Station: WBEB(FM)Market: Philadelphia, No. 4
Owners: David Kurtz and Jerry Lee
Frequency: 101.1 MHz
Information: (610) 667-8400 or Web site www.b101radio.comAs radio makes its transition to digital, one of the leaders is WBEB(FM) in Philadelphia.
The soft rock station makes plenty of headlines. This year it celebrates 35 years as an independently owned station -- unusual if not unique in the top five markets.
Co-owner Jerry Lee is a legend, thanks to decades of innovation and service to organizations like the Broadcast Industry Council. The Philadelphia Inquirer reported that the station owners turned down a $100 million offer two years ago. And in the spring Arbitron ratings, the station climbed to the top of Philly's competitive FM pile.
Chief Engineer Russ Mundschenk has achieved his own share of national recognition. Colleagues seek his expertise on digital matters, because WBEB was among the first radio stations to install digital consoles and to explore the new world of digital switching and synchronization.
Mundschenk spoke to RW Editor Paul J. McLane about the station's recent $375,000 studio upgrade project.
This is one in a series of articles about how radio facilities are implementing digital technology.
RW:Describe this facility.
Mundschenk: We have three full studios: one air studio and two full production rooms, and a voice booth off the main production room. There are two full-time production people, who stagger their shifts so we can cover weekends and after-hours. We have about six full-time announcers and five or six part-timers.
RW:When did you begin the project?
Mundschenk: We've been talking about rebuilding studios since the day I came here in 1983. … The studios had been built for beautiful music. It just wasn't set up.
About 1994 I gave Frank Grundstein at Harris a call. He's a good friend of mind, and I said, "I want to rebuild my studios, but I think it would be ludicrous for me to rebuild them with analog equipment. After all, the equipment we have in our studios is digital, it only makes sense to keep it in that realm."
He said, "We are in fact talking to somebody about building a digital console for Harris." … I went up to Zaxcom in New Jersey, and they showed me their proposed product. It pretty much had all the things that I had requested.
The thing I wanted most was programmability and versatility. You certainly can make an analog console that sounds as good as a digital one, but to be able to get the kind of versatility out of it that you can using DSP technology is what really makes it shine.
RW:When was the project completed?
Mundschenk: The production and air studios were completed in the fall of 1997. Maestro system
Gear at the BA sampling of WBEB's equipment:
Computer Concepts Maestro System
Harris DRC-1000 Digital Console
Logitek PRE-10 Input Expander
AKG 414 Mics
360 Systems Short/cut
Sage Endec EAS
Leitch 12X1SBA Video Switcher
Computer for Internet access for jock
Studio Technology Furniture
Harris DRC-1000 Digital Console
SoundForge Editing Software
SoundForge CD Architect
Denon DN-951FA CD Players
Tascam 122 Mk III Cassette
Otari MX-5050 RR
Gentner TS612 Phone System
Roland Effects Processor and Voice Synth
Eventide H3000B UltraHarmonizer
Technics CD and Cassette Players
Studio Technology Furniture
Cutting Edge Omnia Processor
Moseley PCL-606C STL
Harris Digit Exciter
CSI T20-F Transmitter
ERI Two-Bay Panel
RW: How is your on-air audio delivered?
Mundschenk: It's a live assist format. We run two Computer Concepts Maestro II systems. We operate them in such a way that the left machine runs the commercials, and the right machine runs the music. Either is capable of running the entire format, but that's the way we get our redundancy. Each is in a peer-to-peer configuration, so if one went down, we could go to the other one immediately.
Everything is on hard disk.
RW:How are you storing that? Are you happy with your algorithm?
Mundschenk: We're using the apt-X algorithm, which is what Computer Concepts started out with. In the past four or five years, they've made MPEG available, but we're using the apt-X format.
It's a 4:1 compression system. Each of the machines has 18 GB of storage; that translates to about 150 hours in stereo.
I really like the way it sounds, and that's the bottom line. Yes, it only does 4:1, but in my opinion, high compression ratios are really only important where you have an absolute need to do it, like limited bandwidth on an RF signal.
RW:Your consoles are the Harris DRC-1000s. What lessons did you learn in installing some of the first digital consoles to be used in the radio environment?
Mundschenk: If a console says it's digital, make sure it's totally digital.
I can think of one prominent model out there that is not totally digital. There are only effects on the mic channels; all the auxiliary channels are still analog. They mount their audio processing unit directly underneath the console, instead of in an environmentally controlled equipment room.
The way Zaxcom does it, with a separate control surface and a number-crunching audio processing unit in an equipment room, is a much better way to handle it. You want to keep everything centralized, in an environment that you have control over.
RW:What else does the console do that allows you more flexibility over a traditional analog design?
Mundschenk: It is infinitely programmable. Every channel has multiple effects. Compression, limiting, gating, AGC -- which is a separate compressor, a gain rider, great for on-air. It's got equalization, five bands. Five types of equalization, available on each five bands, and all of this per channel!
RW:Any unexpected problems or complications in using the console, by nature of its digital approach?
Mundschenk: It's been relatively bulletproof. … It would be nice if the console had the ability to download a file from a computer, to be able to configure it, rather than sitting there for hours putting in individual letters.
When they introduce software, the new version is on a PROM that you have to pop out and put in. They didn't give me ZIF (Zero Insertion Force) sockets to do it. And some new versions wipe out all our presets, so we have to redo the whole thing from the beginning.
You can save setups to a PCMCIA card on the console, so you can go from one console to another.
RW:What percentage of your day is satellite fed?
Mundschenk: At 8 o'clock at night on weekdays, we take the "Delilah" show from Broadcast Programming. That is delivered via a StarGuide system, on Satcom C5. It's MPEG compressed, Musicam.
I'm not really happy with the audio quality on satellite, but it could be the compression algorithm, or multiple A-D conversions. The satellite receiver does not have a digital output. I've tried to get them to provide me with one, but they haven't gotten enough requests for that yet. ID vs. 1992
RW: How do your devices talk to each other?
Mundschenk: It's the AES/EBU format. That's pretty much standard. But there are two types of AES/EBU. There's AES3-ID format, which is the same data protocol but with a different impedance and voltage than the AES3-1992.
AES3-1992 uses a voltage of between 3 and 7 volts peak to peak, at 110 ohms balanced.
AES3-ID is a voltage of 1 volt, like video, and it's 75 ohms, like video. As a result, you can lose a lot of video stuff: DAs, switchers, equipment of that sort.
We use the ID format. It's really easy to put on BNC connectors; the wiring is simple.
As far as trying to maintain the integrity of the impedance along your entire transmission path, it's much better. Coaxial line is meant to carry RF. And it is RF. It's 6 MHz. You've got to start thinking RF when you're talking about AES/EBU. At 48 kHz, if you sent all "ones," you'd have a signal that occupies 6 MHz worth of bandwidth.
Look at some of these balanced 110 ohm systems, and the connectors that they use going in and out of the pieces of equipment. You realize that, every time you make a transition, you have the potential of introducing capacitive rolloff, for reflections and all the other problems that a bump in the path, impedance-wise, will cause.
That's why I like the coaxial system. It's clean. And you can transmit AES3-ID up to 1,000 feet, whereas AES3-1992 balanced can only go a few hundred without some type of equalization.
RW: What would I see if I looked between your Computer Concepts system and your mixer, or your airchain?
Mundschenk: Computer Concepts is another one of those companies that has not developed a digital output for all audio channels using their standard cards. They announced at the Seattle NAB show that they will soon be supporting Digigram audio cards, which use MPEG compression and also have digital I/Os. I only wish that the standard apt-X cards supported digital I/Os on all channels.
Anyway, we took the analog output of the Computer Concepts, and fed it directly into the best A/D converter that I could find at the time, Benchmark's AD2004. We have Serial No. 1 of those.
Benchmark knew how to preprocess the analog audio so that it would exhibit the least amount of artifacts after A/D conversion. It's a great box.
RW:What about routing digital audio, beyond the console?
Mundschenk: For studio switching we've made interesting use of a video switcher made by Leitch, called the Xpress Series. We like it because it is a 12x1 stereo analog, 12x1 video switcher.
We have the ability to individually or separately switch either the analog or the video channel. We put the digital through the video side, we put the analog through the other side. Our analog output goes to our analog audio processing, which is used as a backup, and the digital goes to our digital audio processing.
We've built our own custom drivers for remote control panels, and we're able to have a remote control panel in each studio. I put in a little enable switch, so you have to hold down two switches at the same time to switch the on-air switcher, deciding which room goes to the audio processing. Bandwidth concerns
RW: Let's follow that air chain. From there, the main signal is fed out of the Leitch switcher in digital…
Mundschenk: To an Omnia processor, from Cutting Edge Technologies, which then feeds an analog STL.
RW:After all this digital here?
Mundschenk: Just recently I've been made aware that the Moseley Starlink units have been out in the field for a while. They use 48 kHz sampling, which equates to a 24 kHz bandwidth, versus some competitive units, which operate with 32 kHz sampling and a 16 kHz bandwidth.
Why worry about 24 kHz bandwidth, when 15 kHz is the cutoff in FM?
I want to avoid as many rate conversions as I can. I want to keep the signal as clean as possible. If we have a 48 kHz sampling/24 kHz bandwidth system in this facility, I want to keep it 48 kHz. This is fast becoming the adopted standard in broadcast. As I buy new equipment, I want to try to get the most bandwidth that I can, within reason -- certainly something that would support DAB in the future.
I'm thinking seriously about the Moseley Starlink, for that capacity. Right now we're using the PCL606. That goes to the transmitter, into a Harris Digit exciter, which we like a lot. The transmitter itself is a CSI transmitter.
RW:How much of a factor did digital audio broadcasting play in the design of your studios?
Mundschenk: Only in that I wanted to be able to deliver the quality of signal necessary for DAB when the time comes. I didn't want any critical part of my chain to be limited to 32 kHz. Synchronization
RW: House synch is a new phrase for a lot of people in radio. How do you accomplish it?
Mundschenk: House synch is very important.
You want a house synchronization system that is very reliable, and very stable. This is your bread and butter. If you lose it, you're dead.
If you lost your house synch, and you had a device that doesn't automatically switch over to internal synch -- which the Zaxcom consoles do not, but the Benchmark A/Ds do, smoothly -- there's a very good chance that your console is going to start free-running at whatever frequency it will synch to. The final product may end up sounding pitch changed. If you have devices that use house synch on the input, and you lose synch, you're going to get errors.
Every single piece of equipment that can be synchronized to a house synch, should be. …
It's important that you maintain a standard. And if you can do that, independent of the equipment, with a piece of equipment you know to be extremely precise and reliable, there's nothing like it. We're talking about a DARS digital audio reference signal generator that has a jitter spec of 10 picoseconds, or less. Preferably something around 2 would be nice.
Let's go another level. You've got your DARS synch, and you're feeding it to all these pieces of equipment that have synchronization, but how are you going to do that? If you split it with a DA, there's another piece of equipment that could fail.
Really the most ideal way is to have some type of switching arrangement that will switch over to a backup unit, if you lose your main source of synch. I don't have that implemented yet, but it's important.
RW:Are your fellow engineers educated about digital in general?
Mundschenk: Nobody knows. I didn't know. It's all new.
We need to have more awareness of the digital environment, to really understand what's going on.
RW:Digital product manufacturers reading this article might ask what they could do to make your life easier as a user of digital products.
Mundschenk: Just give me a digital input and output, that's all.
It's very simple. AES3. I'd like the ID format, but that's a matter of personal preference. It doesn't matter. You come out AES3-1992 with a 3 to 7 voltage range and a 110 ohm impedance, balanced, and you can convert that with a transformer. Ditto the other way.
But just give me a digital output. I bought a digital satellite receiver. It doesn't have a digital output. I've got a digital commercial system. It doesn't have a digital output. I've got all this stuff that is digital, but they all don't have digital inputs and outputs on every channel!
This is the way things are heading, folks, and you're going to have to do it eventually. Put your money into it now, so you can say that you are digital-ready.
RW:Who did your studio wiring?
Mundschenk: We did it, my assistant Kevin Kilbourne and I. He is very intuitive and does a great job of sorting stuff out for me.
We also have a full-time MIS director. This is unusual, for a station of this size to have three technical people. But if you take a look at our usage of computers, there's no way two people can handle it. Splitting and switching
RW: What observations can you offer to engineers thinking about these digital audio questions?
Mundschenk: The big "if" is how we tie it all together. It's the console. Routing. You can use a video router; there are now AES3 routers available to do the same thing.
We're not doing too much splitting; where we've had to, we use these little block DAs from Video Accessory Corp. They're flat to 200 MHz, and we've done tests to take a look at their square wave performance. They're infinitely better than a lot of others in that price range, about $180 each. One in, four out, nice. They even have adjustable gain.
For switching, the switcher we have is a nonsychronous switcher; that means a switch could happen at any point in the format, in the middle of a digital word, whatever. This means that whatever is downstream requires longer to synch up.
It doesn't sound too bad, really. They make a big stink about wanting to do nonsynchronous switching, but in our case, with the stuff that's downstream, you really don't notice any objectionable noise, big pops or anything like that.
RW: What's in your equipment room?
Mundschenk: We use video patch bays. Everything in the facility, every single audio source is available in the equipment room. You can patch anything to anything there. They all run through patch panels.
RW:An awful lot of steps.
Mundschenk: Yes, but it gives you total versatility. You can do anything you want, and frankly we've used it.
RW:You're serious about radio.
Mundschenk: Radio has been a life-long passion of mine. I don't know why.
When I was in Florida, I did a lot of work in TV, helped install a transmitter, did some repair of cameras and VCRs and things like that. TV wasn't the same to me as radio. Radio has the warm fuzzies that TV doesn't. Not to put down TV as an environment. But there is a warmth in radio that just doesn't exist there.
RW:Is that changing now, post-consolidation?
Mundschenk: I don't see it here, because I work for a mom-and-pop station -- or a pop-and-pop station.
I think it is, though; I'm sorry to see it. There are a lot of people out there who are trying to pinch every penny.
This is not one of those radio stations. We spend money to make money here. And they must be doing something right because they're the No. 1 FM radio station in the city right now.
Station: WBEB(FM)Market: Philadelphia, No. 4