Consider: From now on, just about every video title created, whether for episodic television, advertising, theatrical, home theater, or short-form Internet viewing, needs to be formatted in multiple ways, so as to suit the requirements of the native viewing environment. Watching an episode of “Lost,” for instance, or any other TV show, plays differently on a handheld, than it does on a big-screen TV.
These multi-formatting requirements also extend to the looming 3DTV viewing environments. Watching Avatar in the theater, for instance, carries different resolution and distribution requirements than watching it on home TV. Plus, 3DTV adds some spin to the challenge because of the extra bandwidth associated with what’s essentially a two-camera/two-eye content creation methodology.
That’s why we at Cisco are so focused on what we call the “IP contribution and distribution” part of the video life-cycle. It’s all about making it easy for content owners and service providers to compress, process and distribute their video products, in formats that can be consumed on lots of different viewing surfaces.
Not that long ago, the notion of sending “contribution-grade” video over an IP network was considered unlikely, at best. Packet sizes were too big and unwieldy, people thought, not to mention the “special needs” of video, relative to other data types.
“Contribution-grade” is a broadcast video engineering term meaning “highest possible quality” – not unlike a digital master. When content owners prepare video shows and events for distribution, for instance, they want at the onset the cleanest, highest-grade product because it inevitably gets manipulated and re-processed along the way.
Because contribution-grade video typically connotes an uncompressed, lossless stream, it wasn’t immediately obvious that an IP network could handle it.
That’s all changing, now. In large part, it’s because contemporary media distribution involves carriage of multiple file formats, sizes, and resolutions, all of which are in a constant state of change. IP allows media distributors to deal with multiple content types, so as to provision and dispatch them quickly.
The holding pens for video content entering production facilities for editing are getting pretty full, and this year’s NAB trend lines only add more heft to the scene. More and more HD/3D content, for instance, will add 20 to 30% more bit bulk, per title; high definition streams alone can run as high as 4 Gbps.
Plus, there’s the core issue of video packets, in general. People often talk about how video has “special requirements,” relative to other standard IP applications, like email, voice, and related broadband Internet fare.
It’s true. Video does have more rigorous needs than other types of IP traffic.
Why? One contributor is simply packet size – video packets can be as large as 4 Megabytes – compared to packet sizes measured in mere kilobytes. As George Tupy, Digital Media Solutions Manager for Cisco, describes in the video below, it’s like the difference between cars and trains, attempting to merge on a highway.
In this sense, the cars are “regular” IT data packets, the trains are the video packets, and the highway is the IP backbone. Obviously, merging cars on a highway is one thing, and not a big deal. Merging trains on the same size highway requires one or more to wait (or crash.)
Up until recently, the rallying cry from the annual gathering of broadcast TV professionals centered on the “digital transition,” or, the 2009 deadline by which U.S. broadcasters shifted their analog transmissions to an all-digital format.
At this year’s National Association of Broadcasters (NAB) show, though, “all digital” is assumed, and the buzz of the show is “all-IP,” as a foundation for 3D, HD, and multi-platform television expanded.
A case in point: Cisco’s news with Intelsat, to partner on a first-of-its-kind delivery platform that blends both satellite and terrestrial delivery technologies. That means Intelsat’s customers will be able to specify a delivery path – say, Chicago to San Francisco, or New York to L.A. – and the Cisco-powered Intelsat platform will determine the best and most cost-effective transmission route with built in resiliency.
Specifically, Intelsat S.A., a 45-year provider of global, fixed satellite services, is outfitting what it calls its “IntelsatONE” platform with an unprecedented combination of IP/MPLS networking, fiber, teleports, and points of presence – all integrated with Intelsat’s existing satellite fleet. Our work with Intelsat is a gratifying and important reference for a global IP video network. It shows the importance of combining the flexibility of IP networking with the leading edge video technology, all controlled by one comprehensive management platform.
Actually, I have been thinking about real time data exchange challenges and opportunities between service providers or even “served up” through a cloud. Characterizing “real time” could be in the form of video, telepresence, voice, on-line games as examples, where mitigating latency is a must e.g. between clients-servers; agreement on what the optimal communications protocol should be and so on in order to deliver these services as expected by the consumer.
Jim Waldo, Sun Microsystems Laboratories, articulates some of the scaling challenges related to entertainment e.g. on-line games and virtual worlds, in his article published for ACM and entitled, “Scaling in Games and Virtual Worlds,”
“An avatar exists in at most one location within a shared virtual space. Conforming client applications use the protocol to manipulate and move the user’s avatar, create virtual objects, interact with other users and their surroundings, consume and create media and information from sources inside and outside their simulated environment.”
Entertainment-centric applications like on-line gaming and virtual world instantiation will certainly stimulate innovation on current IP-based networks, in addition to the Internet; ditto for novel client-server scaling mechanisms.