‘The problems with patients is that they always lie,’ according to Gregory House, the doctor in Fox Network’s hospital drama, played by British comic actor Hugh Laurie.
And in this drama they always begin by lying to themselves, which is why we were reminded of House, while walking around the core pavilions within IBC, the International Broadcasting Conference, held this week and last in Amsterdam, in the Netherlands.
There is a distinct aura of self denial and a fair amount of placing of heads deeply into buckets of sand. The IPTV community faces a crisis, partly brought about by Microsoft and Alcatel’s dominance of the major Tier 1 telco accounts, but also partly of their own creation.
Last year the IPTV contingent yelled the loudest of any at IBC, but there was definitely a feeling, just like in the Lord of the Rings, that dark forces were circling, as Microsoft IPTV threw a pall over proceedings.
Faultline asked if a year on there would be mosaic interfaces, rapid channel change and High Definition IPTV to match the Microsoft claims. None were in evidence.
‘We’ll deliver those features when Microsoft delivers them,’ said one middleware company, and much of the talk among the executives behind the 100 or so start ups that drive the concept of IPTV, is still of Microsoft and Alcatel.
‘We understand that it will end in an ugly divorce between Alcatel and Microsoft,’ said Yosi Glick, VP marketing and business development at Israeli middleware outfit Orca, predicting that Alcatel will allow its OMP middleware, which at least works, to re-emerge. ‘Once Alcatel has bought Lucent, it no longer needs Microsoft quite so much and it has the problem of what to do with the Imagenio middleware that it has taken on from Telefonica. It will merge it with OMP.’
He extends the theory further, saying that NDS will become the new DRM for the system, and that Microsoft Windows Media DRM will be out in the cold.
Like most self delusion there’s some truth in what he says. There are cracks in the relationship between Alcatel and Microsoft, but divorce is out of the question, at least in the short term. The most that anyone can expect is that Alcatel will become somewhat promiscuous (to extend the marriage analogy) and share its post-Lucent merger favors around a little more, with Microsoft its favorite, but the door not exactly slammed on either OMP or Imagenio. Like most married couples they have responsibilities.
Throughout the IBC show the age old statements that the Microsoft software won’t scale and that unicast channel change is a mistake, continue to abound. Perhaps we would grudgingly accept that High Definition rapid channel change is going to be something of a challenge. Codec designers at the show reminded us that H.264 not only requires the delivery of the immediately preceding I-Frame, but of multiple past and potentially future I-Frames, before a multicast stream can be rendered, and that when the sums are done on high definition video, this can be extremely heavy task for dedicated channel change servers.
Perhaps it won’t scale. Perhaps the entire Microsoft system is, as was rumored at the show, going into a third rewrite. But all of these problems still weigh heavily on the IBC IPTV contingent, more than they weigh on Microsoft.
Orca CEO Haggai Barel was more circumspect, ‘The shadow of Microsoft is a big problem for us, no question. We have changed our focus from tier 1 telcos to tier 2 and 3. Tier one sales require a higher sales expense, you have to go through a separate IT department, a procurement division, fit in with existing policies, and there is more after sales support. We have to accept that this is not the best way to spend Orca’s money. It takes 1 year to 1.5 years to crack a tier one account.’
But Barel was deflated by what happened at Lucent, the company, that until its merger with Alcatel, was going to carry out the bulk of Orca’s systems integration work and take the tiny specialist into tier one accounts. ‘Orca was in advanced negotiations with Telefonica Brazil before the Imagenio deal was announced, and now in Latin America we are still in bids with Lucent, but not in any of the Telefonica properties.’
‘In Europe we are involved in some deals led by Accenture, at tier 2 accounts, which take more like 6 to 9 months to win. Some of the companies we were talking to with Lucent, we are now talking to with Accenture.’
But he is insistent about the features that are pushed by Microsoft, ‘The set top builders all tell me that there is no set top that will be made in the near future that that will carry HD, run on Windows CE and not cost north of $500. They tell me it is not possible and I believe them.’
Perhaps the most telling point that Barel made was that none of his clients are scaling the way he would like. ‘This year’s revenue was 75% made up of new business. My existing customers need to begin rolling out volumes in order for Orca to thrive. I can’t keep finding 75% of revenues fresh each year.’
It will be of scant interest to his investors if Orca fails and then Microsoft has to scale back its ambition slightly, or raise its set top prices. So will Orca sell up? Mostly Barel shrugs and agrees, ‘Collectively all the IPTV companies messed it up, and we gave this market to Microsoft” and there he trails off. He sees the entire community’s future in hanging onto the coattails of companies like Nortel and Accenture, but not everyone can try that. For a start there are too many of them.
That strategy is really limited to perhaps Orca and just Myrio and Kasenna, who continue to target tier one telcos in Europe, Myrio because it is now owned by Siemens, and encouraged by its successes at Belgacom and KPN, it thinks it can afford to target very large accounts, and Kasenna through the Ericsson deal. But will either of them be strong? Nokia has merged its networks division with Siemens, so will it even continue in IPTV, and Ericsson has yet to cut its first deal.
This year Microsoft never even took a stand at the show, the only major absentee, an indication that it did not want to remain a part of an eco-system that it sees as fixed rigidly in the past, hostile and antagonistic, and perhaps it doesn’t want a deluge of bad publicity from stalled and cancelled projects. It appears to be open knowledge at the show that Swisscom has cancelled its deal using Microsoft software and that Telecom Italia will no longer use any part of the Microsoft strategy going forwards. That still leaves it the best account list in the world.
The truth is that there are too many suppliers in the IPTV space and most of them are small companies. It has proved impossible for them to mimic the feature set claims of the big companies and just staying alive is becoming a full time occupation. So much for the predictions that IPTV will steal some 50 million pay TV customers away from cable and satellite TV over the next five years, it may instead just disappear, or at least get bogged down.
So what’s holding it back and what are all the companies at the show doing about it?
There are three reasons we’ve comes across. Integration, integration and integration.
One exec put it this way, ‘You say that the DRM, the set top, the middleware and the VoD servers are all integrated and then you try to take the entire set up, from four or five different suppliers, to a new customer. Suddenly it all stops working.
‘It’s not that there are 60 set top makers and 10 VoD server players, it’s that there are multiple releases of each of them. No two set tops are the same, and we are always tweaking our software. There are no standards out there.’
So all the answers we came across fall into three categories and all three were on show at IBC. IPTV players can support less integrations and co-market in clusters with the ones they really do support; they can try to create a standard set top, a standard for middleware; or they can move their offerings sideways, a VoD server company or a DRM specialist can offer their own middleware, and be sure they integrate.
Andy Mathieson, a director at Irish DRM specialist Latens told us, ‘The trouble is that we’ve all been doing systems integration work for free for the past three years, and it’s killing the market. We have to offer best of breed, but we also have to get systems integration costs right down.
‘There are 1400 small incumbent operators in the US who represent 10 million subscribers. What we’ve done is carry out a tight integration with Minerva middleware, and go after these customers in that combination. Those customers can’t afford Microsoft’s system and they can’t afford expensive systems integration work, so we tell them what we can offer and if they want to ask us to integrate some new DVR or a VoD server which we don’t know very well, we tell them what the integration will cost. If we have to we’ll just walk away from the business, but not everybody does.’
He points out that the other companies which are chasing a solution to the problem by adding middleware, are just making the problem worse. At the show Kasenna, SeaChange and others all expanded from VoD into middleware, while newcomers have also emerged such as Dreampark, which has managed to kick out Orca at Norway’s iVision.
‘There are now 30 middleware packages out there. The number is not going down, but up. Middleware is just a term for software that looks after anything the customer can interact with, so that’s the customer devices, programming assets, billing, service information, on screen messaging, electronic program guides, content ingest, customer provisioning.
‘As a DRM company if I try to integrate with all of them it would be suicide. There are customers out there that say they want 3 different types of set tops, and expect to order 1,000 of each and for us to do the DRM integration on them all for free. I wouldn’t even be able to make enough money to pay for the invoicing process out of a deal like that.’
Mathieson is even more scathing about companies that want to offer IPTV services that connect to PCs, either playing content on the PC or taking downloaded content from the PC. ‘When a service call arrives at an IPTV operator the help desk can make them switch the set top on and off and if it doesn’t work, send someone round with a new one. What happens when a service interoperates with a PC is that you start doing PC support. There are different video cards, different versions of middleware for the PC, and many of the hardware features of set tops that are running in software on a PC. We will just stick with a cookie cutter approach and try to drive our customers up into volumes.’
One company that is doing just that, opening a link to the PC on its latest release is Kasenna. After years of being a VoD player, suddenly it has a partnership with Ericsson and a spanking new version of its ‘PortalTV, version 2.0, which it says has taken it into the IPTV middleware business.
‘Ericsson is our way into Tier 1 accounts,’ says new CEO Kumar Shah, stating that he already has 10 operators in his sights with the new offering. PortalTV was launched in July last year and the new version offers content management, integration with internet video its own EPG and integration with Amino and Wegener set tops and has successes at Cavalier Telecom in the US, Sasktel in Canada, Jazztel in Spain, TV Sierre in Switzerland, and claims that it has deals in both China and India. It also claims that it has about 50% of the VoD deal at FastWeb in Italy, a fact that incumbent VoD supplier BitBand hotly disputes, saying that it is closer to 80:20 in its favor.
‘All the tier one accounts are not yet tied up. In the Asia Pacific region, although customers pay prices some 50% to 60% of US prices, the market is still set to be larger in revenue terms than the US or Europe,’ said Shah.
So the Kasenna money is on the name of Ericsson opening doors, and taking it to the far east, a place where previous IPTV efforts have virtually all been pieced together with best of breed hardware, with the operator itself acting as the systems integrator.
And Kasenna seems to be ignoring the fact that there have been some spectacular failures in trying to take the web into the living room, but its current campaign is entitled ‘Where television meets the internet,’ and somehow we think that consumer habits haven’t changed enough to warrant this being a success.
The other approach we saw was all about creating new standards. ‘Watch this space,’ says Simon Woodward, CEO of IPTV browser company ANT, ‘we’ll be trying to create an open standard in the IPTV browser market.’
He was talking to Faultline about Galio, its one year old upgrade to its old Fresco browser, ‘The most important thing about Galio is that it allows multiple applications to run on a single instance of the browser. This means we’ve added application threading so that the browser can act as an abstraction layer and manage more apps.’ Woodward talks about removing all of the complexity of a hardware environment from the IPTV development programmer.
By ANT offering a consistent API, he reckons that his company can take the grief out of all the different set top implementations. ‘At some point this industry will be brought to its knees by the complexity of tying all of those applications together,’ he said.
‘Why should an IPTV developer have to know how to change channel on different set tops such as Amino, Kreatel and ADB, which use varying Javascript extensions and different syntaxes.’
Of course Woodward is trying to increase the importance of his company seen purely as a browser supplier, that adds just a couple of dollars to the cost of a set top. But his point is fixed in the same observations that others had at the show, complexity and integration is grinding the industry down.
Linux specialist Softier was launching its own set top at the show, but was also after something similar. ‘We are all small teams, ANT, Oregan, Espial, and us. We put millions of lines of software code into homes and that gives us all a software headache,’ said Haim Behor, VP of marketing at Softier.
‘The set top is approaching the complexity of a PC and it’s time that it was fully standardized. The DVB based pay TV services such as cable support just 100,000 lines of code, and the devices are dumb. If we try to put all the intelligence into IPTV set tops, it has to be a standard, not all the companies can each write their own versions,’ said Behor.
Softier is the company that first put a copy of Linux on a Texas Instruments DSP chip, an achievement that has led to many devices, such as set tops, tying together the benefits of Linux and DSPs into one.
That was 2 years ago, launched at IBC also, and now the company hopes to begin an open source campaign to begin standardizing the software layers in a set top. Not just starting at the browser and moving down to a hardware abstraction layer, as ANT said, but throughout the software.
‘Open source software ends up more stable, better tested, and it needs to be componentized, and with clear APIs. The operating systems should expect standard APIs, standard protocols should be used,’ he said and called for any company interested in helping out to contact Softier straight away.
But that’s the true enigma of this market, it’s unlikely that anyone will contact Softier, or ANT or any other company that needs help and co-operation, because any co-operation in this market is hard won. We noticed around the show that Kasenna’s effort to launch its own middleware was greeted by scorn by all the middleware and VoD players. They would be out in the cold on any Ericsson deal, so they would never integrate to the Kasenna VoD system if they were a middleware company and never use the Kasenna middleware if they had a VoD offering.
The schisms in the IPTV market are likely to remain long after rampant consolidation has set into the market and our guess is that IBC can safely book a smaller hall for IPTV next year and not so many of the suppliers will still be around in 12 months, especially if they keep lying to themselves.