HONG KONG— The humble BNC connector has been around for so long that most broadcast engineers don’t even know what the letters “BNC” mean. Ask around the workshop: “What do the letters B-N-C stand for?” I’ve yet to find someone who knows the answer off the top of their head.
I had to Google it myself. “Bayonet Neill-Concelman” is the correct answer. We can thank Paul Neill and Carl Concelman for designing that compact, quick, easy and reliable coaxial cable interface that faithfully transports our real time synchronous television content between pretty much every piece of broadcast baseband video equipment—and a lot of audio content too.
Neill and Concelman had military applications in mind when they designed this ubiquitous signal interface for the ends of coaxial cables. The military aspect of their design objective makes for an interesting coincidence when you consider that the patent application on which they based their collaborative effort was filed with the US Patent Office just a couple of weeks after World War II ended in Europe in 1945. As the tired, hungry and demoralized soldiers laid down their weapons across Europe, Octavio Salati was in his lab at Hazeltine Corporation in springtime America, enthusiastically writing his patent application, complete with meticulously hand-drawn crosssectional images for his “electrical connector.” He filed his patent on May 19, 1945. His patent protection was awarded to him on January 30, 1951. (The patent approval process took about the same amount of time as World War II.)
The idea behind the BNC connector is older than the author of this article. But not by much. That the BNC connector is still so thoroughly entrenched in our daily lives in 2014 is quite remarkable. Its longevity is testimony to the success of its design. But, there are forces at work to end the 60-year reign of our small round shiny friend.
The inevitable replacement for your old faithful BNC/coax combo is the cheap but rather unsatisfying plastic 8P8C connector on dual unshielded twisted pair (UTP). After 60 years of saying “BNC on coax,” we’re supposed to soon be saying “8P8C on UTP.” It’s not quite the same, is it? “8P8C on UTP” doesn’t roll off the tongue particularly well.
That’s why nearly everyone calls it “Ethernet cable.” Even “Ethernet cable” is too much of a mouthful for some people, so you’ll typically hear them say “Cat 5,” “Cat 5e” or “Cat 6” instead—referring to the relatively inexpensive category 5 or category 6 unshielded twisted pair copper cables for Gigabit Ethernet applications. “8P8C” is rather lacking in aesthetic appeal as far as names go. Why didn’t they ask George Lucas for suggestions? He came up with C3PO and R2D2 for the Star Wars movie. He could have done much better for us than “8P8C.”
And that’s why we all prefer to call it a modular connector, or perhaps more commonly and somewhat inaccurately, the RJ45 connector.
It goes to prove one thing—that the IT and telco industries are definitely at the upper limit of their creativity with three letter acronyms (TLAs).They can’t do four. They need help from someone like George Lucas for that, but he charges too much. Cost is important. Cost is the basic motivation behind the effort to replace all of the broadcaster’s BNC/coax infrastructure with Ethernet cables. It’s already a fait accompli for non-real time asynchronous video transfer.
Relocating or replicating broadcast MXF files around a TV station or production house and even around the planet through Ethernet cables and IT/telco packet switching infrastructure is already the done thing. It makes sense to do that now. Enough bandwidth is available at the right price, and for simple file relocation and replication, we can accept some degree of variable transfer speeds, interruptions and latencies.
But streaming synchronous live uncompressed high definition video as a playout signal or pass-through signal or a live camera source back to the studio, remains a very tall order for a TCP/IP network. It isn’t only a matter of having enough bandwidth and not too much latency. It’s also a matter of having a whole ecosystem of IP friendly devices that support all the operational functions of television production and broadcasting—and all the devices in that ecosystem need to obey the network protocol rules so that traffic flows smoothly—without bottlenecks, hiccups or crashes.
THE AVB EXAMPLE
An ecosystem of IP-friendly streaming video devices can only come about if three things exist. First, there has to be the will to do things entirely in the IP domain. Second, there has to be a set of standards to follow for the bit streams and electrical interfaces and network protocols. And third, we need a marketplace of manufacturers to invest in the launching of a whole new family of products that conform to the new standards, and those manufacturers must have some hope of selling gazillions of said products.
So far, in May 2014, enough of a will has existed for a long enough duration to have brought about a set of standards. The standards can be found under the umbrella term “AVB” - Audio Video Bridging. The AVB set of four IEEE standards is comprised of the following: IEEE 802.1AS (precise time synchronization), 802.1Qat (Stream Reservation Protocol), 802.1Qav (“traffic shaping” to prevent bottlenecks and packet clogging in queues) and 802.1BA (identifying which devices on the network are AVB-capable and which ones are not).
A surprising discovery when one looks for information on these standards is the fact that the family of four standards was finalized and published about three years ago. It begs these questions: “Why is it taking so long? Why hasn’t AVB taken the television world by storm?” The answers to questions like those are rarely simple. But there is an important side note.
There was one more important standard required, and that standard wasn’t ratified until August last year: IEEE Std 1722.1-2013. That’s the one that provides a kind of intelligence for a sort of plug and play AVB device to just work when you connect it to the bridge. Without that, a lot of manual configuration on setup menus would be required before you could plug an AVB device into the bridge. In reality, the AVB standards were not really a complete set until about nine months ago. That could partly explain why AVB systems aren’t all around us yet.
As part of our further analysis for the purpose of finding more answers, let’s compare the upcoming rollout of AVB to the rollout of digital video over coaxial cable 25 years earlier. The Serial Digital Interface (SDI) first saw the light of day when Sony and Ampex launched their D2 format 19mm composite video tape recorders. To make D2 VTRs simpler to install within the existing TV station cabling infrastructure, the parallel data streams on the cumbersome and distance-challenged parallel digital interface used on the early D1 VTRs had to be serialised by the inclusion of a new dedicated chipset behind the male BNC connectors on the D2 VTR. It was that humble pair of serializer and deserializer LSI chips that allowed a standard BNC connector on 75 ohm coaxial cable to pass SD digital video streams at 270 Mbps.
Sony had the market for that SDI chipset very much to themselves in the early days. They sold the serializer/deserializer chipset to all the other broadcast video equipment manufacturers. That low-cost chipset was a major contributing factor that allowed the whole broadcast industry to quickly and cheaply adopt the new SDI digital video interface.
Today there are many semiconductor manufacturers producing HD-SDI transceiver chipsets. Within a very short span of time, video monitors, audio monitors, audio recorders and players, CGs, still stores, test equipment, switchers, routers and every other video-related device was “in the game.” SDI was adopted with great enthusiasm because the SDI ecosystem was quick and easy to create. And that SDI chipset remains, with its BNC-oncoax bedfellow, in wide and stable use today.
WRITING THE SCRIPT
That story of SDI is a hard act to follow. But follow it AVB must. Writing the script of the AVB story is a group called the AVnu Alliance (www.avnu.org). The AVnu Alliance is the AV industry body responsible for certifying products for compliance to the AVB standards. Passing all
the compliance tests permits the manufacturer to attach the “AVnu Certified” logo to their product. That logo is what you look for when you’re shopping for the parts to build yourself a BNC-free AVB zone.
The first question on my mind as I navigated my way to the AVnu Alliance web site was “If I’m building an AVB facility today, how many components of the system can I buy that are already AVB certified?” The answer surprised me. The answer is “one.” Uno. Un. Ichi. To be fair, the official list of AVnu Certified products as of May, 2014, has 14 items on it. But those 14 items are the various models of one product line: the Extreme Networks “Summit X440” series of Gigabit Ethernet Switch. It’s appropriate that the first item to pop out of the AVnu certification process is the switch. Without a switch, there’s no AVB. And in the Audio Video Bridge world, the whole system revolves around the bridge. In AVnu language the certified Gigabit Ethernet switch falls under the product category called “bridge.”
Okay, we’ve bought ourselves a bridge, thanks to the good work of Extreme Networks and AVnu Alliance. What do we start saving our pennies for now? What will be the next products to earn their AVnu Certified logo? How long will it take us to build our BNC-free AVB platform? What other products are coming down the AVnu certification lab’s pipeline?
At NAB this year, Arista Networks was making noises about the upcoming AVB certification for their Ethernet switches. They gave a very convincing demonstration of frame accurate real time glitch-free switching of live HD streams. That demo was in partnership with Fox Networks Engineering & Operations. So this isn’t just marketing smoke and mirrors. This is a real world practical demonstration—of an AVB bridge, with a real broadcaster. So we have another bridge coming for sure when Arista’s products get their AVnu logo.
We can cross “bridge” off the shopping list. What about cameras, microphones, monitors, test equipment, amplifiers, loudspeakers, legacy videotape players, audio mixers, real time digital effects, realtime graphics, video mixers with keyers, and etc.? When will they come out with an 8P8C socket and AVB compatibility? Sennheiser demonstrated a prototype of an AVB microphone in 2012. There’s no more recent news than that on their web site.
Beyerdynamic has partnered with XMOS to utilize an external AVB gateway for the microphones in their wireless conference room systems. So yes, we can now get a microphone signal into an AVB bridge, thanks to the XMOS “mic in to AVB out” box. But that’s a conference room microphone network. What about a high quality voice-over booth microphone? Or a studio shotgun boom mic? “To be confirmed,” is the current reply.
Wohler announced an AVB interface card for their AMP2 series of video/audio monitor just before NAB this year. This gives us the ability to monitor, mix, level adjust and remux the audio channels in an AVB stream. Now we can cross “audio monitor/level adjuster” off the shopping list. That’s about it. A bridge, some conference room microphones and an audio/video monitor/adjuster are what we can easily find today. That’s not much. AVB promises to save us so much money. Why is the television world not overtly beating a path to the AVB door? If we look at the list of AVnu members on the AVnu web site, the most familiar names to us television people are Avid and Dolby. Sony isn’t there. Neither is Panasonic, nor Grass Valley, nor Imagine Communications (nee Harris), nor Miranda, nor Blackmagic Design, nor JVC. Yamaha is there, but broadcast television is only a small part of their total business.
Avid is on the AVnu membership list, mostly it appears, with their audio hat on. Avid’s ProTools and SC48 mixing console are definitely candidates for a place in an AVB world. Shure, Sennheiser, Bose, Harman and Beyerdynamic are AVnu Alliance members. So it appears that the audio world shows more interest in networking their real-time baseband payload over Ethernet than the video world does.
Barco is an AVnu Alliance member. Barco’s presence among all the audio names in the list is a hint that AVB might have more relevance in non-broadcast AV applications like auditoriums, concert halls, stadiums, houses of worship, and the like. But somewhat like a fox in a flock of sheep, appears the name of Audinate in the list of AVnu members. Audinate has already established a market for their Dante product line, which is a credible Ethernet-based near-enough-realtime audio networking solution over standard commodity Ethernet switches. Audinate is surely wise to be in the AVB group so their current headline product, Dante, can take advantage of the new capabilities offered by AVB bridges.
But the success of Dante on existing non-AVB Ethernet network infrastructure is, one might expect, also a factor that could be taking some shine off the whole AVB concept. The audio world already has a solution of sorts in Dante. Why do they need AVB?
AVB FOR CONSUMERS
To make matters worse for AVB, when we look at the consumer AV device market, it appears to be far ahead of AVB with their own networking solutions as part of a quite mature ecosystem nurtured by the HDBaseT Alliance (www.hdbaset.org). The HDBaseT Alliance was founded by LG Electronics, Samsung Electronics, Sony Pictures Entertainment and Valens in 2010. Their secret weapon is what they call “5Play on a Single Cable”, which means they can carry all of the following on a single Gigabit Ethernet cable: Up to 4K video (one way) PLUS digital audio (one way) PLUS 100BaseT Ethernet (two way) PLUS control signals (two way) PLUS up to 100W of DC power (to the receiving device).
The Ethernet cable for HDBaseT is a standard commodity 8P8C on Cat5e/6 UTP cable from your local computer store. And up to 100 metres in a single run is feasible. But the HDBaseT switch and data transport method on the cable is proprietary. It isn’t based on packet switching like Ethernet. It’s based on a superset of HDMI. There are already enough HDBaseT devices to build a system. The reason the HDBaseT Alliance could establish an ecosystem of devices so effectively is because the whole ecosystem is based on an extension to the HDMI chipset. In other words, it’s just like 25 years ago when SDI easily established a foothold in our business thanks to the SDI chipset.
The HDMI chipset already handles high bit-rate video and audio. And most importantly, it already handles the HDCP copy protection protocols for Hollywood HD content. By combining the HDMI chipset with the new Valens HDBaseT chipset, the “5Play” killer app could be realized. That brings us to AVB’s Achilles heel. They don’t have a chipset that makes a microphone or other AV device connect easily to an AVB bridge. It’s going to take a long time to write all the device AVB driver software and have every device AVnu certified. A chipset negates the need for all that time. AVB doesn’t have a chipset. They only own a set of IEEE standards. They really need a chipset.
So it looks like we might be in for a long wait for a BNC-free zone from AVB. The good news is, if you buy that Extreme Networks X440 switch today, you can still get value from it today as a normal Ethernet switch, and take comfort in the fact of being “AVB ready” for when those AVB devices eventually do come out of the certification pipeline.