at least it's a local phone call from here, so I have really challenging Internet options. I have two DSLs here, I have the cable, and I have a wireless provider. Each of them has its own strengths. I've been using the DSL for voice because it has the lowest jitter and it's the most consistent, and I suspect also that a lot of people aren't trying to watch Netflix on their 1.5 down/384k up DSL which may free us from that, from some problems.

But I had problems about two months ago, maybe three months ago. It was terrible all the time and it was a frustrating thing to do, to use the phone. And then all of a sudden, maybe six or eight weeks ago, it improved dramatically. I heard from a friend that AT&T increased the back-all size and made some other changes in their network that improved things. It's been perfect since, and it's been surprisingly good. The other providers, the cable is terrible. The cable has horrible jitter, although they've increased their speed because they have new ownership.

The wireless people are the most consistent, but just to give you an idea, late afternoon and into the evening, the speed drops from about eight megabits down to about one or two megabits down. It's predictable, and it's got to be Netflix, people coming home and watching TV. So I can't really use that for the phone except for emergencies, although I've got to say the call we're on right now is through that circuit. As long as we don't start doing it at 3:00 or 4:00 in the afternoon, we can probably do this.

Kirk: Hey, Chris Tobin, he's with us and standing by. Chris, I've got the same question for you. See, I'm very optimistic about the public Internet. I'm actually a bit more optimistic about it if we don't put rules on it because I think rules stifle innovation, even though we might think that they solve some other problems. But Chris, in your work, and you've done a ton of IP codec work which is extraordinarily similar, almost the same thing as voice-over-IP telephone calls, Chris, what have you seen about in your area? Public Internet connections, are they getting better, getting worse, or staying about the same at this point?

Chris: I would say they've improved in some areas. In others, they've stayed the same. What I'm discovering working with a lot of folks is the providers, the ISP providers of the traditional legacy telecom world, seem to be the ones that have more difficulty providing relatively decent service whereas the folks like a wireless local loop company or these other I'll call them competitive local exchange carriers, CLECs, they seem to be more I'd say responsive to the market. And when you call up and say "I'm trying to do X, Y and Z with this device and this technology," they take a moment like "Oh, oh yeah, okay. We see what we're doing. Sorry, we can fix that." And then I have no problem.

When I've called up the others, like say - I'll say them, Time Warner Cable; Comcast; Verizon; FiOS; Verizon FiOS and the ilk; or CenturyLink, formerly Qwest; SBC, now AT&T, I get the typical what we're all accustomed to for the last 30 years telephone company response. And it's interesting the dichotomy or the difference between the two but both providing a similar service.

I'm with you, Kirk. I prefer no regulation, leave it as is, let the marketplace move along, and sorry Comcast, if it costs you more money to be competitive that's the way it goes because I like the other three companies who are trying to vie for the same service or same customer base like me. They should have a hand at it, and you not have an upper hand.

So you know, I had a similar experience when I was up in Canada on the east coast, over by Montreal, Quebec and stuff and in Ontario working with a couple of the carriers. I worked with a few private - I'll call them private - same results. Private guys were like "Yeah, I've got what you're doing. It makes sense. Pretty cool. Never saw it before, but we'll work with you. Here are the parameters we have; here are the parameters you have. Thank you very much."

Then when I worked with Bell or Telos or the others, it was the usual

"Yeah, we'll give you what you get. That's it. That's what you pay for."

 

I'm like guys, come on, work with me here. So I'm not sure where things are going to go. I like to be optimistic.

 

The problem here in the States for us, we have the FCC, Federal Communications Commission. The chairman is a former cable guy; he's a former lobbyist. So his interests, I'm not sure where they lie and I'm curious to see where things go. But that's okay, because apparently what's what you want. You need somebody who's been in the business to fix the business. I think Albert Einstein had the best quote. "The same people that got us into this trouble are not the people who are going to get us out."

Kirk: Yes. So guys, okay, so let's imagine that you're a broadcast engineer and you've made the decision to move. Let's say, okay, I keep using myself as an example because I'm an engineer for a group of small market stations, very typical, small town radio stations in Mississippi, in Greenville, Mississippi; Cleveland, Mississippi; Indianola, Mississippi. We have two different cable Internet providers and we have one DSL provider at these stations.

To my knowledge, if I was going to get anything else, any other methodology of Internet, it would have to be a T1 through AT&T. I don't know of any other alternatives, and I don't know of any dedicated wireless providers. Of course there's Verizon 4G LTE in that area.

So here's my point. Whatever I'm going to do, it's probably going to be through the public Internet. It's going to be through whoever the provider is, whether it's Cox or I can't think of the name of the other cable provider, or AT&T. It's going to be through one of these local providers, regional providers.

Okay, so other than ordering the fastest line that I can get from any of them, Bob, is there anything else I can say to a typical Internet provider that would help me get more reliable service? Help me get lower jitter?

Help me get a more consistent speed that is perhaps less affected by other customers in the area? Or am I really suffering at the whim of whatever they built out and what their other customers are doing at the moment?

 

Bob: I don't think a lot of them are going to be too willing to make any changes. Chris mentioned that the smaller guys are more likely to be a little more flexible, whereas the bigger guys are more cookie cutter, do it – they do what they do, and that's about all they do. There are a few things, though, they can be asked. So if you're looking for a DSL service, ask the person who you're dealing with, be it your account rep or a call center person, ask them what the distance is from the CO - they can find out, and that means a central office - or the DSLAM, which is the device that actually provides the digital service. It puts it on the twisted pair.

Ask them what the distance is if it's a long line, because you're dealing with copper as a medium. There's going to be attenuation. As you increase distance, there's more chance for noise, etc., etc. Every broadcast engineer knows this, these theories or these concepts.

So if you find out that it's a really long loop, depending on the technology they're providing, it could be a problem. Things like VDSL and VDSL2, they can go not as far as the older technologies just because of the fact that they're using more spectrum, wider spectrum, but those DSLAMs because of that limitation are usually closer. So you've got to find out and make sure you're not too far away, and they'll have no problem telling you that.

The other thing, too, is these newer technologies, if you are lucky enough to get on something like a VDSL2 or 2+, that kind of thing, they have a lot better error correction. They can do forward error correction based on parity information that's kind of stuffed into the - if you go down into the lower levels, they actually stuff parity information into the headers, and you can essentially have your packet rebuilt, or the damaged cells, because it's all just ATM cells so it's really small. If you lose a cell, it's easy enough for them to rebuild it.

With that, though, comes buffering. We'd call it buffering. I forget the proper term we use, and I see it all the time when I'm dealing with DSLAM profiles, but basically what they do is they put delay in there on purpose. They actually create lag on purpose that gives you enough time. So it kind of works like a jitter buffer or any of the other buffers you're used to. It's not like you would see at TCP level. This is lower down at the cell level, which is just a chopped up Ethernet frame.

So those sorts of technologies are good. As far as what else you could ask the provider? This is one I like to use. Ask how many spare ports are on the DSLAM. If there's lots of spare ports and it's not a busy area, then you're probably in a good situation because you're going to be vying for that bandwidth a lot fewer people. This comes into the difference between the technologies.

With cable, it's a broadcast medium. Everybody is on the same piece of copper. So although you're dealing with coaxial cable that can handle much broader bandwidth, and I mean that in the spectrum sense, you're sharing it with more people.

Whereas a telephone line, you've got a medium that attenuates faster, possibly, maybe not. You've got a technology that is going to be longer-line because the density . . . the way the old cable plants are laid out, unless your vendor has chopped things up a lot better and segmented a lot better, you're probably going to have a longer line which could be a good thing or could be a bad thing.

It could mean that you're sharing with 48 of your neighbors a single gig link. That's good. Or, if you're in a dense urban area, you could be pulled back to a 512 port DSLAM. So the question is often going to be on the right track. If you ask them how many spare ports they have, it could be helpful if they say "Oh yeah, there's lots of spare ports."

Kirk: Okay, that was really good information about the Telco and the DSLAM. I know there's a DSLAM just a block from my house. It seems like they're in it all the time. But there's fiber to the DSLAM. That sounds good. And I believe that, here in Nashville, it appears to me . . . I know a couple of the Comcast engineers here in Nashville because they're members of our SBE chapter here in Nashville. That's really nice.

You've got a former broadcast engineer who then goes to work for the cable provider, in our case Comcast, then still remains a member of the SBE. Hey, great, you get to ask the guy questions. "Oh, really, where's this done?

Where's this done? Oh, you mean all of our data in Nashville goes through that one little hut on the north side of town? That's interesting. Do you have a backup? No?" Those are cool things to find out.

 

So here's my next question, though, and whoever can answer this please jump in. With a cable system, my own experience improved with my provider quite a bit when they searched me from a DOCSYS 2 to a DOCSYS 3 cable modem. And, wow, a lot of problems went away. Maybe it's just a brand new modem; maybe it's a better design. But it's just been flawless since they put this new one in. So there's that. Maybe you can answer that.

I guess what I'm looking at is what's the state of cable Internet providers? Are some of them just laggards and not moving to upgrade their equipment? Just waiting until there's enough complaints? Or are a lot of them just really moving ahead forward with providing the best service and speed they can, which has been my experience? Who would like to begin addressing that question about cable providers?

Joe: I'd say it depends on where Google Fiber is. Seems like every place they announce, everybody starts competing. So other than that, I'm in a place where I have very few choices and it hasn't improved very much. It's not really adequate. But the competition in a market - a market with competition is always much better.

Kirk: Joe, you've got competition there, but you've also got a pretty small population there.

Joe: We've got 40,000 people here. It's a very spread-out area. When I moved here two years ago, all I could get was 1.5 down/384k up dynamic DSL. It's been pretty terrible, but it's been improving somewhat. I can now get 6 meg down/768k up DSL.

The competition from the wireless people seems to really be the driver here. There were a bunch of independent companies, smaller guys, and they were bought up by another company who actually streamlined things a bit and improved everything pretty dramatically. Now the cable people are responding to that, so it's slowly coming around.

But yeah, I guess 40,000 people isn't a whole lot of business. I know the city's trying to get more bandwidth into town because it would actually create jobs and make things better for the people that live here.

Kirk: Who was next? Sorry. Who jumped in right after Joe?

Chris: Oh, this is Chris.

Bob: I guess we all doubled.

Chris: Oh, Bob, go ahead. You're the guest. Go ahead.

Bob: Okay, so basically to answer the question, DOCSYS, when they went from 2 and 2.1 to 3, there were quite a few improvements in the protocol itself. Everything from density, so the number of nodes that could be on a single CMTS I believe it's called, that increased as well as spectral width. A lot of it had to do with the change to DTV and the different signaling. Once they got rid of the analog carriers, or the majority of the analog carriers on the actual coax, they were able to use a lot more bandwidth or kind of move things around.

So it improved a lot of things as far as what the actual protocol could do. So noise immunity, error correction or at least error detection. I'm forgetting all the bullet points here, but yeah, it did quite a bit.

Kirk: Better tech. Chris, what do you have to say about that?

Chris: Well, from what I've read, it's correct. DOCSYS 3 has definitely improved considerably. I can say in my area, where I am, I'm in an urban center and I'm in a building. I only have access to one carrier to get services. Anything beyond . . . I have DSL and cable at the house, and I will say this, the DSL's been the most solid for the last eight years that I've had it, and I'm sure it's more the technology. I'm also one block away from the central office, so it helps. I can actually see it from my window, so it's even better.

However, I will say the cable side of things, whether it's DOCSYS 3 or not, a lot of the problem I have in my building is the quality of engineering and maintenance of the infrastructure. I can tell you that during the wintertime, when the temperature dropped below I guess, yeah, I guess 30 or 29 degrees Fahrenheit, my cable service would start to become macro-blocking and the video would break up and whatnot.

I have the access code to get into the box to look at the signal strength of the channel so I can see what the signal strengths are, and sure enough, they drop considerably. I also have a cable line tester I can use, the same one that's used by Time Warner when they go into the field, and I can look at the signal strength, carriers, and everything on the line. You can see the skewing, distortion and distribution, everything. As soon as the temperature goes up, warms up, and everything's back to what we consider a comfortable environment, signal strengths increase, quality increases, everything.

I've contacted the cable company on several occasions to tell them this. Every time they come out, they replace my cable box with another of the same model, just brand new, and they tell me everything is fine and you don't understand how the system works. I said "Okay, fine, be that way." So I went down to the basement. Since I'm a member of the board, I have access to parts of the building that most people don't. Me being a techy-type person, it's even worse to have access to the building, places like that.

So I went to the place where the cable comes in from the street, a small room, dingy, dirty, as typical. I looked at the box, the distribution amp and the way they break out all the cables and connectors. I took several pictures and sent that to the folks with my other report and said, "Guys, I know nothing, you're right, whatever you want to call it. However, from a visual inspection, these cables that are cut in half with nothing terminated on them, some of this what looks like white calcium deposits on the connector coming out of the wire, it might be an issue but I couldn't be certain. I failed chemistry and electrical engineering, so I just thought you should be aware of this issue." That was it.

Well, three weeks go by and suddenly I see a truck outside. The building manager tells me "I don't know what you did, but there's six guys downstairs in the basement trying to figure out what happened to their distribution amp." It turns out that yes, the distribution amp has been there a long time. Nothing has been changed, because they did find the date codes on it in a tag. But it turns out that's not the actual distribution amp. That's being fed from another amp in a building next to us. That's where the actual feed is. I guess, what should I call it? The proper . . . that's the branch feed that's actually coming from the cable system.

So I'm saying to myself, "Wow, they've got two amplifiers, one amplifier feeding another amplifier." So you can just imagine if the levels are not right and everything's changing. Yeah, it was fun. So now my cable service has been stable because it's springtime, so I can't tell you if it's going to work or not come the fall. I'll have to wait and see, because they didn't change out the distribution box. They put new connectors on the cables.

Kirk: Okay. All right.

Chris: They snipped off the connectors, pulled the cable closer, put a new connector on, and stuck it on the unit.

Kirk: I'm sure they know what's best.

Chris: Yeah, what I'm saying is without the competition, nobody's forced to fix things. Yes, people can complain. But I can tell you, in my building when people complain, all they do is come out and put a box in and claim it's nothing else and just . . . I don't know. I wish there was a better answer, but I'm suspect of where things will go with that.

Bob: The biggest mistake a troubleshooter can make is to ignore the information from the actual user. They're often the ones that will go down there with a flashlight. They'll take a look at what's going on. They've had the exact experience. Nine times out of ten, what they're saying is completely valid. And although you don't want to take it 100 percent because they are prone to exaggeration, don't completely dismiss it. That's a huge mistake that a lot of troubleshooters or field technicians, etc., will make.

Kirk: Good point. Good point, yeah. Hey, we're going to tell you that you're watching This Week in Radio Tech, or listening to it, whether you download the audio or watch the video online. This is the show where we talk about broadcast engineering, and today, our subject on this 212th episode with our guest Bob Holowenko, our subject is trying to wrangle all this new technology, the Internet, and it's been around a while, sure.

But it's getting better and better for the most part, and we're trying to figure out how to make it work for us for the stuff that we need to make it work for. I mean, POTS lines are going away; ISDN lines are going away, and we're going to need to depend on this pipe or several pipes that are Internet connections. How can we best make that work as technology improves and changes?

Joe Talbot is with us. He's with the Telos alliance. He's the Telos Products Manager and a telephony guru for many, many years. Also, Bob Holowenko works for an ISP in Canada and also does station engineering. And of course, Chris Tobin is with us with a great deal of station engineering experience and IP audio experience as well. So that's what we're rolling through today. Hopefully we're giving you some good information about how to help insure that you can get the best out of your ISP and deal with any remaining problems that you have.

Our show is brought to you by our sponsor, and I'm glad to say my employer, the Telos Alliance. And specifically, the Axia Radius Audio Console. If you haven't experienced what an Axia Console and live wire can do for you, I've got to tell you, you're in for a treat, an absolute treat. I was on Facebook late last night and happened to notice that one of our co-hosts, Chris Tar, he's not with us today but he was busy last night installing a new Radius console at a station in Wisconsin. They had suffered a lightning strike. They were about to make some changes anyway, and Chris put in a waiting Radius console there.

And I can tell you, you can put a Radius console in in just a few hours, less than a day. And if your wires are already in the right place, literally just a few hours to get that thing put in and configured. Now Chris and his post on Facebook last night, he said this is he thinks his 12th Axia Console installation. Once you get used to it and you understand what to configure, it becomes very easy.

I tell people if you can order a book off Amazon, you can configure an Axia Console. It's all web-based, lots of click and dropdown menus. You choose what kind of input. Is it a mic input? Is it a line input? Is it a telephone hybrid or a codec? Is it a computer player? You choose what kind of input you've got coming in. You give it some characteristics. Do I want to back feed to it or not? All of these things become very easy and quick to configure. You can add audio processing to mic sources or other mono sources on our consoles.

The Axia Radius Console is a really low-priced console. I'm sorry, I don't have the price list in front of me right now. It's in the range of six to seven thousand dollars for both the eight fader surface and for the brain that rack mounts nearby. It's called the QOR. The QOR 16 is the one that comes with it. It's got mic inputs, line inputs, line outputs. It's got an AES input and output, and really cool, the QOR 16 and its bigger brother, the QOR 32, they have a built-in Ethernet switch.

So this allows you to take the gigabit connection and connect it back to a QOR switch if you have a multiple studio installation, or it also has 100-megabit connections to hook up things like a telephone, a Telos VSet phone, or an audio node or anything else that would be live-wire connected. Even a program delay from 25-7 can be hooked up with an Ethernet connection. An Omnia processor, say an Omnia One, hook it up with that Ethernet connection and you've got control and you've got audio over the same connection.

So I would encourage you to go to the Axia website, it's axiaaudio.com, and look up the console that's called the Radius. It's just as cool as can be. Hey, this show right now at the GFQ Network is being mixed on a Radius Console. The mix minuses to each of the participants - there it is - to each of the participants on the show. Mix minus is automatic. I mean, Andrew Zarian's a smart guy, but mix minus is one level above his thinking. So it just happens automatically, and it's great, because we're all here. All of us participating remotely from all over the world get the right mix minus. It's just as cool as can be.

I've got two Radius Consoles at my radio stations in American Samoa. They went in in a couple of days. I've also got a small console, the rack, the RAQ console in our newsroom in American Samoa. I can access them remotely and make changes if I need to, add a source, delete a source, change the way a source behaves.

Check it out on the web, axiaaudio.com, and check out the Radius Consoles. If your needs are small to medium, you'll be very glad you did because they work really, really well. Thanks again Axia and the Telos Alliance for sponsoring This Week in Radio Tech.

Okay, hey, I want to bring Joe Talbot back in. Joe, boy, you've heard a lot of this stuff. Joe, I know a lot of the telephony you deal with with bigger stations are circuits like PRI and T-1 circuits, and those are well-established. A lot of folks know about them. So I want to stay on the subject of Ethernet, Internet and SIP circuits. We asked Bob Holowenko what you can talk to with your ISP about trying to get the best service that you can. Joe, what do you find out in the field? What's important for the incoming service from a public Internet ISP provider?

Joe: For me, the last-mile options that you have. When I start out a conversation with a customer, I want to find out what options they have, if they've got fiber coming into the building, if they've got wireless coming into the building. Just do a little bit of research to find out if there's any other provider options that they have.

A lot of times, you'll be limited to the utility phone company and it'll be a pair of copper wires and a T-1. And if they're having a lot of problems with their outside plant already, or one of the reasons they're making the improvement or change in the first place is to make things better, if they've got copper problems, outside plant problems, you may have a problem. You may have nowhere to go with that.

A lot of times, I mean typically if there's a fiber provider coming into the building, that's usually where I like to go because the fiber will often have possibly multiple providers available on it. It may be like a metropolitan area network, and more options is always better.

Kirk: The fiber's probably . . .

Joe: Copper's less likely to go down.

Kirk: The fiber's likely new equipment. It's not old, a decades-old plant. It's new equipment installed by freshly-trained techs. That could be good or bad. Well, hey, a friend of mine, a guy who's been on this show before a long time ago, Larry Fuss, he's now got in Las Vegas fiber into the home. He's got a gigabit from . . . I guess, is it from Time Warmer?

Joe: CenturyLink, I think, is who he's got it from.

Kirk: CenturyLink, that's it, you're right. CenturyLink coming into his home, a gigabit. And that's really cool. Is that where we're going, Bob? FiOS is not being built out so much anymore to my understanding. Is there a call, need, and supply going on for gigabit connections?

Bob: Yeah, providers are trying to get closer to the home with more bandwidth so that they can provide more services. A lot of them are starting to go with glass, with fiber with new locations, new sites, big high-density residential areas. When you start to get into the burbs, they've already got a lot of money spent on copper, bringing copper to each individual drop if it's been there for a while. So they want to leverage that as best they can, because that's probably the biggest investment over time has been their copper network.

So you're going to see them bringing the DSLAMs closer to the customer if that's cheaper than bringing fiber directly to them. Often, it probably is, because there's a lot of construction involved with fiber. They're trying to sell more services. They're trying to do TV, faster Internet, dialup. Or sorry, not dialup, dial tone services, all over that single piece of glass or pair of glass.

Kirk: You know, Bob, you bring up a very interesting point. A lot of folks are wondering, okay, what happens when ISPs are no longer motivated to improve their physical plant? And that's why they're trying to sell you triple play. If they sell you triple play, they probably have to improve their plant to make all that work.

People don't want their cable TV . . . they don't want their Oprah show dropping off because there's too much congestion in the way. They want the stuff to work. They want to sell you many, many tens of dollars, even hundreds of dollars, of services every month. So that's motivation for them to build out their network. That sounds like a good thing. I mean, these are for-profit companies. They need to be motivated. They can't lose money selling this kind of stuff.

Bob, earlier you said you're the kind of user that probably ought to pay more. I can tell you here in Nashville, Comcast is doing something that I think is pretty reasonable. I pay about $90 a month for a higher speed home Internet and I guess it's the highest speed home Internet that they offer, and they have a cap of 300 gigs per month. They have had grace periods about that. I busted through all the grace periods, and last month I got my first bill where they charged me extra for the extra bandwidth that I used. I think it was $10 for every 50 gigs that I used during the billing period, and I got dinged for 20 extra dollars. That's 100 extra gigs.

Now in my mind, that's pretty fair. Their base rate, I get 300 gigs. That's enough for most people to watch plenty of Netflix and of course do all their email, that's nothing. But to use other people's services like Netflix, then they charge for the extra. Bob, you mentioned that you think that's fair. Chris, is that how it works in New York or do you have a different system there with the providers there?

Chris: Well, currently where I'm at with Time Warner Cable, they don't really offer anything like that just yet, but I'm sure that it'll be coming down the pike. I'm still on the fence when it comes to the cost-per-byte ratio. I don't know if 50 gigs, $10, there's still something not adding up with all the caps and costs involved and what goes on, but I'm going to reserve that for another time. But I did have a question for the group.

Kirk: Yeah?

Chris: The reason I'm reserving it for another time is because I can't get any other service in my building because the local cable companies have petitioned the municipalities and sued them to prevent them from doing other things. I have a problem with that, but nobody else seems to have a problem.

Kirk: But that's everywhere. I mean . . .

Chris: That's right. That's right.

Kirk: Decades ago, the cable companies convinced city fathers across the country that, "Hey, we'll wire your city out because we're great guys, but you've got to give us exclusivity."

Chris: And times have changed.

Kirk: That deal is everywhere, and that's why you have one cable TV company strung around your city, whoever that may be.

Chris: That also came down from the FCC to subsidize a failing master antenna TV system back in the '70s, and it just continued to grow. That's where it all started from. Municipalities that do have the rights, they have the rights still to do what they want. They can revoke those franchises any time. That's in the charter; I read it. But apparently, again, it goes back to lobbies. They do very well to prevent things from happening.

Kirk: People complain, and probably rightly so, about monopolies. They complain about that. But remember, a monopoly can't occur without a governmental body saying "Yes, you have exclusive rights," in exchange for money or whatever it may be. A company by itself can't create a monopoly. It requires legislative action from politicians to create a monopoly, and that's why for decades we had one phone company and one cable company, and the argument being "Well, you don't want the mess of wires and it's so expensive to build this out that if we don't have exclusivity, we're not going to provide phones to your community or cable TV to your community."

Chris: That's a load of bunk. We all know that.

Kirk: Times have changed. I'm sorry, did you have something else, Chris?

Chris: Yeah, for the group, something I recently read about and I'm curious to see what's going to happen, Spring has begun a push to build out their network so that they can begin providing people with Internet services over their wireless network. So instead of you . . . you could do a 4G modem right now, but they're looking to actually make it a service, almost like say instead of going to your cable company for Internet, you can come to us and we'll provide you a similar service. Anybody have any thoughts on where that may wind up going, or if that's even feasible in this current regulatory climate?

Bob: It's becoming more popular in rural areas. So I know the company that I work for is also a mobility carrier, so they've got 4G like LTE HSPA networks built out. And because they're going to be increasing their footprint in these rural areas even more, mostly because they're upgrading equipment because CDMA is out or older technologies are out and now they're moving to these newer, faster, better protocols that run on these towers, taking down old stuff and putting up new stuff, they're building bandwidth in anyway.

And it's a lot cheaper for them to try to get the most out of that spectrum for people that are in that area. It's cheaper for them to do that than to run copper to these places where maybe they didn't have copper, or where it would be too expensive to upgrade them from anything that can handle more than just dial tone.

Joe: I don't trust those kinds of carriers to provide the level of service that a business actually needs. So far, it's been treated like an optional way, a way for 15-year-old girls on smartphones to watch cat videos. It's not mission critical. The tower sites tend not to have backup power. When there's an earthquake in California, it's really interesting to see what stops and what keeps working, so I'm really suspicious of all that right now.

Kirk: Now Joe, you've got a wireless ISP you said at your house. It's one of the four sources of Internet that you have available. But you said your local wireless ISP is pretty good.

Joe: Wireless ISP is not the same as a cell carrier.

Kirk: Well, you're right.

Joe: They treat . . . the cellular or the wireless providers, cellular companies, tend to treat this stuff a whole lot more casually. I think that sort of - that and the size of the country is why we have such a completely random expectation of performance on that stuff. I've just learned not to trust it. If it works one day, it may not work the next day. I've just not had good luck with it.

Chris: That's true. I would agree with Joe. I would agree with Joe on that.

Kirk: Bob, have you got a follow-up thought on that?

Bob: I tend to agree. The one thing I did find a little bit different is Joe, you mentioned that the tower sites don't necessarily have backup power and that caught my attention because here in Canada we're actually required to have a certain number of hours of reserve battery for telecommunication services simply for the 911 requirement. So that's for anything that could be used for telephony, standard telephony as defined by our government. So for us, that's a little bit different. A phone call, though, will have precedence on the network at any time, particularly a 911 call or anything that's emergency in nature.

With something like a data connection, you'd have a little device which operates just like a phone where it's got a data modem to the carrier and then also either maybe you'll have an Ethernet out or Wi-Fi hotspot on it. That traffic you're going to be paying a heck of a lot more for and have much smaller bandwidth caps than you would with like a cable or DSL connection that's actually wired to your home where you've got kind of more guaranteed bandwidth, air quotes for that one.

Joe: When I bought the house here, obviously being in the business I'm in, Internet is kind of a big deal for us. And factored into the cost of the house was putting in a $30,000 point-to-point microwave system to bring it in. Things have been improving steadily, so I've been holding off making that investment, but it's still not out of the question.

I like the idea of controlling our own destiny, so to speak. I still have to buy the Internet access from somebody down the hill. I would have a 40-mile shot to Mount Potosi then another 11-mile shot down to Las Vegas to get decent service here. So that still may be happening this summer; I'm not sure, yet. But if I can get a grant of 40 grand, I will.

Kirk: If you did all that, Joe, it sounds like you could be an Internet slumlord yourself and resell to your neighbors.

Joe: That wasn't really my goal. I really didn't want to be in that business because you have to deal with customers like that, customers that don't pay their bills and things like that. But with a lot of the different radio systems that are out there now, the ubiquity radios and things, those were basically designed so that the cost of the installation would cover the cost of the equipment.

So if you had a subscriber that didn't pay, you could turn them off and not really lose out on anything. So I suppose it wouldn't be a bad business to be in, and the folks here are making money, but it wasn't really my plan.

Kirk: We've been spending most of our . . . oh, I'm sorry, go ahead, Bob. Go ahead.

Bob: I was just saying I sympathize with the requirement that you've got to have a very stable Internet connection. I'm just about to move into a much more rural location. You know, I'm down to my last 15 days where I am now, then I'm going to be going up to the interior. With my personal requirement, because I work from home a lot - well, I will be now 100 percent working from home - I need stable; I need something that's going to work, because I do a lot of voice-over-IP over a VPN because I take calls or I do presentations online all the time. So I need to be able to have a decent Internet connection.

So I am going with three ISPs. I'm going to be going with our local cable provider, because they've got maybe not as reliable of service, but they have more capacity. I'm going to be going with the Telco that I work for because it's cheap, it's reliable, and that's where I'm going to put my business traffic out, my VPN and stuff. And then I'm going with a WISP, a mountaintop ISP sort of. And I say sort of, because where they are located can't actually see my house, so I'm going to reflect off them using a tower site that one of the local ham clubs has and I'm going to have a pair of ubiquity radios setup so that I can shoot down to my place.

So I'm going to have lots of options, but for that, you need a pretty decent firewall that you've got to be putting a lot of effort into. So I've been putting my money there, and I'm going to be paying it in subscription fees for three different ISPs.

Joe: I have an 8,000-foot mountain that would be the midpoint where mine would have to be, and it's not the kind of place . . . I like going up mountain roads, but I don't like going up this one that much. So if I can possibly avoid doing this, I'm going to, but there are days when I say, "That's it, I'm putting it in. We're done with this."

Kirk: And I've got to tell you, that is a rough mountain Joe's talking about. If Joe doesn't like to be up on that mountain, it's rough.

Joe: Eight thousand feet up, and the road, it's got a single-wire ground return power feed, okay? So in other words, if they add more capacity up there, and there's four class-C FMs up there by the way. If they add more capacity, they have to go and drive a bunch of ground rods. It's very interesting. The whole place is full of rotary phase converters.

Kirk: So hey, guys, we've just got a couple minutes left and I want to cover one more quick question. I'm going to address this to Bob then see if we can get Chris Tobin to follow up with it. Bob, we've been talking about telephony and getting telephony services via your local Internet provider. Okay, and we haven't talked much about the providers of the telephony service. We mentioned a few names, but we'll have to leave that for another day.

So your local provider, what considerations are there that are different, if any, when we switch from talking about radio station telephony for business or studio calls to talking about radio station IP audio communications? High-quality, higher bandwidth. Instead of 64 kilobits per second for a phone call, we're talking about maybe 128 or 320 kilobits per second for a pretty good quality or even high quality AAC-encoded audio. What consideration might be different?

Let's say instead of occasional phone calls or regular phone calls, I want a connection that's going to be on almost all the time between say my studio and my transmitter, or from my studio in one city to a studio in another city where they do a great morning show, and I want to pipe that in and I've got IP codecs on each end. Do the things that we talked about for reliability and low jitter and good connectivity, do those things apply just as well to that scenario for IP audio?

Bob: Yes and no. I say yes because of the fact that if you're dealing with a company that is providing your Internet as well as your RTP SIP session, or even if they're not, if they are a competitor to whoever you're using, chances are they've got something in place where they know how to treat RTP packets and you might be able to get them or they might already be treating your telephony service with a higher grade.

Audio streaming services, those be it a UDP-based service or an HTTP-based service like an Icecast or something like that, it goes up like any other packet. So the answer to the second part is no. You'd have to be doing your own protection. You can't do anything about the downstream other than limiting the connections that exist downstream, so don't pull a bunch of stuff down if you want to protect your downstream. Leave it pretty empty.

Upstream, you have control over. So maybe prioritize your outbound bandwidth. Find a time in the middle of the day when your connection is the absolute worst. Baseline it. Do a bunch of speedtest.net tests throughout the day. Find out what the worst of the worst is, like this is the 7:00 everyone's coming home and torrenting and Netflixing and saturating, then get a more advanced router, shape your traffic so you're using just a little bit under what your actual capacity is, but make sure the priority outbound is given to the things you want to protect: the SIP, the RTP, your streaming session.

Anything that if your secretary is pushing photos of her puppy up on your business connection, make sure that those types of sessions are less prioritized. Really, all you can do is adjust your upstream.

Kirk: That's a very good point. Chris Tobin, real quickly, wrap us up with your experience with ISPs and IP audio.

Chris: I would agree with Bob, and the best thing you can do is when you're talking to your ISPs, explain to them what you're looking to accomplish and see if they have any restrictions or anything in place that might prevent, as Bob pointed out, the different types of packet protocols. And also just manage both ends. Like Bob mentioned, you shape the data packets, do managed switches, and understand your limitations and work with it.

I'm sure if you have bandwidth of 1.5 megs up and you're going to do audio IP, try not to get close to 1.5. Try to do one meg and keep yourself with some headroom, things of that sort. You have all the controls; going up, you do. And even down, you can make something happen.

Kirk: Something we haven't mentioned . . .

Chris: What's that?

Kirk: Something we haven't mentioned, but it's probably worth mentioning, is if you're going to do telephony over the public Internet to bring your phone calls in, and we do that at my stations in Mississippi and it's working fine for us, if you're going to do that and if you're going to use an Internet connection for IP audio transmission, say to your transmitter site or to another studio or to bring in voice talent, if you're going to do all that, it might be worthwhile to have a separate drop, a separate Internet connection for that.

Chris: Oh, sure.

Kirk: Separate from your business connection. As an engineer, you may not always have control over what the other employees, whether it's 5 or 500 employees in the building, are doing. And so if you can have your own little walled garden of Internet connectivity, and I realize you still can't control what's outside of that.

But as we've discussed, a lot of the problems, and I would argue that probably 90 percent of the problems with Internet connectivity occur either in the last mile or inside your own building and in your own router. Get your own connectivity for telephony and IP audio codecs and don't let all that other very bursty traffic or even high-density traffic from unknown business users out there, there's no telling what they're doing, affect your critical on-air stuff. Then you've got the best chance of getting your data at least to and from your ISP. What they do with it after that is up to competition and up to how bad they want to serve your data correctly. Any comment?

Chris: No, absolutely. That's the best approach to take. If you can do that, then you control it yourself. You control the last mile on both ends. If you're in a studio transmitter environment, studio-side, put aside a separate path. Put your revenue-generating codecs on there. At the transmitter end, it's by itself, or if not do the same thing there. Then you . . .

Joe: Yeah. I've got a quick question for Bob, too, when we get a chance here.

Kirk: Yeah, go ahead, quick.

Joe: Yeah, any particular tools you like to use to isolate problems or to find out where pairing-point problems are? To find out, you know, jitter tests that you like or anything you find to be required? That you really need or like to use?

Bob: The basics, so like ping, trace route, those are good ones for very simple things. I'm more of a baseline kind of guy, so something like a ping plotter or MTR, these are things that - you know, get your pen out and write these down. Ping plotter and MTR are two really good tests.

Another one that I really like is called IPERF, it stands for IP Performance. That's a good one because you can simulate what your codec might use. Say you're running a non-variable bitrate at 128 up to your CDN, why don't you just stuff that 128 bits using a test tool, upstream, and see what kind of packet loss you get? That's probably something for a different show, but the tool belt, you should always be familiar with the basics: the pings, the trace routes, and then the other more advanced tools, MTR, ping plotter, there's so many of them.

Kirk: And speaking of ping plotter, by the same company, there's a tool that I use here at my office. It's called MultiPing. It's by this company, Nessoft, but it's for Windows, MultiPing. And it's cool because you can setup . . . I wish I could turn my camera around right here. I can't at the moment. But I plot ping time over time to seven different places around the world of interest, like the zip servers that Telos operates, a SHOUTcast server that my radio stations operate. I ping to our stations in American Samoa to see how good or bad the traffic is there, and to our other stations as well.

It's really interesting. Every evening, US time, from about 5:00 or 6:00 p.m. until 11:00 or 12:00 p.m., the ping times rise slightly and then they go back. Sometimes they rise significantly, and I'm assuming that's due to all the people using their Internet at home and kind of bogging down networks and peering points a bit. But MultiPing is great. And what Bob said, you've got to have a baseline. You can't just measure it once and be done with it. Measure it over time. And so a program that plots that out and measures it repeatedly over time can be really helpful.

It can be expensive bandwidth-wise to measure bandwidth persistently, but you can measure it occasionally throughout the day. Pings are no problem. That's just a small packet to measure the ping time. Guys, we've got to go. I really appreciate all your enthusiasm, your great questions, and your fabulous comments. Bob, thank you for joining us. Bob Holowenko from Vancouver. Are you in Vancouver?

Bob: Vancouverish. I'm in a suburb of Vancouver for now. I'm moving to the interior soon.

Kirk: All right, well we'll keep track of you. Don't let us lose track of you, please. We'll have you back on. And I'd like a report from how you get along with your new situation with several ISPs in a rural area. It's going to be interesting, and to compare that with say Joe's experience there in Area 51. Joe, thank you for being with us as well. I appreciate your time.

Joe: Thanks, Kirk. It's always fun.

Kirk: All right. And the best-dressed engineer in radio from Manhattan, New York, IP solutionist, it's Chris Tobin. Chris, where can people reach you?

Chris: Oh, it's real simple, just support@ipcodecs.com. Drop me a note. I've actually had a couple emails recently this week.

Kirk: Support@ipcodecs.com, sounds good. All right, we've got to go. Our show's been brought to you by the Axia Radius Console. If you would, check it out on the web, it's very affordable. Hey, even I bought a few of them. Check it out at axiaaudio.com then find the Radius Console. You'll like it.

Hey, thank you very much to Andrew Zarian, our producer for the show, and the GFQ Network for distributing This Week in Radio Tech. Please patronize our sponsors, and we'll see you next week on This Week in Radio Tech. Bye-bye, everybody.

Topics: IP Telephony

Subscribe

If you love broadcast audio, you'll love Direct Current! Get it delivered to your inbox weekly!

Posts by Topic

see all