<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

Audio for TV with Gibson Prichard

Posted by Kirk Harnack [TWiRT] on Dec 15, 2014 11:34:00 AM

Find me on:

TWiRT 238How has the move to digital audio changed audio routing and workflow at TV stations? Gibson Prichard should know. He’s the Chief Engineer for Journal Broadcasting’s WTVF-TV - News Channel 5 - the CBS affiliate in Nashville, Tennessee. Self-educated in broadcast engineering, Gibson started at a local FM station and has worked through digital and HD upgrades at two television stations.

 

 

 

Watch the Video!

 

Read the Transcript!

Kirk: This Week in Radio Tech, Episode 238, is brought to you by Lawo and the new crystalCLEAR Virtual Radio Console. crystalCLEAR is the radio console with a multi-touch touchscreen interface; by the new Omnia.7, FM and HD audio processor with Undo technology. Omnia.7 is a big-priced audio processor with the sound and features you love; and by Axia Audio and Livewire soundcard replacement. Get your PC out of the XLR era with Axia Audio over IP.

How has the move to digital audio changed audio routing and workflow at TV stations? Well Gibson Prichard should know. He's the chief engineer for Journal Broadcasting's WTVF TV NewsChannel 5, the CBS affiliate in Nashville. Self-educated in broadcast engineering, Gibson started at a local FM station and has worked through digital and HD upgrades at two TV stations.

Hey, welcome in to This Week in Radio Tech. I'm Kirk Harnack, your host. I'm so glad that you're here with us. This is the show where we talk about everything from this thing, the microphone, to the light bulb at the top of the tower. You know, I've got tone of those here in the office. Sometime I've got to turn the camera around and show it to you. We've got a 300 millimeter code beacon here in the office. We'll check it out another time.

Our show is brought to you by the folks at Lawo and the Lawo crystalCLEAR Console, the virtual mixing desk; also by the folks at Omnia, and the new Omnia.7 Audio Processor. It's amazing, and it's family-priced; and also by the folks at Axia and the Axia IP-Audio soundcard replacement. We'll tell you about that in just a few minutes.

Let's bring in our guest. It's just me and our guest today. That's it. That's all you're going to get. Our guest is Gibson Prichard. Hey, Gibson. Gibson, you are the chief engineer at I guess the number one television station in Nashville, the CBS affiliate, WTVF. I guess do you still call it TV or do you call it DT now?

Gibson: Well, I think the DT went away when we shut the analogs off so most stations are just called their call letters, or their call letters dash TV. And that usually stemmed from if there was a co-owned AM or FM, if the call letters were ever also associated with an AM or FM, whether they are today or not. Sometimes a TV suffix is indicative of an AM or FM with the same call letters, but the DT stuff went away in 2009 when we turned the analogs off and everything was digital.

Kirk: Gotcha.

Gibson: But you still see it pop up from time to time.

Kirk: So the show is called This Week in Radio Tech, and so some people may be asking "Why do you have a television engineer on there? That's just radio with pictures." And Gibson brings to the show experience in both radio and television. I met Gibson, gee, some years ago. Actually, I probably met you, Gibson, when you were already at a TV station. But just prior to meeting you, you had been the chief engineer at one of my favorite stations, Lightning 100, WRLT here in town.

Gibson: Yeah.

Kirk: How did you get into broadcast engineering of any kind?

Gibson: Well that's a very interesting question, because it's not a path that is one that is probably shared by many other people. I started in sales. I started selling radio advertising. I was out of college about 25 years ago or so and was looking for a gig like everybody else is at that time. I applied for a sales job figuring that's what was open, and I had had some sales experience in college, so I got a job in sales.

To put it politely, I was not blowing the doors off the sales numbers, and at that time the station for whom I was working, which was Lightning 100 here in Nashville, was in receivership, owned by a bankruptcy court. They had a couple people leave and I became a traffic and production guy. The chief engineer was a position held by a contract person which is common in radio engineering, of course.

And I sort of taught myself the engineering. We needed things done. We needed CD players hooked up or microphones fixed or things done on remote, and when the engineer was only being paid to keep the station on the air, I was the one who sort of... I picked it up by default. I taught myself engineering. I think I figured most of it out along the way. I had several people mentor me, and I was there for about five years.

I moved the station and built two additional transmitter sites when we purchased two other stations, and I learned quite a bit. I learned computer networking; I learned audio networking; I learned analog audio cabling; AES cabling; RF engineering; and I was blessed with the ability to pick things up rather easily. Then I got a call from the chief engineer at a TV station in Nashville who said "I hear good things about you. Maybe you'd like to come work television?" Honestly I never thought about it.

Kirk: I've got to interrupt you. You're blowing my mind. I had no idea that you didn't have formal education in some kind of electronics, because Gibson ...

Gibson: No. Well, I have a college degree. I have a college degree, but it's in communications. It's in TV.

Kirk: Right. What does that mean?

Gibson: It's in public relations: how to be a disc jockey, how to edit video, how to write commercial copy, how to do PR. Nothing to do with technology.

Kirk: Now you went to Mississippi State. Do you still follow the Bulldogs?

Gibson: Yes I do, Kirk. I do. I still have a cowbell. Yes, my kids are very proud of the cowbell.

Kirk: Okay.

Gibson: I use it to wake them up sometimes.

Kirk: I'm glad we had this conversation. I've known you since almost the moment I moved to Nashville, maybe within a year of my moving here almost 14 years ago, and I'm blown away that you are not classically educated in electronics and broadcast gear and that kind of thing. Which is just like me. I have no formal education in this either. I tell people I got shocked enough times to where I learned where all the electrons like to hang out.

Gibson: Well, sometimes that's what happened to me. I got shocked today just walking down the hall. I drew an arc with some static electricity, and it just made me think for a moment, wait a minute. There's electricity everywhere we go. It's all around us.

Kirk: Now I've got to tell you, it took me... you kind of had a baptism by fire over about a five-year period, you said, that you worked for Lightning 100.

Gibson: Right.

Kirk: And of course when you moved into TV you had a bunch more stuff to learn. That's much quicker than me. It took me a lot longer to learn whatever it is that I know about broadcasting, so that's just awesome. You said that... now this is really interesting, because I always thought I'm a little bit gifted with being aware enough to figure out what puzzle pieces I don't have yet. You know, I'm able to figure out what it is that I don't know yet, and I could go after those things. How have you thought about that learning quality for yourself?

Gibson: Well, if you think about it like a puzzle, you always start with the corners. You know where the edges are. You can look in the box and if they're only cut on one side you know that's probably a corner piece and you start there. You know, honestly I think I just have a sense of intuition so I can look at something whether it's an audio processor or an audio console or a transmitter and I can look at it and say "All right, I know it's got to end up out one end. How does it get there?" And somehow I can read the schematics and read the owner's manual and I'm able to pick things up as I go.

I'm sure I've made a lot of boneheaded decisions along the way and probably did some really dumb things and some of my former employers might say "Well yeah, he wasn't the smartest kid when he first started here." But honestly, I've just been able to pick things up as I go. I'm pretty good at reading owner's manuals, pretty good at finding things on the Internet, asking questions. I know kind of what questions to ask. People ask me how I figure it out. Honestly, it's just something I do. It's kind of like walking or chewing gum. I just figured it out and don't really know how to explain it.

Kirk: We're talking to Gibson Prichard on This Week in Radio Tech, and I can tell this is going to be a really fascinating hour or the 45 or 50 minutes we have left. We're talking about broadcast engineering. You know, not necessarily all the hands-on specifics, although there may be a few things that we get to in the hour. But this is going to be a really fascinating hour, because Gibson, it turns out I didn't know this, is very much self-educated. I know a lot of other engineers are as well, so we'll get into that.

Our show is brought to you in part by the folks at Lawo. Lawo is a console company called L-A-W-O, so go to lawo.com. Check out this great console. It's an idea that I've always thought is terrific and it's about time it's come along, and that is a touchscreen console. It's the Lawo crystalCLEAR Virtual Radio Mixing Console. And by virtual, it means that the surface that you touch - you see that picture right there? The surface that you touch is a multi-touch computer monitor. You can have up to ten touches on there at once.

It's actually a good HP computer with this ten-touch multi-touch screen. You can be touching several faders at once plus buttons, move things up and down, move things around. It reacts really well. It's a fantastic touchscreen. And the software has been written to look just like a Lawo Audio Console, so stuff is where you'd expect it to be.

What's cool about this is buttons behave in context to what you're doing. You're not limited by a piece of hardware anymore where a button has two wires going to it and only does this thing and it only lights up this way. Because it's written in software, the buttons can do anything you need them to do, and so they're very contextual. If you're working with a microphone channel, a microphone fader, and you touch an options button for that channel, you get microphone-based options right there.

You can put up to eight things through the console at once, at the same time. There's eight faders on the control screen. But the engine itself that you plug things into will handle up to I believe 24 - yeah, 24 sources altogether: mic, analog line, and AES inputs, plus you can get inputs from an AoIP network, from either Ravenna or AES67. Ravenna is AES67-compliant. So if you have other items in your facility that are either Ravenna or AES67 AoIP then you can bring them in as well. And they have eight up on the console at the same time. Of course you can switch things that are on the console instantly with no interruption in your audio whatsoever.

If you mess up the console, "Whoops, I put all ten fingers on at once and I'm not sure what I touched," there's a panic button. It clears any changes that you've made and returns you back to the console's default position for that shift.

It's got three stereo mixing groups, program one, program two, and a record bus. Very intuitive GUI. You can really use this thing quickly. Advanced DSP for microphones, so there's mic processing built-in and four external sources as well. Two low-noise mic pre-amps are built in, and of course you can use external pre-amps if you need more than two mics. And there's power supply redundancy that's built-in as well plus GPIO for on-air lights, for muting speakers, and muting other things around the studio that you might have to.

So if you're interested in this kind of technology, and I would hope you would be, this is really cool, check it out. Go to lawo.com, L-A-W-O. So if I say it the American way, lawo.com, and look for the crystalCLEAR... it's under the products and radio consoles area. Then look for crystalCLEAR. I sure thank Lawo for their support of This Week in Radio Tech and for really acting on this great idea in audio consoles. All right, Gibson Prichard is our gest. He's the chief engineer at Channel 5 here in Nashville, but he started out in radio, a disc jockey, selling ...

Gibson: I don't think I was a disc jockey.

Kirk: Okay.

Gibson: I probably filled in for a weekend shift once or twice. I was a disc jockey in college, so I guess that would be accurate. Yeah, I was a college disc jockey.

Kirk: So making the move, Gibson, from radio to television, I guess I met you at FOX 17, the FOX affiliate here in Nashville. So how did that work? What was your confidence level in transitioning to an environment where audio is kind of almost an afterthought sometimes?

Gibson: Well, it was interesting because as I said earlier, I got a call from the chief engineer at the FOX affiliate who said "Somebody told me about you and said you might be a good computer engineer, somebody I might want to have on-staff to work in IT And engineering because you have RF background." I said "To be honest, I never really thought about working in television."

My first love was radio. I went to a job interview. We talked about the position and talked about what would be involved and what would be expected. After I took the job, what occurred to me was well, suddenly I had a budget to work with. I had been working at a very small radio station where I didn't have very much of a budget, if at all. I suddenly had a budget. I had things I could buy. I didn't have to go beg and borrow for things I needed.

And I picked up the video piece. I mean, at the time it was all analog. I had to learn about analog video, but I knew how to put a BNC connector on. I didn't know all the Belden part numbers for the different video cables, but of course I know them now. I understood some of the basic principles. Having gotten a college degree in TV and radio broadcasting, I knew about how to edit and how to make a tape edit which is what we were doing at the time.

You know, I understood the basics. I understood high-powered amplification. I'd been to a radio transmitter school, so at least I understood the ground-to-grid transmitter design, although please don't ask me to come neutralize a tube or work on your CCA transmitter or your Energy-Onix transmitter. I wouldn't be the guy you'd want to have since I haven't touched one in 15 or 20 years.

Kirk: A long time.

Gibson: But you know, once you understand some of the basics, whether it's a Klystron or IOT or whether it's a power tetrode... I mean once you understand the principles, and I was just I guess somehow lucky in that I was able to make the leap from one to the other. I did a lot of audio work. Like you might expect, I was connecting audio cabling, doing a lot of audio patch-panel wiring.

What I was struck with when I moved to television was I had options. I had a routing switcher. I had the ability to send audio from one place to the other without a direct cable. I mean it was a direct cable, but I didn't have to make a dedicated cable. I could use the facility's routing switcher, so that was a real blessing.

Kirk: Yeah. In small-market radio, even medium-market radio, we don't always have routing switches. In my 20-year career as a contract engineer, I never once worked for a broadcast facility that had a whole-house routing switcher.

Gibson: I did some work for a local... or it was the Tennessee Radio Network here that did have an analog audio routing switcher, so I was fortunate. I got to understand that and got to wire it. You know, I had some people along the way who taught me how to do good cabling. It was all 66 style telco blocks and I graduated to other forms of interconnect.

But you know, I just sort of picked it up along the way. As I said earlier, there's probably some stuff I'd sort of turn my head at and look at if I did now. If I were to go back now and look at some of my old work, I might not be so proud of it. But you pick it up as you go. You certainly recognize the bad. I don't know that I recognize the good so easily, but I recognize the bad work. You try to do your best.

Kirk: In a few minutes, I want to talk about your experience in the television world with the transition from analog television to digital television and the upgrades and studios, and of course audio often goes along with that.

Gibson: Sure.

Kirk: I want to get there. But first, one of the things you and I talked about in a number of SBE meetings here in Nashville was when Nextel was involved in making some changes in the ENG spectrum, the two gigahertz spectrum.

Gibson: Right.

Kirk: And Nashville was one of the markets, I think the smallest market, where stations got provided with some new gear. Right?

Gibson: Actually, no. It was actually every ENG market across the country. So for those of you that don't know what ENG means, Electronic News Gathering. In television, generally there are a couple of different bands, frequency bands. But generally speaking, there is a little chunk of spectrum around two gigahertz, from 1990 to 2110 megahertz or 1.99 gig to 2.11 gig. That's what it was, anyway.

And Sprint Nextel, of course it was just Nextel at the time - it was before Sprint bought them - they said that's adjacent to all the PCS channels right below it. So the transmitters, the handheld devices, the cell phones that we're already building would work just beautifully in that spectrum. So "Dear Mr. FCC, couldn't we have some of that spectrum?" And they said "No, you can't just have it. You can buy it." They said "Well, if we buy it ..." But the problem was you had to get the incumbent users out. You had to vacate the spectrum of all the TV stations who were using it for news gathering, and some for STLs, although most STLs are out of that band by now.

So Sprint Nextel was after five megahertz of spectrum, but they had to clear 20 megahertz because there were 15 other chunks the FCC was going to sell. And so Sprint undertook this enormous plan of paying for all the TV stations to change out their hardware and change out all their analog FM-modulated news gathering microwave for new digital gear.

It was an incredible boon for the people who made the equipment and for the guys who were hanging transmission line and antennas. But like a lot of things, when it happens out of a mandate, it's a sort of feast-and-famine situation. For about a year or two years it was just going gangbusters, then once everybody got converted it all dried up.

But if you think just dollar values, with each transmitter being say $30,000 that goes in a van, and the reason they're so much is because they had MPEG encoding built-in. So if a station's got four or five vans and each transmitter is 30 grand, that's $150,000 just in transmitters for the station. Then you've got the receivers and then you've got control systems. Basically, Sprint Nextel was replacing everything, because that stuff as you might imagine had lasted for ages. TV stations had this stuff around from the '60s. At the time, it was 25 or 30 years old. It was still working, and Sprint had to replace it with all-new, all-digital gear. It was quite an amazing transition to be honest.

And so what happened is all the TV stations had our band shrunk by what was that? I guess about 35 megahertz. We lopped off this big chunk at the bottom of the band and then kept the same number of channels by moving from 17 megahertz channels to 12 megahertz channels.

Kirk: Oh, okay.

Gibson: So each channel got five megahertz narrower, which 17 megahertz is what you needed to pass analog FM video, but you could do that just fine in 12 actually or six or eight megahertz is what you need depending upon the bitrates you do. So that was what happened, because the TV... you couldn't just reduce the number of channels except in the very, very, very small markets like markets number 150 and above where there was just one or two TV stations doing news. In almost every market, there's at least three, maybe five or six stations doing news, each with a varying number of trucks and operators.

But it was a big ordeal. I mean I was involved in my station's work, and at some level for the Nashville market. I am the frequency coordinator for the market through the SBE. But it was an amazing task. I'd kind of forgotten about it. That was part of the analog transition for television when we turned off analog transmitters, so it seems like eons ago, but it really was about five or six years back.

Kirk: So with all this new ENG gear, was it capable of high-def transition?

Gibson: No.

Kirk: Ooh.

Gibson: Well, not directly. Well, yes in a way. It all was standard definition, so all the manufacturers were including standard-def modulators. Basically MPEG2 modulation or MPEG2 encoding, COFDM modulation. So you could put analog video in or you could put standard-def digital video in and it would be passed through. If you wanted to do HD, you needed to buy some other form of encoder and then connect it through one of the external inputs, usually called the ASI input where you can take a digital data stream from an external encoder.

So you were essentially using the RF portion of the radio but not the video components. You were not using the pieces built-in. It's sort of like using your computer but bypassing the soundcard and doing something. You know, putting audio in through a livewire interface or something. You're not using the built-in soundcard because you want to improve the quality. So that's basically it.

Kirk: Ah, okay. Okay. All right. Well cool. I remember you explaining to a lot of folks what's going on and be in touch with your Nextel rep and all this. It seemed really complicated. Hey, we're talking to Gibson Prichard. He's the chief engineer for a TV station on This Week in Radio Tech. He's the chief engineer for WTVF, the CBS affiliate here in Nashville, Tennessee. I'm always pleasured to have a conversation with Gibson, usually after an SBE meeting here in Nashville when we both get to attend. I'm always finding out what's new in the market and what's new in technology.

We've got a couple important subjects to still get to. Gibson, besides the transition over to HD and how that affected audio in your TV stations. Also, I want to end up talking about the Internet, the IT router subject that you and I talked about a few weeks ago after the SBE meeting. That's pretty fascinating too. I know lots of people who watch the show are interested in such things, so we'll get to that in a minute.

Hey, our show is brought to you in part by the folks at Omnia Audio, and I've got a brand new processor to tell you about. I don't think it's shipping yet, but you need to find out about it. For a lot of radio stations, the Omnia.11 and even the Omnia.9 are just a huge financial step. These are $10,000 plus audio processors. Well, our terrific technology partner Leif Claesson, and you guys know - if you're familiar with Stereo Tool or any of Leif's work, you know how good it is. Well Leif has now - Leif and Omnia together have come out with a new audio processor called the Omnia.7.

I'll put a link to it in the show notes. There's a picture of it right there. It's a smaller version of the Omnia.9 with many of the features that are most desirable still in it. And let's just say it's very popularly priced. You're going to want to get with your Omnia dealer to find out about this processor.

Now it doesn't do all the multiple paths of processing that the Omnia.9 does, but what it is for, it's for FM and it's for HD. It does have that exclusive Undo technology that Leif Claesson is so famous for. You know, records are - music is mastered nowadays with such clipping, and it's just ugh. I guess it may be okay for your iPod or for just listening to it, but my goodness, it doesn't code well in the bit streams, in bitrate-reduced streams, and it's no good for FM because with FM we're just going to process it more and clip it some more in very intelligent ways. But this Undo technology is just amazing. It can take the worst-sounding recording that's just clipped to heck, you can look at it right off the CD and it's just clipped as can be, and Undo rebuilds very intelligently the peaks that got clipped off.

And Leif has some demos of this on YouTube. Of course the best thing is to see Leif demonstrate this in person. You can really hear it well. So the Omnia.7, just like the Omnia.9, has this Undo technology. It's a two-step process that restores peaks and dynamic range and removes distortion from source material that's been damaged by over compression and clipping in the mastering process typically.

It has a cycle acoustically-controlled distortion masking clipper. Boy, that's a big mouthful. Let's just say it sounds terrific. It has Omnia Toolbox built-in which includes a digital oscilloscope, an FFT spectrum analyzer, and a real-time analyzer to help you adjust your processing and to see what you're hearing.

It also has - this is cool - speaker calibration. I haven't tried this out myself, but as soon as I get my hands on an Omnia.7 here at my lab, I'm going to get this speaker calibration going so I can make sure my speakers are EQed, flat, and everything sounds good.

It has a dry voice detector that helps it with adjusting the processing moment-by-moment when there is dry voice being put through it. A remote client that is a joy to use. It's absolutely terrific, a very real-time remote control client. So check it out. This Omnia processor, the Omnia.7, is just amazing. I did get to look at one and play with one for a few minutes at the Telos headquarters in Cleveland back a couple weeks ago and I'm just very excited about it. If you go to the website omniaaudio.com/Omnia-7, so it's omniaaudio.com, the usual Omnia Audio website, and slash Omnia dash or hyphen seven. Check this thing out. So if you couldn't afford Leif's high-end processor before or an Omnia.11, this can do the job for you. I think you're going to like it. Thanks to Omnia Audio for being a sponsor of This Week in Radio Tech.

It is episode number 238. We are talking with Gibson Prichard who is a television engineer, and we're going to get around to talking about audio and some IT issues and subjects that can affect you at your broadcast facility. So by the way, coming up on another show in the very near future... it's already slipped my mind. We've got a guest next week that I'm very excited about. Sorry, I'm having a senior moment.

Gibson: It's Ol' What's-His-Name?

Kirk: It's Ol' What's-His-Name, yes.

Gibson: He's great! I can't wait for that, Kirk. I'm so excited.

Kirk: Oh by the way, Gibson, thanks to your suggestion my booking agent is reaching out to your suggested guest.

Gibson: Oh, okay. Very well.

Kirk: The superman of broadcast engineering, and I didn't even know it. So, let's get around to talking about transition from analog television to digital. And along with that, most people think it was a transition to HD, but really it was a transition to digital and HD came along for the ride. Is that kind of...

Gibson: That's right. Yeah, that's right. The FCC said you have to change your modulation from analog NTSC to this 8VSB method that we're currently using.

Kirk: Yeah.

Gibson: And with it, there were some specifics about it. You had to be compliant so that any tuner could receive it, or any tuner that could receive it could decode it. So that means MPEG2 video, Dolby AC3 audio. You had to include certain parts of metadata. You had to be able to send things like guide information and clock data. There's a relatively thorough list of what was included. But one of the things looking back in hindsight, since what is known as the ATSC or Advanced Television Standards Committee, since that system was first proposed and of course adopted by the commission, that's been... well, since 1992, '95, somewhere in there?

If you go back and think about what was going on in radio technology 20 years ago, there's a lot that's changed. So if you think about MPEG2 video being the standard, that's not keeping up with what's current. I mean most people are using MPEG4. That's what you get on Blu-ray players. That's what you're getting streaming over the Internet from Netflix and Hulu. And then there's of course H.265 video which is coming, which is actually being used by Netflix and Hulu and others.

Kirk: Ah.

Gibson: So what we have is this standard that provides great pictures, and if you transmit high-definition you can transmit great pictures, either 720p or 1080i video. Two major networks chose 1080i, two chose 720p. To most people, they can't tell the difference. There are some reasons they'd choose one over the other. But that transition for the TV stations required us to go out and purchase new transmitters and purchase new systems to encode the video to make the MPEG2 video and make the Dolby AC3 audio.

And then in the case of my TV station, we actually had to change channels. We were transmitting on channel 5 in analog. Our what was known as pre-transition channel for our digital signal was on channel 56, which was outside of the core. That was part of the spectrum that the FCC was going to auction off to all the wireless telecom people, so we could not stay on 56. We had to go on another channel.

There was a period of time that we looked at it and we went back to channel 5 as our primary signal, which is a low-band VHF channel. It worked great for analog; it covered a long distance. But because of its susceptibility to impulse noise, noise from lights and motors and power lines and things, it really was not all that well-suited for digital, particularly people with indoor antennas because of both the noise and the long wavelength of a channel 5 signal.

Kirk: Yeah.

Gibson: So if you didn't have a rooftop aerial, sometimes you had trouble getting our station even though we were transmitting plenty of power. So we put a translator on. I put a translator on in UHF, then over a period of time made an application to change our primary signal back to UHF. Of course we couldn't be on 56. We chose 25 as the channel number. And so that's where we are. We've had to change channels four different times. We've had four different channels. That meant replacing transmission line, replacing antennas, retuning transmitters, changing the mask filter.

Those of you with HD radio, you know what your mask filters are. People in television, of course, it's the filter that takes the signal from your transmitter and makes sure that there's no emissions outside of the six megahertz channel that you're supposed to occupy. So they're expensive sharp-slope filters to cut the edges off the channels so there's no spurious emissions coming out of the transmitter that would affect other people.

Kirk: This kind of filter you're talking about, is this done at high-level RF? This filter?

Gibson: Yes, it's done at high-level RF. So in the case of my station, our TPO is say 56 kilowatts or so. We put about 57 into the mask filter. We lose a little bit through the filter. But it is a large assembly that is about 20 feet square probably, maybe 17x12. Probably 17x12. It's not 20 feet tall. It takes up a huge chunk of a room. It includes a combiner to combine two transmitters through a magic T, a switchless combiner, and then through a big mask filter that performs the magic and chops off the things that are outside the six megahertz channel. So it's channel-specific, yeah.

Kirk: So let's move the conversation a bit to audio.

Gibson: Sure.

Kirk: How did your audio world change in the digital transition or your HD transition?

Gibson: Well, we had installed, several years prior to turning off the analog transmitter, we installed a standard-def digital video plant with an AES audio plant. There was an AES router and an SDI video router. And while the SDI video, SDI means Serial Digital Interface, but while the SDI video can carry the audio along with the video in what's called embedded audio, we built a facility that was all discrete audio which meant that you had to run two cables to everything. You had a video cable and an audio cable. Even if the video cable had audio on it, because it was embedded, we weren't using that audio. We were using a separate cable. So we made a separate color code, so you have two cables going to every piece of equipment. Then if it was a tape deck, you had two coming back out of it as well.

When we migrated to HD, of course some time had transpired from the time we converted to just standard digital. We made the decision to do all-embedded audio, because other than just a few places in the TV station, there's really no reason to separate the audio from the video. Everywhere you separate it, there's a possibility of creating a lip sync error where the audio is not matching what people are saying. So it also makes it simpler because it's one cable going to a piece of equipment and one coming back. It's not two. It's half as much routing equipment, half as many distribution amplifiers. It's just 50% easier.

It also can carry more audio, because an AES audio standard, the AES3 standard that we live under, carries one audio pair on a cable. Whether it's balanced or unbalanced, you get one left and one right. You get a pair of audio, whereas an SDI signal can carry up to 16 audio channels, eight AES pairs. So it's much simpler.

So if you needed to carry say television, most things we do have surround sound so that's left and right, center low frequency, then two surround channels in the rear. So a total of six channels, or 5.1 as you hear it described. Well that's three AES pairs, so that easily fits in SDI. That's one cable. That's your video and three audios. It gives you space for other audio or other ancillary data.

It's quite elegant, to be fair. It's a very elegant and simple system. People might say, "Well, aren't there reasons to separate your audio?" Well, sure, of course there are. If you have an audio mixer, an audio console where you're doing newscasts and you've got microphones and sources coming into an audio console, a traditional board like a radio board? Of course, that's all discrete audio. There's some analog; there's some AES there.

You know, so we handle it like a traditional console would and then we put it back in the video. We go back into an embedder. In this case, ours is a frame synchronizer that ties the video back to our house time code standard or our house time standard. It puts audio back on it as well. And that's where we can adjust the lip sync issue. So if we have a lip sync issue, we can adjust it there.

Kirk: Tell me about lip sync. A lot of us as consumers, as radio guys that aren't necessarily in television, we've seen this. We've experienced it. Where does the problem normally ...

Gibson: We all did. It was called the Godzilla movie.

Kirk: [Laughs]

Gibson: Yeah, it was going on 45 or 50 years ago. It's called Godzilla.

Kirk: Assaulting that audio, Gibson.

Gibson: That's right.

Kirk: So at what point is it most typical that lip sync errors creep in, and what is the point that it needs to be right? Or where do you correct it?

Gibson: Well, the point that it's most likely to creep in is anywhere it's separate, anywhere the audio is separate or takes a different path. So if you process the audio, say you've got a compressor or limiter or you've got something that maybe you're doing some delay in the case of the frame synchronizer, you might add a frame of video delay, so 33 milliseconds of video delay, to make your video match your house timing standard. Then you need to also account for that 33 milliseconds in the audio path.

Kirk: Ah, okay.

Gibson: Or if you go through a console. Every digital console has some delay, and it's usually a published spec and you can look at it and you can work from the published spec. Generally what we do is we just watch it. You know, we'd sit there, get somebody to sit in front of the camera with a microphone, and just watch it. I'm talking about the things that we control like our newscast, our news audio.

Kirk: Yeah.

Gibson: So we just try to make sure that it's closer and that it looks right. But the other place that it can get messed up is when it comes from a network because often it comes in on satellite. It comes in, say in the case of CBS, CBS delivers audio with the Dolby E standard. Are you familiar with Dolby E? It's an AES-compliant data stream, but it supports up to eight channels of audio.

Kirk: Okay.

Gibson: It's a compressed data stream that meets at 1411 kilobits per second maximum, so it's compared to say AAC for HD radio. It's a lightly-compressed signal, but it is still compressed and puts eight channels into one AES signal. Well, we've got to decode that. We've got to break it back into the AES, and then we've got to embed that back in the video. So that's two separate things. There's the Dolby decoder and then there's a frame sync to put it back with the video. All of those are places where you can have timing errors. Those are places where you can create that Godzilla moment where things just don't line up.

Kirk: Is it more typical that in routing around a plant that video gets delayed? Or is it more likely that audio gets delayed? Which one do you usually have to make up for?

Gibson: Usually you have to make up for video delay. Usually you have to make up that the video is behind the audio. That's the most common. Something is processing the video like you're running through a video switcher to put in the logo. Just like you're putting the logo at the bottom of this screen, well there's some processing time there, so you have to account for that. Usually the video has more bits to it so it's more computationally-difficult, and so that's usually where the delay comes up.

Kirk: And as you mentioned earlier with the frame synchronizer, one frame of video is you said I think 33 milliseconds, right?

Gibson: Right, yeah.

Kirk: And that's a significant amount of audio delay. Not for an IP codec, but for a live audio mixing system, whether it be a MADI or AES-based, or whether it be IP audio-based, that's a lot of time for audio.

Gibson: Yeah, it is.

Kirk: And exacerbating that is the fact that our brains have less problems seeing the sight first and hearing the audio a split second later, because ...

Gibson: If the audio trails, you can make it up. Something about your mind can perceive it.

Kirk: The brain is okay with that.

Gibson: The brain's okay with it. It can say "Yeah, that's not quite right, but I can live with that." But if the audio comes first, it looks like a chicken talking. It just looks funny.

Kirk: It'd be like hearing the thunder and then seeing the lightning. Or it would be like hearing the crack of the bat from the stands, then seeing the batter hit the ball.

Gibson: That's right.

Kirk: It wouldn't be right at all. The brain goes ahh!

Gibson: Right, it's not supposed to be that way. That's right.

Kirk: Yeah, that's fascinating. So talking about audio, and I mentioned IP audio there, I'm curious about video over IP. I take it from everything you've described, you're using SDI. That's not IP. Those are...

Gibson: It's a serial stream, that's right.

Kirk: Serial stream systems. So your switchers are not inexpensive. They're purpose-built. That's the phrase I'm looking for. They're purpose-built.

Gibson: That's a great term. Just like an audio processor, it's purpose-built.

Kirk: Yeah. So tell me about the world of television though internally in a TV station transitioning to IP video or IP transport and routing.

Gibson: Well, IP comes into a traditional TV station probably in two ways in video. The first and foremost for stations like mine that are in the news gathering business are cellular backpack video. So you're using IP codecs. You know, you're compressing the video out in the field with a small computer mounted on someone's back or put on a box or somehow mounted near a camera that takes the incoming video and compresses it to fit on several different IP networks. Usually those are cell phone networks like cellular data cards. Maybe five or six or seven or eight cards all carrying some data back to the station.

So it packetizes the video, breaks it into different chunks, seven different chunks, and sends it over seven different cell cards where they're received back at the station on a common Internet connection and then put back together. It's kind of like Humpty Dumpty. It puts it all back together again and then it spits out the video.

Of course it's bit-reduced, and so it's not the same quality that it started with at the camera out at the field, but it's passable and that's IP video. But that's really an IP bridge, if you will. That's a purpose-built appliance to bridge a problem, which is to get video from a remote network when you have very limited bandwidth.

Kirk: Yeah. The cell carriers don't carry SDI very well do they?

Gibson: Not at all. And if you don't have a satellite truck and you don't want to pay for some sort of high-bandwidth link, that's how you do it. Now cable networks, most cable networks are all IP internally and so they route everything. Everything is packetized and put on a gigabit network. Almost every bit of video inside a television cable company, and this extends to satellite companies like DIRECTV and Dish Network, they are all IP-based. So it's serial. It's continual data. It's mostly UDP packets, but it is IP video.

So we have a little bit of it. We have some places, usually at the edge of our building, at the edge of our signals, where we encode our video. We take the SDI and we compress it to send to our transmitter through our MPEG encoder. Well that encoder also has an IP out. Currently it's not plugged into anything. It could be. I could pass it over a common Ethernet network if I had a private Ethernet network. I can make a 19.39 megabit per second IP stream. Or actually, I do make it. It's just not going anywhere. I could send that to a cable provider and they could put that directly on their network.

Cable networks love IP. Cable companies love it because they can route it around. They can take it from one head into another, and it's very efficient when you're talking about things that are 10 megabits to 50 megabits. They can put a lot of those on a gigabit Ethernet pipe. So it's really about the ease of distribution I think would be the easiest way of putting it.

IP video routing in a facility is still quite new. There are some companies building equipment for it. I believe ESPN in Bristol, Connecticut has a big IP router that they have purchased. I don't know what status it's in or whether it's in service yet. But the big TV equipment vendors are making IP routers, but really they also have to make the backplanes. It's not just buying equipment from Cisco or Foundry or somebody; it's building a backplane design to switch at least three gigabits per second.

In serial packets, probably more like 12 gigabits. Because if you're going to build a router, if you're going to build a facility, you want something that's going to last into the future. So HD video is 1.5 gigabits per second for 1080i or 720p. To do 1080p, you double it, three gigs a second. To do 4K, you're at 12 gigs a second. So it takes an inordinately large amount of bandwidth, and so your IP routers, your IP video routers, have to be sort of pushing the edge of what the silicon can do. So that's new technology.

Kirk: Gotcha. Wow. Wow, okay. Do you see TV stations that have in the last few years built out with a new SDI plan? And as we mentioned, a purpose-built SDI router. Do you see any reason for them to transition to an IP routing system internally? Because their gear is pretty new at this point.

Gibson: Not at this point. You know, my SDI router is 10 or 12 years old and it still functions very well. It's still serviceable; I can get parts for it. But at some point it's a conversation that we'll have to have. What is the next router, and do we move to IP? Is the answer to that question that you move to IP simply because it's new? Because you've got to do something, and why not? Kind of like changing the color of your coat or changing your tie, you want to change just because change is out there? I don't know.

I've not yet looked at the economics of it to know whether it makes sense for a TV station, because in terms of the matrix size, I think where it really shines is where you have an inordinately large matrix where you're trying to switch over 400 or 500 sources both in and out. My plant is not that large. I'm about 256 or so and 256 out. So while it's quite large compared to say what an average radio station might have in terms of number of sources, it's still small compared to what a network like ESPN or NBC or CBS might have in their facilities.

Kirk: Ah, that makes sense. Good deal. We're going to take a quick break here and hear from the folks at Axia because we're on the subject of IP and they're a sponsor. Then we'll get back with a couple more things to talk about before we close out the show. We're talking with Gibson Prichard on This Week in Radio Tech, and he's the chief engineer for Channel 5 WTVF TV here in Nashville.

Hey, Axia is a sponsor, and I've got to tell you one of the things that I really love about audio-over-IP and the pioneering work that Axia did on this is soundcard replacement. Hey, at my own radio stations I've got six, maybe seven PCs, some running Linux and some running Windows that have no soundcard and yet I have absolutely bit-for-bit perfect real-time audio transfer in and out of those PCs. The applications that are running, they don't know the difference. They think it's a soundcard, either in the Linux operating system, part of the ALSA mixer, or part of the Windows operating system, because the Axia IP Audio Driver looks to the operating system like a soundcard.

And so you can make an Axia driver the default soundcard, so applications that don't otherwise choose a soundcard device just use it. Or if you have maybe a multitrack editor or some other application which lets you choose what soundcard to use, well you just choose the Axia IP Audio Driver.

And this means you get absolutely - let me say it again - bit-for-bit perfect audio transfer in and out of the PC. You don't lose anything, not even let's see, it's 24 bits so that's like 132 or 136 dB down. I mean it's two levels of quantum. It's enormously better than CD audio. It's good stuff. In fact, at station-after-station, when either I've done an installation of an Axia AoIP system or one of our other staff members has done that, they report "Wow, we can't believe how much better the station sounds."

I mean there's a big national radio show that's produced just down the road from me here in Nashville with I want to say six million listeners a week and we put it on the air, the Axia system, this was years ago, like seven years ago. And bam, I mean instant. It was obvious, even over the satellite, over the affiliates. It was obvious that the audio quality was better, and they didn't have bad equipment beforehand.

I just heard from one of our engineers that just did some work here in Tennessee, a good-sized market here in Tennessee, that said we replaced a control room with all Axia IP Audio gear and, bam, you could hear the difference on the air. You wouldn't think so. It surprised me. Well, the Axia IP Audio soundcard is a big part of that because that soundcard... a traditional soundcard in a PC, man, I mean quit fussing really with 3.5 millimeter plugs and unbalanced stuff or even XLR connectors and balance stuff. All that expense, the drivers that go out of date when Windows makes you upgrade to something new like Windows 7 or 8 or Windows 10 coming along? None of that changes. All you do is use another network card. Stick a network card in there that does gigabit and install the Axia IP Audio Driver and bam, you have got perfect audio in and out of that PC.

If you're just doing a soundcard replacement, if you're putting PCs on a network to play audio for maybe automation work, for a production machine, for a logging machine? Hey, I've got a logger at my stations in both American Samoa and in Greenville, Mississippi, and the PC just runs. It's got a second NIC on it plugged into the Axia network and we have access to 24 different stereo channels on the network. If we want to log what's happening with one of our crazy satellite receivers 24/7, we can do that just by typing a few things into the logger. In other words, no additional wiring. My wiring for the logger consists of one Ethernet cable from the Axia switch going into that PC.

Gibson: Shazam.

Kirk: Shazam. It's been the biggest boon to wiring to not have to futz, and that's the nice word, with audio cards. I can't tell you. So go to Axia Audio. It's one of the big reasons that you should go to an AoIP system. And Axia has been making IP audio drivers longer than anybody else. They came out with those things I guess 10 or 11 years ago and they're just terrific. I can't say enough about it except now I'm running out of time.

Just one of the benefits of going to AoIP, especially Axia AoIP... oh, I should mention the IP audio drivers also handle virtual GPIO. And that means that contact closures don't require additional wiring. It's the same piece of wire. It's the same RJ45, the same network, and you get contact closures in and out of your automation system, your logging system. They're just there and they just work. Really amazing stuff. Check it out at axiaaudio.com, and you'll just be so delighted that you got into that world of AoIP with Axia because it does work so well.

All right, we're just about to wrap up episode 238 of This Week in Radio Tech. I'm Kirk Harnack. I'm talking with Gibson Prichard. He's the chief engineer at Channel 5 WTVF here in Nashville, Tennessee. Self-taught in engineering, he started in radio and is now a chief engineer at probably the number one biggest station in Nashville. So Gibson, what was it? We were going to talk about a couple more things, communications law blog, and what was the other subject? Oh yeah, in the IT world.

Gibson: Yeah.

Kirk: We were talking about routers, and years ago I had built up a Linux-based - it was called Coyote Router.

Gibson: I remember that. I think I played with that, yeah.

Kirk: Yeah. As far as I know, they stopped supporting it. No more updates were available. We used a Coyote router that I built out of an old beige box PC at our stations in Samoa for years and it was so easy to deal with. You told me about another one though that you really like. Tell me about what that was.

Gibson: Yeah. I use a system called PF Sense, like Packet Filter Sense, S-E-N-S-E. It's not based in Linux per se. It's based in a Unix called BSD Unix which is very similar to Linux in that it's free. PF Sense is free. There is commercial support available. But what I like about PF Sense is it takes what is or what used to be rather arcane methodologies to a router and firewall where you have to learn everything at the command line and you can do all this stuff.

If you want an enterprise-grade router that doesn't have a big networking name's company on it like Cisco or Foundry Networks or something you had to pay big dollars for that. What PF Sense does is put it in a GUI, all accessible through a webpage, and it's quite fantastic. I used it for a long time here for NewsChannel 5 here in Nashville. I use it at my house and I'm able to do great things with it.

It has all sorts of features. You want to do filtering? You want to take Suzie the receptionist and say that she can't go to her eBay site? Well you can block that. You can block it per-user; you can block it for everybody; you can do logging; you can tie it to Active Directory, and it's all free. So you can take the people's username, take the name that they login to the computer, authenticate them, give them Internet rights specific to their username. You can log it. It makes a great VPN endpoint.

So if you want to be able to access your building remotely or access your site remotely, it's great. I can't say enough good things about it. There's commercial support available, but you don't have to pay for it. If you don't want to pay for it, you can do it yourself. There's a lot of tutorials online. You can build it from an old beige box. You can take any machine that's basically a Pentium 2 or faster. You know, if you need to pass more than 50 or 60 megabits of traffic, something better than a cable modem, you need something like a Pentium 4.

Kirk: Yeah.

Gibson: But you can buy appliances these days. You can build an appliance. Most people have cast-off machines that used to run XP that they've replaced for some reason. If you've got a Pentium 4 processor in it and 512 megs of RAM, I mean it only needs like 32 megs of RAM maybe to operate. Some of that matters with how much encryption you're doing. If you do a big VPN system, you'll need some processor just to handle the encryption and decryption process, but it's really simple. It's a simple installation. You can back the settings up to an XML file. They're human-readable. You can open them in a text editor. You can copy it. You can make a system. You actually can have two of them in a failover with a third NIC that ties the two together for a heartbeat so if one should fail the other one comes online.

Everything your corporate IT department says you need to pay $10,000 for for a commercial firewall you can do with PF Sense. So it's as robust as you want to make it. Dual power supplies? Well, buy a PC with two power supplies.

Kirk: Sure, yeah.

Gibson: Get two PCs and tie them together.

Kirk: I've long been a fan of the MicroTik routers, the RouterBOARD OS, but programming them can sometimes be a little cryptic. I think you and I got started talking about this because I told you remember the RouterBOARD OS, it's a little difficult to do packet prioritization based on the content of the packet. There are a number of criteria you can use.

I may be wrong on this, but I haven't yet found how you determine "Oh, it's an RTP, a Real-Time Protocol packet, maybe for a VoIP phone and I want to prioritize that kind of packet on an outgoing basis." And I haven't been able to do that in a MicroTik router, and many other routers let you do that.

Gibson: That's true.

Kirk: I love the idea of building up my own router with, like you said, a cast-off PC. But like you also said, you can buy appliances that are already preconfigured with PF Sense if you want to.

Gibson: Yeah, you can find them on Amazon or eBay. PF Sense has a store where you can buy them. They're reasonably priced around $200. You know, it's more expensive than one of the common routers that you might buy at Wal-Mart or any sort of other store, but it does a lot more.

So you get what you pay for. It's certainly less expensive than a commercial firewall and does about what a commercial firewall will do, sometimes more. So it's certainly something you should look at if you are responsible for your company or your TV or radio station's connection. It's a great way to protect you from the things on the Internet as well as to do all kinds of things you never dreamed possible.

Kirk: I want to close out on a note that goes back to what you said earlier about being self-taught in broadcast electronics and broadcast systems. I think probably the way that I've learned, and I wouldn't pretend that I know everything about it but I know a lot about IP, about packet handling and routing and filtering and that kind of thing, is by doing it myself.

Gibson: Sure.

Kirk: And I've always had this crazy interest in routers, or at least enough of an interest in routers to get what I want to do done successfully. And I'm a little surprised sometimes at how much misinformation that there is out there, or people get used to using one router and they think that's how it's done when in fact there may be lots of other ways and maybe better ways to get something done. Well, if you will just take a cast-off PC and build up your own router using Coyote, which I used ten years ago, or this, PF Sense, you're going to educate yourself quite effectively in how this stuff works.

Gibson: And you'll find, if you're trying to do IP audio, you'll find all the tools in PF Sense to do it, to be able to watch the packets, to see what's going on, to prioritize voice-over-IP or audio-over-IP packets. You'll be able to teach yourself the things you need to do to safely and securely administer your own network. All the tools are there.

If you can Google for the question, you can find the answer. If you're willing to go to your toolbox and pick up the screwdriver, the hammer, and the set of diagonal wire cutters and go fix a problem in your radio station, you probably also are the type of person who wants to be able to fix their own computer problems or fix their network issues. I think a lot of it goes - what's at the heart of it is understanding, and being confident and feeling like it's not black magic, like it's not the guy behind the curtain, but it's you. It's you being in charge of it. And I think that's something that learning your own router teaches you a lot about what's going on.

Kirk: I kind of like being the guy behind the curtain.

Gibson: Well, you do that well, Kirk.

Kirk: [Laughs] Hey, unfortunately we are out of time. We have sped through an hour of chatting about television and audio and IP and some IT stuff with Gibson Prichard. Our show has been brought to you by the folks at Lawo and the crystalCLEAR Audio Console, the touchscreen console, and also brought to you by Omnia and the new Omnia.7. Go check it out on the website at omniaaudio.com/Omnia-7.

You'll be hearing more about that; it's really affordable. And it's been brought to you by the folks at Axia. And one of my favorite things about the Axia AoIP system, about livewire, is soundcard replacement with just a $15 network card and the IP audio driver. I use it in all kinds of places. Check it out if you would at axiaaudio.com. So Gibson Prichard, thank you for being with us. I really appreciate your quick notice of joining our program tonight.

Gibson: It's been fun. It's been great, Kirk. Thanks for having me.

Kirk: I'll have you back again sometime. Especially if you rewire the whole TV station again, be sure you tell us about it.

Gibson: By the way, just like you, I have one of the 300 millimeter tower beacons at my house too.

Kirk: Really? [Laughs]

Gibson: I do. I have it sitting on my back porch. When we re-lamped our tower I kept a beacon.

Kirk: Do you still have the full-size bulbs in it?

Gibson: It is still... I actually put a cord on it. Yes, I can plug it in. It will put 1300 watts of light out over my neighborhood, yes.

Kirk: Well, I don't have mine on the back porch. We don't live in a red light district like you do. Mine's here at the office. It lights up the office pretty darn well.

Gibson: It's a great sign that Rudolph is on the way.

Kirk: Say, that's a good idea right there. We should put it out for Christmas Eve. Hey, again, thanks for being with us.

Next week, I remembered who our guest is next week. It is a guy... he won an award from Radio World Magazine. It is... who is it? What is it that he won here? Oh, the Radio World Excellence in Engineering Award. Okay, all these awards have different names.

The 2014 Radio World Excellence in Engineering Award has been bestowed upon Wayne Pecena who is an educator and he does these SBE webinars where he teaches all about different aspects. Typically it's IP systems, IT, Information Technology networks. Wayne is going to be our guest on the show next week. So the guy who usually does all these webinars, he's a great educator and a fine southern gentleman from Texas, he'll be on our show next week.

Then we're going to take a couple of weeks off for Christmas and New Year's, so a couple of programming notes there. I hope you'll be sure and join us for next week's show with Wayne Pecena. So there you go. Gibson, thanks again for being with us and we'll see you all next week on This Week in Radio Tech. Bye-bye, everybody.

Topics: Broadcast Engineering

Subscribe

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

Recent Posts

Posts by Topic

see all