itvt

As director of interactive delivery for NDS Americas, Steve Tranter is based in the company’s Southern California office, where he heads up its Interactive Delivery group. In this recent, wide-ranging interview with [itvt]’s Tracy Swedlow, he discusses NDS’s interactive TV games strategy, which, among other things, sees it working with broadcasters to develop branded games that are designed to keep viewers tuned to their channels; its Xtreamplay platform, which uses the DVR hard drive to enable video-rich games, advertising, and other interactive applications; new features that it is incorporating into its MediaHighway middleware, including peer-to-peer capabilities; the new version of its Synamedia Metro IPTV platform, which is currently in trials with SES AMERICOM; its efforts to develop a next-generation EPG for HDTV; how it is working with chip vendors to ensure that their future chips will support the interactive TV services that are being conceptualized today; and much more.

[itvt]: What has NDS been working on recently?

Tranter: Among other things, we recently launched games on DirecTV and we’ll shortly be launching games on DISH–so we’re really active in the games space right now. This summer, we’ll be launching our new IPTV middleware, which is currently being trialed by SES AMERICOM. We’re working on a lot of gambling-related interactive TV projects. And we’ve also branched into something new for us–developing a security system, called RadioGuard, for high-definition radio.

Part I: Games

[itvt]: So let’s start by discussing NDS’s recent activities in the interactive TV games space.

Tranter: Most of our games are developed out of Denmark, through our NDS Denmark division–which was formerly an independent company called Visionik. As I mentioned, in North America, we recently deployed games on DirecTV, and we’ll shortly be deploying games on EchoStar. We also have games on Cablevision in the US and on Bell ExpressVu in Canada.

On DirecTV, for example, we have around 10 games. Seven of those games are generic games, but we also have branded games from Mattel–so we have Uno on there, which is always in the top five games on DirecTV–and we also have games based on various Nickelodeon brands. We actually work with a lot of kids’ brands when it comes to games.

As you know, we’ve offered games in the UK for some time now, and we have branded children’s games there that are the result of partnerships with Discovery Kids, Nickelodeon and Cartoon Network. We’re actually planning to bring those games to the US.

[itvt]: Which type of games do you find are proving more popular and lucrative: generic games or branded games?

Tranter: They complement each other. What usually tends to happen in this space is that platform operators tend to start with generic, non-branded games. That’s because you have more control over when to launch those games and how to package and market them–for example, you’ll typically want to rotate them to keep them fresh. So, generally, generic games serve as a good starting point for differentiating an operator in the marketplace, and for getting people used to interacting with the set-top box–and they also generate revenue on a standalone basis.

Now one of the big differences between generic games and branded games–at least in our approach–is that generic games tend to be launched from a portal or some other dedicated location in the EPG, whereas branded games–because most of our branded games are the result of partnerships with broadcasters–tend to be launched directly from the channel you’re watching.

[itvt]: Just to clarify, you’re not yet offering branded games in the US that are launched from channels rather than portals, correct?

Tranter: Correct. I’m referring to our deployments on BSkyB, Sky Italia and Foxtel–although I expect this approach to be adopted in the US as casual gaming becomes more established. Anyhow, how the approach works is that, if you’re watching "Scooby Doo," for example, you can launch a "Scooby Doo"-branded game directly from the program. In our experience, these kinds of branded games get more usage: from the viewer’s perspective, you’ve got to be pretty motivated to proactively find and play a generic game, but, if you’re already watching a particular channel or program, it’s likely that you’re going to be interested enough in that channel’s or program’s brand to want to play a game based on that brand.

Now one thing that we do with these branded games is ensure that you can carry on watching the channel you’re tuned to: so the channel appears in quarter-screen mode or the game is a semi-transparent overlay and the channel continues to broadcast underneath it. Let me give you an example of how we ensure that you continue to watch a channel while you’re playing one of our games: in the UK, we have a game called "Miniature Golf" on Nickelodeon. The broadcast feed becomes a quarter screen and is incorporated into the game in a way because it appears in a different part of the screen, depending on which hole you’re playing. In effect, the broadcast feed becomes one of the obstacles you have to hit the ball around. The result is that you continue to watch the channel you were tuned to and, very importantly, that you continue to watch it through the commercial breaks. So the game actually keeps you on the channel longer: what we’re finding with these branded games is that people tend to play for at least 20 minutes. And what happens is that a lot of the channels we work with are actually scheduling games in the same way that they schedule programs: Discovery Kids in the UK, for example, actually schedule their games throughout the day and promote them avidly: they found that something like half of their viewers at any one time were playing the games. Now, that’s at the high end, and usually the percentage of viewers playing a game is going to be smaller than that. But anywhere between 20% and 50% is pretty typical.

[itvt]: Can you give an example of how Discovery Kids programs its games in parallel with its shows?

Tranter: Yes. For Discovery Kids we developed a series of episodic games for the "Mystery Hunters" program. The game episodes gave the viewers the chance to immerse themselves in the game as the on-screen characters, searching hidden temples and solving puzzles. The game episodes were released every few weeks, while the program was airing, to keep them fresh, and they were regularly promoted to viewers along with the actual program. In some cases, Discovery Kids provides as much promotion for their game schedule as they do for their programming.

[itvt]: Now the ability to shrink the screen while you play a game…

Tranter: Is dependent on the middleware. Obviously our own MediaHighway middleware can support this, as can OpenTV and Power TV. We’ve been doing this on all those platforms.

As I mentioned, we also do some clever things where the broadcast feed remains in full-screen mode and the game appears as a semi-transparent overlay on top of it. For example, we’ve developed a game called "Chomp" for Discovery, where players eat away at the overlay to reveal the video beneath it.

[itvt]: As I recall, one of the first games NDS–or rather Visionik–offered that was designed to keep viewers tuned to a channel was the "Ad Break Tennis" game you built for MTV.

Tranter: Yes. That was a long time ago. We’ve been doing these kinds of games for several years now.

[itvt]: What are the business models for your games?

Tranter: They’re offered under a direct subscription model, under a pay-per-play or time model, or under an advertising-based model, such as I just described, where the games are designed to keep you tuned to a channel and watching its commercials. Games that are offered for free tend only to be available from the associated channel, as their purpose is to retain viewers for longer on the channel and to get greater reach for the ads. That’s the case on BSkyB, for example, for Nickelodeon. Though some broadcasters do have a slight variation on that model: on Discovery Kids, for example, the games are free, but if you want to submit your high score, you have to pay.

[itvt]: And branded games are consistently played more than non-branded games, correct?

Tranter: In general, yes. Branded games tend to be played more, simply because of the channel’s promotion and the fact that viewers generally have more opportunity to become aware of them. That’s especially the case when a channel directs viewers to its branded games.

[itvt]: You also offer multiplayer games, correct?

Tranter: Yes. We’re starting to get a lot of requests for those from our clients. People definitely like gaming in a community environment. We’ve just launched multiplayer poker in the UK, for example, and have been offering that in conjunction with betting. We’re also getting a lot of requests for multiplayer and multiplatform games from IPTV providers.

[itvt]: What kinds of community features do you offer around your multiplayer games?

Tranter: The important part of community-based games is the portal, where viewers can see who is currently playing, learn about scheduled tournaments and events, and join or invite others to play. It’s also the place where you compare high scores and earn bragging rights against your friends. We’re now starting to include these features in all our portals deployed on platforms with two-way connectivity, which includes satellite and cable, as well as the obvious IPTV platforms.

[itvt]: You mentioned earlier that you’ll be launching games on DISH Network? Can you say what that offering will consist of?

Tranter: We will shortly be launching a branded games service for a major network carried on DISH, similar to those I described earlier. We will be refreshing the games every couple of months in order to keep the service fresh and interesting.

[itvt]: You also mentioned that you’re working on a lot of "gambling-related interactive TV projects." What are those projects?

Tranter: We recently launched a sports gambling service on the Sky Italia system. In addition to the usual feature of being able to place wagers prior to an event, the service includes dynamic bets that can vary throughout an event as you watch, such as predicting the first goal scorer in a soccer game. A critical component of the system was being able synchronize closing the bet once the outcome was resolved, as well as making sure the bets presented on the service were kept up to date with the multiple bookmakers for the service, each managing a different Italian region.

More recently, as I just mentioned, NDS also launched a multiplayer poker game on BSkyB, which hooks into the single-account, multiplatform backoffice system from Orbis, an NDS subsidiary company. The service allows multiple players to play against one another whether they are accessing the service from an interactive application on the BSkyB set-top or via a Web site. The service extends the reach of interactive gaming across multiple platforms, but keeps the simplicity and security of a single account.

Part II: Xtreamplay

[itvt]: One gaming technology that you’ve been demo’ing at a number of shows over at least the past year or so is called Xtreamplay. Could you talk a little about that?

Tranter: It’s not available commercially now, though it’s complete and we’re demoing it. It’s going to part of our MediaHighway middleware, which is now in over 50 million set-top boxes. We’ll be launching it later this year.

Let me explain how it works: if you take any interactive application–be it a game or any sort of enhanced TV app–the interactive element is drawn by the set-top box, using graphics and images, which are, as it were, simply placed on top of the video. What that means is that, if you’re developing a game, you’re limited by the graphics capability, the palette, and the animation and processing capabilities of the set-top box–which are obviously going to be less powerful than most PC’s. So your games are going to be limited to some degree.

What Xtreamplay does is that it allows players to interact directly with the video layer–so with true-color, very rich, moving video. In other words, it takes the video and the on-screen display area–the interactive layer that sits above it–and allows the two to function in tandem.

Now this enables us to do a variety of things. One example would be a golf course game–which is something we demonstrated at CES. The way you would create such a game is that you would film a golf course from different angles–or just use rich 3D graphics like "Tiger Woods" and other console golf games do; then, on top of that, you’d have a sprite character–so a little guy with a golf club–which is drawn by the set-top box. Then, when the sprite character hits the ball, different video plays, depending on the direction the ball is going: so, if you hook the ball, the game will play a video going off to the left; if you slice it, it will play a video going off to the right; and if you get it dead center, it will show a video going down the center. The ball, of course, is also drawn by the set-top box. What we do is mark the video in production, so that players can interact with it: so, on the video of the fairway, you’d mark the trees and the bunkers, so that, if the ball hits the trees, it will rebound off them, or if it lands in the bunker, it won’t bounce any further forward. Obviously, allowing players to interact with the underlying video layer like this makes for a very rich experience. The possibilities are endless: for example, we have a first- person shoot-’em-up where you can fight with baddies who are all in the video–you can actually shoot them directly. So, if you succeed in shooting one of the baddies, we switch to a video of him falling down. Or if there’s an explosion, we switch to video of a real explosion, rather than just using on-screen graphics of an explosion. The Xtreamplay technology enables seamless, frame-accurate switches between the main video and alternate scenes such as this, which makes for a much richer experience than graphics from the on-screen display, and can also be used by the viewer to control the game’s narrative.

[itvt]: Did NDS develop this technology in-house?

Tranter: Yes. It utilizes video production tools for generating interactive areas within the video, and then converts this into a proprietary language, delivered as private data along with the broadcast MPEG video. In other words, as you encode a video, it includes descriptive language of elements of that video, which allow you to interact with it later on when it’s played out on the set-top box. The final element is a set of API extensions within the middleware to enable an application in the set-top box to interact with the masked video areas. Generally, interaction can take the form of a collision, such as a golf ball drawn in the OSD layer hitting a tree in the video and bouncing back, or an occlusion, which is where the OSD image is undrawn to give the appearance that it is going behind an object in the video, such as a golf ball drawn in the OSD layer passing behind a flag in the video.

[itvt]: When you roll out the new version of MediaHighway that contains Xtreamplay later this year, should we expect to see your customers, such as DirecTV, launch the kinds of video-rich games you describe shortly after that roll-out?

Tranter: I hope so, though there’s nothing in the pipeline today. Obviously, we have to get the middleware out there to seed the market. But I do think we’ll see operators taking advantage of it: that’s because it doesn’t only enable rich games, it also enables rich advertising and video-centric information portals. If you look at the Web today, you’ll see that there is a lot of very immersive, Flash-based advertising. Xtreamplay allows you to reproduce that kind of immersive experience on the set-top box.

[itvt]: Could you give an example of how Xtreamplay might be used to enhance an ad?

Tranter: One example we have demonstrated is of a car ad. The video is of a car rotating on the screen, which loops after each rotation. The viewer can then select parts of the car in which they are interested as they pass by, such as the interior, the alloy wheels, engine, etc. When selected, the video branches to a separate video, zooming in on the selected feature and displaying more information. In all, the total length of video is no more than about 45 seconds; however the resulting immersive experience is found to be between two and four minutes in duration. This type of ad would only be available off a DVR’s hard disk, but could be telescoped to from a regular 30-second spot, while the live video was paused.

[itvt]: Now Xtreamplay-based games are delivered to the DVR by simple broadcast, correct?

Tranter: Exactly. Delivery is through a straight broadcast, so it makes no difference what type of set-top box you’re delivering to, whether it’s satellite, cable, or IPTV. It’s basically similar to recording a regular TV program onto the DVR: an Xtreamplay game is a regular, encoded TV program that just happens to have a data element transmitted with it, and that just happens to be much more immersive. Now, you could push it down a separate pipe as well, if you needed to. But it’s still going to be played out of the set-top box in the same way that a regular program is–it just has this extra data that describes the areas of the video that you can interact with.

Actually, one of the nice things about Xtreamplay is that it makes it easy to constantly refresh and update a game. So, in the case of the golf game I was describing, the operator could update it with a new course every two weeks or so, so that you’d want to play it over and over again. It would probably take around half an hour to deliver the update, if you broadcast it at two in the morning. So you can easily see how you could have a subscription-based business model around this kind of constantly renewable game.

[itvt]: Do you see Xtreamplay has having any application for enhanced TV services?

Tranter: Well, the technology requires you to mark the video in advance, so you wouldn’t be able to use it to offer games that allowed viewers to play along with live video. On the other hand, an operator could push down, say, a course for the golf game that was based on whatever course a big ongoing tournament was playing on, so that viewers could watch Tiger Woods play a hole, and then pause the TV and play the same hole themselves on the game.

[itvt]: Do you have any thoughts on Xtreamplay’s applicability for scripted or unscripted programming in general?

Tranter: There are no reasons why you couldn’t provide a more immersive version of a program for viewers with an Xtreamplay-capable set-top box, that could be enjoyed from the hard disk rather than live. It would be a great way of providing alternate video scenes and letting the viewer control the storyline at certain points or branch off the video for a more in-depth explanation of, say, a cooking process, etc.

[itvt]: Do you have any plans to use Xtreamplay to enable 3D VR-type environments like Second Life on the set-top box?

Tranter: No. We believe in walking before we can run. Although that’s an interesting idea–I wouldn’t discount it completely.

[itvt]: What other Xtreamplay games have been developed?

Tranter: Well, we have a version of the golf game that we developed with a Canadian company called Electric~Spin and demo’d at CES. They have a golf training device that is comprised of a ball on a tether, and it’s connected to the set-top box via USB. So instead of using the remote control to play the game, you can use a real golf club and ball. You could actually use a number of different external devices to play Xtreamplay games in lieu of the remote control: we’ve developed   games with exercise bikes and kids’ toys, for example, and connected them either via USB or infrared. Allowing people to play interactive TV games with other devices than the remote can definitely make for a more immersive and interesting experience.

Other examples of Xtreamplay games include "Rogue Trooper," which is a game that’s been available for some time on the PC and which was developed by a company called Rebellion. We provided them with Xtreamplay Studio–which is our toolset for marking up the video–in part just to demonstrate how easy it is to use. They took their existing PC game, marked up its video, and developed an Xtreamplay version of it, which worked on an HD set-top. By the way, we’re certainly very interested in talking to any of your readers who are interested in using our toolset to create Xtreamplay content.

Part III: MediaHighway

[itvt]: Let’s talk about NDS’s middleware. What new features, in addition to Xtreamplay support, should we expect to see in the new version that’s scheduled to launch later this year?

Tranter: Most of the things we’ve been working on have to do with interoperability, so that’s where I think you’re going to see a lot of change with MediaHighway. We’re looking at adding more home networking capabilities to the middleware, as well as supporting more broadband capabilities in the set-top box. One example of this   improved interoperability is this ability–exemplified in the joint project we did with Electric~Spin–to support the connection of various peripherals to the set-top. We’re also making it a lot easier to move content from the DVR hard drive to other devices, such as personal media players. Our goal is to provide easy and seamless mobility of content, while maintaining the security of that content and the rights of the content owner in such a way that the viewer isn’t aware of being restricted in any way. We understand the importance of DRM–after all, it’s our core business. But at the same time, we understand that the viewer doesn’t want to be hampered by it. So we’re implementing things such as the Secure Video Processor (SVP) alliance, which is an open standard for passing digital rights between devices. If protected content is recorded onto a DVR equipped with our middleware, we can pass on the rights associated with that content to other devices you might transfer it to. Previously, if you’d recorded a pay-per-view movie to your DVR, you wouldn’t be able to move it from there onto another device; now you can do so because the "life" of that movie goes with it, ensuring you can access the content for the period of time that was purchased.

[itvt]: Any other new features you’ve recently added or are planning to add to MediaHighway?

Tranter: One new feature that was launched in the last version of the middleware and has now actually been implemented on a number of pay-TV platforms is our audience measurement system. We have a very robust way of measuring usage on the set-top box, and then securely storing and reporting back that data. Basically any specific type of button press that you want to track can be tracked: it’s down to the platform operator how this feature is utilized–and how they utilize it on each channel usually has to do with their specific carriage deals and legislation.

Among other things, it allows them to get very accurate measurement of interactive services. If you’re offering a game, for example, you want to know how many people are actually playing it; how often they’re playing it; whether they’re playing it after the second level or just stop at that point because it becomes too onerous; whether they’re skipping various features of the game; and so forth. Obviously, this kind of detailed measurement of viewers’ interactions with an application provides you with very useful feedback. So our middleware allows operators to collect all this data on the set-top box, and it’s usually encrypted and anonymized, depending on the privacy laws of each country, and then fed back via the return-path mechanism, such as a telephone line or a DOCSIS modem.

[itvt]: Which operators are taking advantage of this measurement feature?

Tranter: DirecTV are using it and BSkyB has been using it for at least a year now. They’re currently using it to measure usage in a focus group of about 20,000 set-tops, which I believe they are going to ramp up to about 100,000 homes. They’re using it to ascertain viewing of their programming, of course, and also to track things like how their guide is used and how their interactive applications are used.

[itvt]: Why are they only planning to measure usage in 100,000 households? Why not measure usage for their entire subscriber base?

Tranter: They could implement this for their entire audience base, but 100,000 is such a statistically significant sample that they don’t really need to.

[itvt]: Any other new features you’re adding to MediaHighway?

Tranter: Well, there’s distributed storage. The new version of the middleware will have the ability to identify an available hard disk on the home network, and store content on it, and then retrieve that content from it when required. That disk won’t have to be in a set-top box; it could be on a PC; it could be a network-addressable disk. So the end-user will have a storage potential that won’t be restricted to DVR’s and that will be easily and massively expandable. We implemented this feature because we believe home networks are the wave of the future, and people are going to want to move their content around the home network easily.

Another feature we’re working on–and we’ve actually only just proven the concept–is peer-to-peer file sharing: Let’s say that you missed "Heroes" this week, and you also forgot to record it–but that a lot of the other users on your cable or satellite network did record it. Our middleware would allow you to pull that content from those other users’ set-top boxes, rather than from a central server–the latter model obviously having legal issues associated it with it, as can be seen from what has happened with Cablevision’s Remote Storage DVR. This kind of implementation of peer-to-peer is an interesting concept–though I don’t yet know where we’re going to go with it. I really like the idea that each set-top box now becomes a mini-server for the whole network, which in turn becomes what you might call a community of content. So, anyway, we’ve demonstrated this concept, but again, it’s still very early days. We’re trying to seed the market, ascertain the degree of interest in it, and then see where the technology takes us. As a company, we’re actually very much into trying out technologies.

[itvt]: Could this peer-to-peer technology also be used for user-generated content?

Tranter: Absolutely. If you took pictures or videos of your kids, for example, you could then share them directly off your set-top box with other family members who happened to be subscribers of your cable or satellite provider.

[itvt]: Now, how would the rights of the content owners be protected if this kind of P2P functionality were implemented in your middleware?

Tranter: Well, the content would be encrypted, so that only people who are on the same broadcast platform–and who therefore have the same decryption system and the same rights to view that content–could actually make use of it. That’s the beauty of this: the rights to access that content would still be fully controlled. For example, somebody with a DirecTV set-top box could feasibly at some point in the future grab a program off a BSkyB box in the UK via the Internet–in fact, you could do that today on the PC. But the content would be encrypted for a BSkyB box, not a DirecTV box, so they couldn’t actually view it. At the end of the day, any feature like this has to maintain the rights of the content owner. Consequently, a content-owner might decide that they want to make their content available in this way–but for a fee.

Part IV: Synamedia Metro IPTV

[itvt]: Let’s talk about your IPTV platform, Synamedia Metro…

Tranter: Yes. Synamedia Metro is part of the overall NDS Broadband product offering. We use the Metro branding to identify the combination of two existing NDS technologies–NDS MediaHighway middleware and NDS VideoGuard CA/DRM–integrated into IPTV devices and networks.

[itvt]: I understand that you will shortly be unveiling a new version of your IPTV platform, correct?

Tranter: Yes. The new version is in trial now, and it’s going to be launched this summer on the SES AMERICOM IPTV distribution system, IP-PRIME, which uses Amino set-top boxes.

[itvt]: What do you claim are its advantages over competing IPTV platforms from Microsoft and the various IPTV technology providers that specialize in open standards-based IPTV?

Tranter: First of all, it’s a much more capable and open system, full of interesting features. Now all those features are on SES AMERICOM’s deployment roadmap, but, obviously, they’re not going to deploy everything on Day One. However, from the start, it will certainly give them a lot more flexibility when it comes to which set-top boxes they can use. This is because NDS designed the Metro middleware solution very much at the chip level in the set-top box. Therefore, as Metro is ported onto each system-on-a-chip (from Broadcom, ST, Sigma, TI, etc.), any set-top box using that SoC inherits the completed integration, thus eliminating long delays waiting for drivers.

Another major advantage of Metro is its centralized architecture. The issue we’ve seen with many of the other IPTV offerings on the market today is that IPTV operations, much like cable operations, are highly franchised. A single "telco" may actually be operating many telephone franchises in a widely distributed geographic region. So if the IPTV solution has a non-centralized architecture, then it is highly challenging to tie individual franchises together into a centrally managed hub/spoke architecture (which is often where the OSS/BSS systems reside–centrally). NDS Metro leverages the centralized architecture we built around our cable offering, so that we deploy a very small system at each telephone franchise, and the brains of the Metro system are located in only one centralized location that manages and regionalizes the edge locations. This is better for monitoring and control, for managing software downloads, for OSS/BSS integration and reconciliation, for hardware upgrades, and for a whole myriad of other aspects of day-to-day operations.

Metro will also give operators a very flexible user-interface experience, which can be easily customized with interactive features using an EPG customization toolkit; and it will provide them with a lot of interactive games and other applications; and–because it’s a very open platform–it will make it easy for third-parties to provide them with interactive applications as well.

One of its strengths will be home networking. We recently acquired a company, called Jungo, that specializes in residential gateway software. Their technology plays into our IPTV offering, by facilitating control of content coming into the home–not just to a set-top box, but also to a PC. The Jungo "OpenRG" residential gateway software can be installed in any off-the-shelf residential gateway, at which point uPnP technology is utilized to enable any device–such as PC’s, laptops, and other types of network storage devices–to enter the home network. Once on the home network through OpenRG, these devices are all mutually visible to one another. This extends the storage capacity of the DVR across the entire home. A non-DVR set-top box can use PC or laptop devices as DVR hard disks and record or play back content stored there just like a resident DVR. Given the pricing difference between a non-DVR and a DVR set-top box, you can immediately see the economic payback of using OpenRG in a Metro IPTV household.

So, I would say that our IPTV platform offers everything that Microsoft’s offers, but it’s a much more open environment–which means that it gives you much greater plug-and-play capabilities. You’re not beholden to one vendor. You can pick and choose the best of breed.

[itvt]: How are you ensuring that your IPTV middleware is scalable?

Tranter: Well, Microsoft and a few others put pretty much all of their software on a server in the headend. So it’s very much like you’re on a PC and you want to go to Amazon: you go and you pull the page off Amazon. The problem is, of course, that there are only so many people who can connect and pull pages at one time, so you end up with delays: somebody has to finish getting their page before you can get yours.

The way we solve this kind of physical limitation is by introducing caching in the set-top box through more of a thick-client approach. The centralized Metro headend will ingest, parse and regionalize all national and regional/affiliate EPG information and push it to the telephone franchises, where it will be cached for a configurable forward-looking time-period, such as seven days. Any information desired beyond the cached amount is pulled down by each set-top box in need on an on-demand basis. Thus the network is spared on-demand EPG requests for each and every set-top box. Therefore, most of the EPG information required by the majority of set-top boxes in the network is immediately available and locally cached, and the operator can more evenly manage their network infrastructure.

[itvt]: Can you give us an example of the relationship between multicast and unicast in your IPTV solution?

Tranter: Here’s an example: the common forward-looking data for a seven-day EPG is multicast to each telephone franchise in a regional format. The on-demand EPG information, for whatever isn’t already cast, is pulled using a unicast model. So we’ve taken advantage of the best of both worlds. VOD catalog information is pulled using unicast, as it is much more variable per household and per region than linear broadcast channel information.

[itvt]: What’s the significance of your recently announced IPTV partnership with Nortel?

Tranter: In many markets, the major factor in deployment success of an IPTV system is the presence of a skilled, end-to-end system integrator. This is especially true in large telco operations. Nortel is a key player in the global IP arena with respect to access line technology. NDS offers Nortel a full-featured suite of integrated technologies for IPTV middleware, CA/DRM, interactive applications and interactive infrastructure, plus home-networking, all under one roof, thus limiting the integration complexities inherent to multiple vendors. A key factor in our integration relationship with Nortel is the use of open standards, such as IMS. Nortel offers NDS a large and experienced access network provider with a global services organization that can take on projects of any size.

Part V: HD and the EPG

[itvt]: Let’s talk a little about some of the things NDS is doing in the HD space.

Tranter: Well, something that’s very interesting to us is that today the majority of set-top boxes out there are standard-definition, but over 50% of the television sets that are being sold now are high-definition. So what we believe is that, over the next couple of years, the take-up of high-definition set-top boxes is going to increase. Now, a high-definition set-top box is generally what we would call an "advanced" set-top. It usually has an Ethernet port, and it certainly has USB–so it’s network-friendly and interoperable, and end-users are therefore naturally going to want to use it to share content. So, for us, when we talk about high-definition being important, it’s not just because it provides a superior display. It’s because it will accelerate the uptake of advanced, interoperable set-top boxes, and will thus bring about a situation where viewers are increasingly moving and sharing content.

Now, the aspect of the shift to HD that interests us the most at the moment is the guide: one of our big pushes right now is developing a guide that will answer to the shifts in viewer behaviors that we believe the availability of advanced HD set-tops and other devices will bring about. So we have a major project underway, wherein we’re working with a number of content providers, platform operators and other industry players, in order to better understand what the EPG of the future is going to look like. We believe it will no longer be about providing a simple grid system, but rather about coming up with an interface that fits in with the viewer’s habits.

Up until now, the viewer who’s using the guide has always thought in terms of channels and times: "I want to watch this channel, at this time." We, of course, believe that this is changing: for the viewer, it’s going to be more about finding a particular piece of content and deciding what they want to do with it. So they’re going to be saying things like, "I want to watch this particular piece of content, regardless of when it’s broadcast or where it’s located," and "Now that I’ve found this piece of content, I want to bring it from my PC to my TV," and "This content is really good: I want to share it with my friends." We also believe that viewers are going to want to receive promotions for content that seems likely to interest them.

So, in short, we believe that the guide of the future will be multidimensional and multifunctional–not the one-dimensional, grid-type guide you see today. Now, in designing this kind of guide, one obviously takes advantage of high-definition: so you get more pixels, you get more information on the screen, and so on. But we’re also looking at some of the things that wow people about the rich guide experiences you get on, say, your PlayStation 3 or your Xbox. We felt it was time we introduced some of those things onto the TV. So we’re creating a master guide and master user-interface experience that runs on top of the middleware and allows you to get the most out of the set-top box–not just for viewing TV content, but for consuming content from elsewhere on your network, for moving content around and sharing it, and for interacting with it.

[itvt]: Could you be more specific about how this new EPG will differ from the grid-type EPG’s common today?

Tranter: One obvious thing we’re taking into account is that people don’t like reading TV. So there’s going to be a lot more promotional video. Now this isn’t just a matter of look-and-feel: there’s going to be a lot more content available to the viewer, which could be coming to them through a broadband pipe and various other places, as well as from satellite and cable. So promotion–especially from a trusted recommender–is going to be increasingly important, and you’re going to see a lot more promotional and recommendation-oriented content on guides. You’ll see things like barker channels, where a trusted source promotes programs to you, and where you have the ability to bookmark those programs–to press a button to schedule a recording of something that interests you when you see it on the barker channel. We also believe that search is going to be important, but we feel that effective promotion is going to be even more important.

Another thing we believe is that the guide of the future is not just going to be limited to a TV set. The same guide–or at least a variation of it–is going to be available on your PC, your mobile phone or your WiMax device. So, you might be sitting at your PC or using your mobile phone on the train, and see something you want to watch, and schedule a recording on your TV; or you might be watching TV, see a URL displayed that provides some information about a vacation that’s being advertised, and bookmark it so that you can view it on your PC later. If you have increasing interoperability between these different types of devices, the guide is going to have to take this into account. We also believe that guides are going to be increasingly personalizable.

[itvt]: When can we expect to see NDS launch these kinds of next-generation guides that you’re describing?

Tranter: Work on our next-generation guide is well underway. We will be launching the new guide at IBC in September, and we have our first customer scheduled for early 2008.

Part VI: "Project Rothko"

[itvt]: Now presumably all these advanced guide features that you’re developing are going to take up a fair amount of set-top processing power…

Tranter: Right. If you run all the features we believe are going to be necessary in future EPG’s on today’s set-tops, you’re basically going to have to run them in software, like an application on the PC. They’re not going to be in the chip itself–which means that you’re going to need more memory, more processing power and so on. Which, in turn, means that, if you’re not careful, they’re going to make the set-top box much more expensive.

So one thing we’re doing to address this problem is a project that we call "Rothko." It basically involves us feeding the requirements for the high-definition guides of the future into the chipsets themselves: we’re feeding them into OpenVG, which is an open standard that’s used by the chipset vendors, and we’re feeding them directly to the chipset manufacturers, such as Broadcom and STMicro. The outcome of the project is a written specification and an associated demonstration showing the features we believe will be required three or four years down the road, which we present to the guys who are designing the chipsets that are slated to be released in three to four years’ time. The goal is to get the chipset vendors to support these features by incorporating the necessary hardware accelerators into their chips–which in turn will enable these features to be supported without requiring very expensive set-top boxes. This is an ongoing project, and we’ve been doing it for something like two years now. Some of the suggestions we’ve been making are now being implemented in chips that are coming out today or that will be coming out shortly.

So the things we’ve been suggesting to the chipset vendors that they support down the road include different ways to manipulate and transform video as part of the UI experience, similar to what studio post-production systems can do today. But our suggestions don’t all have to do with the video display by any means. A lot of them have to   do with new ways of controlling your interactions with the TV that supersede the old-fashioned remote control. For example, we’ve started looking at the possibility of using a Wii controller to control your TV experience: flick your wrist to move up a channel, that sort of thing. I already mentioned how you can use an actual golf simulator to play our Xtreamplay golf game; we’re also looking at controlling games with exercise bikes, and at using your mobile phone as a remote control. Actually, we feel that the traditional remote control is one of the most restrictive parts of the interactive TV experience: a lot of people have been put off interacting, because of the difficulties of using the remote control to do so.

Of course, this is all somewhat speculative. We can’t say for sure that the things we’re suggesting are exactly what people are going to want in three or four years’ time. However, we work very hard to get people’s feedback as to the kinds of things they’re working on that will eventually need to be supported by the chipset vendors. It’s all about trying to anticipate where the industry will be in three or four years’ time, and then helping the industry move there, by ensuring that there will be cost-effective set-top boxes out there that support the kinds of things people are developing today.

[itvt]: What kinds of announcements should we expect to hear from NDS over the coming months?

Tranter: Well, as I mentioned, we recently announced the acquisition of Jungo, who provide a software solution for unifying the software and extending the features of residential gateway devices in the home. NDS are actively working on simplifying the interoperability, set-up and control of devices in the home to further improve the portability and enjoyment of content within the home ecosystem. You should expect to hear more announcements from NDS related to this endeavor.

URL: NDS

Originally Published: May 18, 2007 in [itvt] Issue 7.31

Click: http://www.itvt.com to subscribe to our free email newsletter, which contains all the news stories you see on this Web site, and additional breaking news and scoops, in-depth features, interviews, screenshots, videos, and other exclusive content you will not find anywhere else.