Videonet - news, blogs and analysis for the pay-tv industry

• 08-Oct-2009 by Benjamin Schwarz
• Post IBC Analysis: CAS and Middleware
• IBC 2009 News

Why not two pieces might you ask? Fallacious answers would include that not enough has happened in this field since last year to justify two pieces or that I don’t get paid enough for writing this, both of which are not false. But if the truth be told, this part of the industry is merging, so this piece is a really taste of the future. Enjoy!

Take, for example, a company like the UFO Russian start-up called NetUP. I say UFO because at least on my radar they really did appear out of nowhere and do not fit into any known category. Their new ‘IPTV Combine 4x’ illustrates what I mean about this space converging. NetUP, who incidentally could do with some help naming their products, are pushing a single $14.5k box that can do everything from the head end through to middleware to VoD and conditional access. Konstantin Emelyanov told me that their own CAS is not included in this price and that they can integrate with any other CAS vendor. The single unit box is not targeted at Telcos but more at the hospitality market. I would worry that like SAP in the early days of business process re-engineering, the enticing price tag of the product could be largely offset by the cost of consultants to get it up and running. Maybe someone out there can let me know.

There’s a bunch of other companies I met which are not of the UFO class but don’t really fit into any category easily. I spent some time with Anevia who are one such oddball. They provide various components for IPTV without being either middleware or CAS vendors.

Anevia’s video servers can enable delinearization services like start-over TV and pausing live TV. Their specificity in this crowded market is a mobile focus. They demoed a three screen version of the “follow-me” demo where you start watching a football match on one device, say the sitting room TV, then pause for a pee. But once in the loo you just can’t wait so you carry on watching on a PMP. Men get killed by their wives by the end of this use case because a few drops will always end up on the seat when watching football.

Just like NetUP, Anevia is also focussing on the hospitality market. When I asked them about the future, I was impressed not to get the usual PR mumbo-jumbo. Pierre-Yves Leberre told me Anevia actually believe IMS can make a difference maybe even by next year’s IBC. Well that’s a gutsy prediction so I look forward to our rendezvous next year to see if you’re right, (BTW yet another Ericsson IMS demo doesn’t mean you’ll have been right).

I also passed by the Orca stand at IBC. In a clear demarcation from their parent CAS company Viaccess the booth was a lot fun especially with beer on tap. The link between these companies is perhaps the best example of the merge I opened this piece with.
Orca’s ‘Compass’, a recommendation interface demo, was on show as usual. The ‘Compass’ interface lets operators plug different recommendation engines simultaneously to deliver a compelling content navigation experience. Having a potential killer app is good, even great. But showing it again and again kind of turns it into a commodity.

The less impressive-to-see but much more significant demo at the Orca booth was of the Soft@home STB middleware. Soft@home is a joint venture between Thomson, Sagem and Orange - who so happens to be Viaccess’s parent company also (yes it’s even confusing for the M&A bankers). Dubai based Etisalat is part of the picture but it’s not clear to me what stake they actually hold in the joint venture.

Soft@home CPE software runs on both STB’s and residential gateways (read modem-router if you prefer). The idea is to bring to the IPTV world what OpenTV or NDS brought to the DVB world in terms of de-facto standards and “portability”. CPE middleware has always been about solving portability problems. So although there isn’t all that much to see yet, the potential impact could be enormous.

Detractors would say that this initiative is fixing the previous decades problems. High-end chipsets will bring standardisation around Linux or/and Windows at the OS level. At the same time major open standards like OpenGL or SVG (mentioned below) are making inroads alongside proprietary ones like Flash or Silverlight. All this will create a world where solving the low-level portability from one device to another should get easier, dispensing with the need for CPE middleware.

But if these detractors are wrong or if Soft@home delivers value high enough up the value chain with real functionality not just software glue, they could well end up king of the connected IP-home one day. A bit like roulette; I don’t think they have many chances of winning but if they do, it’s going to be big-time.

Dreampark

I was surprised to meet a middleware company who as far as I know do … middleware and nothing else. Like with all theories, the premise that middleware alone no longer works needs a good counter-example to prove the point. This is it.

Dreampark is a vibrant small Swedish company with 23 installations representing 150k subscribers in total. They have a few hybrid deployments with DVB-T and have just started with the rarer DVB-C hybrid variant.

It was also refreshing to meet with a company that don’t say “yes we do” to all my questions. Dreampark have taken some clear technical decisions. They are a browser based middleware company that has taken a gamble on Scalable Vector Graphics or SVG to display impressive graphics without using up too much bandwidth or processing power. That means there’s no Flash player so if you’re an Adobe addict, move on. SVG is available in most browsers except Webkit, but the demos all used the eKioh browser, which Dreampark told me have the best SVG rendering at the moment.

Dreampark have designed their core product in two parts. The SVG based front end is basically a user interface that is now separated from the heavy lifting IT centric back-end. The idea is that you can use an existing back-end with its legacy user databases, OSS/BSS stuff and plug in their sexy new interface.

The SVG portal generator was the most impressive demo new to Dreampark’s booth at this year’s IBC. The idea was to create a tool for the marketing department who use a GUI environment to move things around on the interface. When they are happy with the layout they just press a button and the new interface goes live on a staging server to be tested on real boxes by the same people. Although this was dreamed-up for smaller operators where small teams do-it-all, the Tier 1’s I know could greatly benefit from Dreampark’s approach too.
When confronted with the 2010 question I asked all my interviewees this year, Dreampark pointed to 3D that was the entire buzz on the trade floor. They also expect to bring some demos next year with recommendation solutions, which they believe are key to growing ARPU.

Verimatrix

Very much like in the Matrix movie, talking to Steve Christian at IBC was disorienting. Steve could easily have been a marketing VP from another reality. I’d come to hear about the latest trends in content protection and he was talking of seamlessness and business logic using the three-screen experience as an example. I wasn’t expecting either to hear about getting the best of the Internet on the TV set without the pitfalls.

This Verimatrix vision I was being presented was not dissimilar to what I’ve reported from the like of EchoStarx, Pacex or Motorolax.
But rest assured, the industry hasn’t yet gone completely crazy and Verimatrix isn’t about to build boxes. Steve is still in our universe.
I tried to provoke him on what happened to downloadable music with MP3 and that even Apple was getting rid of DRM. Wasn’t this going to be the way for video too? Steve remained nonplussed pointing to the fact that video is no way as versatile as music; i.e. consumed in a much smaller number of ways with a much shorter life span. The barriers to usage that DRM created with music must and can be avoided. Verimatrix may not be able to describe the right model but hey can point to many many wrong ones.

The newly formed Digital Entertainment Content Ecosystem (DECE) consortium will try and steer the industry away from the pitfalls music fell into. He concluded that larger file formats would still require protection for the foreseeable future. After I spoke with an executive from a film distribution company, all I can say to that is ‘Amen’.

So after pioneering the card-less or software based approach to conditional access Verimatrix’ contribution to the story is now about maintaining the business logic where ad-supported Hulu models meet the walled-garden pay-tv world.

Verimatrix also want to play a role in keeping the user experience seamless, which brings us back to the CAS-Middleware merger.

Verimatrix used the Telia Sonera case to illustrate much of what was said. With 500k IPTV users they are extending the experience into the mobile world as well as Wifi in the home with a seamless handover from one network technology to the other. I was frustrated not to see any of this in demonstrations. Maybe none were on show because the boundaries in this space, although moving, are still recognizable so that CAS vendor booths like Verimatrix' are still more about CAS than anything else. CAS, however important in the value chain, has always been pretty boring to demo. But as the blurring will continue inexorably, I look forward to some exciting CAS vendor booths in 2010!

Otherwise Steve’s vision of what might be on show at IBC 2010 from Verimatrix included the ubiquitous TV-everywhere and more interestingly a kind of “meta DRM”. “Multi-rights” will take into account the user, the file, the device, the time and maybe the location too. I really will make sure to book a slot next year, but this time I’ll ask to see the demos before the chat.

NDS

Jonathan Beavon talked me through the content-aware widgets that were part of a totally new interface where everything is contextual. This was the demo to see if you only saw one at the show.....and I missed it! Jonathon was kind enough to give me a catch-up session over the phone. One of the demos showed a twitter feed synchronised with a TV program. Another one showed a widget to get extra Internet based content while watching a nature movie; the widget would bring other animal documentary content with the press of one button.

Jonathan argued that it is necessary to ‘see’ the system from end-to-end to implement this kind of feature. That’s a bit of an NDS sales rationale because in this case I’m sure you could get away with just having access to the EPG metadata.

Another demo similar to the Dreampark demo described above was of the new EPG framework for MediaHighway customers. It lets them modify their java based EPG user interface on their own without needing a new release by NDS. It also can enable the introduction of ads into the EPG, but I’m not sure how well that will go down with subscribers. With most existing EPG’s this would just be a little money saving gadget, but if you take the longer view where EPG’s become sophisticated content portals, then having control and a fast turnaround will become vital to remain competitive.

In a totally different direction, opening towards new markets, NDS announced a pre-paid CA System. It’s surprising that this approach, which has been so instrumental in the mobile phone market has yet to make a blip on TV content’s radar screen. Pay-per-view, pay-per-time, pay-per-channel, pay-per-bouquet or pay-per-subscription sounds very much like what I heard from other CAS vendors. There’s a race to get the mother of all DRM/CA’s.

When I pressed Jonathan about the future of CPE middleware he calmly answered that the number of new customers NDS is seeing was all the reassurance he needed for the now.

He jested about the “give me just Linux and a browser and I’ll get an STB ready in no time” attitude, that first arose back in 1999. There is plenty of hindsight and yes; a Linux-box-with-a-browser can indeed do most of what is wanted. But according to Jonathon, operators will typically spend five years re-implementing everything they need (that of course comes “out-of-the-box” with a middleware like NDS’). He went on to add that some things end up being almost impossible without middleware like, for example, synchronising video with advertising, or controlling the hardware inside the STB to implement DVR efficiently.

Another line of defence Jonathan used when I kept up the onslaught on CPE-middleware’s uncertain future was to point out that some things do indeed need to be done at the head end. Aha, so CPE middleware isn’t the only truth after all. Unsurprisingly NDS has a lot of work going on at the “head-end” like, for example, the recommendation engine demo they showed at IBC. Like Orca, NDS believe their job is to let operators’ present different recommendation engines to subscribers depending on a multitude of parameters. At IBC the 4 engines under the NDS hood were:

•    The Israeli start-up Jini incidentally founded by ex-Orca CMO
•    The video indexing trailblazer, Blinkx,
•    Predictive analytics technology company from Scotland Think Analytics,
•    Finally NDS’ own editorial recommendation tool.

That last one is going to be first on the list of real world deployments because you must flog the high margin videos rather than blockbusters that are subsidised or those assets you haven’t reached the minimum guarantees for yet.

When I pressed Jonathon on how to decide where to put functionality; i.e. in the CPE-middleware of as he diplomatically says “in the head end” we started talking about philosophy. Operators with a broadband culture have a “head end philosophy”. I suppose NDS don’t want to call that by its real IPTV-middleware name because that would confuse everyone. Companies with a Broadcasters’ view of the world like to put functionality in the STB. As some of the larger IPTV deployments go beyond the million-subscriber mark, scalability & QoE issues are also changing this debate from philosophy to architecture. Companies that don’t control the networks they use are faced with a specific challenge in mixing these two paradigms.

NDS furthermore expect broadcast to continue well into the future. It is still the most efficient way to deliver linear TV. I must say that early feedback from the first 50k FTTH subscribers in France will not contradict him yet. So NDS see On-Demand making big inroads into “one-way systems that are here for many many years yet”. Jonathan spoke of the On-Demand world integrating into the broadcast world with a hybrid approach. But he by no means sees broadcast being replaced.

At home I just received my personal Canal+ cube. It has they sleekest of interfaces and is jam-packed with features. But alas it is more of a mule than a racehorse :o) I suggested to Jonathon that NDS’s heavy middleware might bear some responsibility in the slowness. He gave me the obvious answer that Pace, the STB maker and Canal+ made their own price/performance decision. A faster chipset would and will bring a faster box. I’m still a novice journalist and I slapped my wrist later for not having the repartee to ask him why some boxes like Netgem’s FNAC box (Fetch TV in the UK) pack as many features, with a same generation chipset, and yet are much nippier. Answers on a postcode readers.

We wound our discussion down by revisiting the question on convergence of CAS and IPTV middleware. NDS hasn’t made any decisive moves in this space (like Viaccess’ acquisition of Orca) because they believe that the pure IPTV market is very thin. All the low-hanging fruit have been picked. The big bucks in the next few years will come from hybrid projects and Over-The-Top. I understand that their “head end” work will let NDS pick up some smaller fruit along the way. The market has already started to focus more and more on smaller projects as T2s and T3s from both the broadcast and broadband paradigms see convergence within their reach.

NDS portrayed to me the image of a CPE-Middleware company who so happen to own a major CAS solution and incidentally dabble in some “head end” functionality. I’m impressed at how they’ve avoided blurring their boundaries contrary to the rest of us. The contrast with Verimatrix was striking in the end.

I don’t think NDS will buy their own “Orca” because even if it’s not officially road mapped, they’re gradually building one.

Back to previous page