We said around March last year that Microsoft Mediaroom would lose its grip on leading Telco TV operations around the world, and a feeding frenzy for leading Telco contracts would ensue. So far the frenzy can be seen to be going on, but so far there is no feeding as such. It has just simply not materialized and there are reasons for this, mostly associated with caution ‘ nobody wants to screw up again and end up on a defunct, under developed platform that hits a wall.
Right now there are a handful of upgrade requirements for Mediaroom customers ‘ the biggest problem is targeting iOS and Android clients with some form of adaptive bit rate scheme over the open internet, although this area is being more and more clouded (excuse the pun). Many pay TV operators on other architectures began approaching this problem by either replicating their VoD servers or making it possible for existing servers to serve the open internet with the addition of transcoding to multiple bit rates, IP encapsulation, segmentation and sending it out as if it were common HTML using HLS, or one of the other sets of ABR protocols. This played well to the existing fad of using Apps on smartphones, tablets and TVs to playout the video but it leaves an almost separated strategy from the original TV service.
Since then an increasing number, including Comcast and Liberty Global, are embracing a dual approach, that of using transcoding inside the home gateway to offer the full linear content lineup in the home, plus VoD, and leaving this cloud based, but app dependent, version to run when people are away from the home, mostly offering just limited VoD. This is due to the way rights issues seem to take longer than technology issues.
The implication for set top vendors is to pull the market in two directions ‘ operators need larger and larger DVR capable Home Gateways, with processors about three times as fast, with transcoding onboard and this requirement has led to complex SoCs and the arrival into this market of Intel. The SoCs on the market now are faster partly due to the transcoding requirement, which is usually offloaded from the main CPU anyway, and partly due to the requirement to segment and stream the output, and also so the device can run Apps. Slave IP devices, attached over in-home IP services, usually WiFi, using DTCP-IP stream protection, with far simpler and cheaper set tops, have also popped up as part of this whole home architecture. But not on Mediaroom.
Finally the aggressive arrival of Netflix has made each pay TV operator have to rethink the way in which their originally small catalogs of VoD could swell to rival that of Netflix, and at the same time use recommendation software, from what had previously been an underused and virtually fallow recommendation software market. You need the recommendations to reduce the size of the VoD catalog you need to view in order to make a sensible choice.
The typical Mediaroom installation has taken a look at this and a year ago cried out a very public ‘help’ in the form of RFPs that listed all these requirements, but in a manner too vague to actually build a system around. In other words last year’s RFPs were RFIs ‘ requests for information, effectively a holding action until these major tier one operators could sense a pause in the rapid development pace of future multiscreen TV. Once they had a standing, rather than moving, target, they might actually do something about it. The Mediaroom 2.0 upgrade that came out in the Spring of 2010, was considered woefully inadequate for tier one operators and still did not support iOS devices or any smartphones.
We think that the target is not only still moving, but the pace of change is actually accelerating. It is particularly worrying that genuine Cloud TV is starting to reach the kind of maturity, where it looks like it can offer everything that the inordinately complex and clunky three part solution can. We call it a three part solution although you might say it is four or five part, because you need a new TV UI, that UI needs to be replicated or at least emulated on other TVs in the home, this same set of choices needs to be expressed, with the same content, on general purpose portable devices inside the home from the Home Gateway, as well as to Smart TVs, while something similar needs to come entirely from the cloud with less content (due to rights issues) to devices on the go. It’s this complexity that makes Cloud TV so compelling ‘ one service fits all on an exceptionally thin client basis.
As we move towards this divided, but hopefully not fragmented TV experience, it looks more and more like a revolutionary move away from Mediaroom will be just as attractive as perhaps an evolutionary one, which leaves Mediaroom in place and places new systems around the edge. But large Tier 1 telcos plan revolutions far more carefully than they do evolutions, hence the caution and inactivity.
There is an expression that we have heard from database administrators back in the early SQL development days ‘ it goes, ‘Two copies of the data equals two versions of the truth, which equals corruption.’
And in a similar way, any Mediaroom evolution strategy which insists that Mediaroom sits there unchanged and another system feeds in new capabilities to tablets and phones, is destined to end up with consumer satisfaction issues and a feeling that the system is corrupt or at least very different on different devices.
But the first thing you have to accept is that the Mediaroom systems out there, just take AT&T, Deutsche Telekom and Swisscom as examples, are still acquiring new customers and still taking market share away from every other form of TV ‘globally IPTV (in percentage terms) is growing faster than any other technology format ‘ ie. Cable, DTH and encrypted DTT, and as such it doesn’t need to do anything urgently. Or at least that’s the impression we are getting from telcos.
In fact the ONLY major telco that is being bold about its next generation video strategy is Telefonica, and that has headed TOWARDS Mediaroom, instead of heading away from it, but only on the understanding that Ericsson will make all the changes that it requires happen. But will it? The dilemmas of Telefonica ‘ with an IPTV system in Spain, IPTV and DTH services in parts of Latin America, with even a spattering of a decaying cable business ‘ are almost identical to Deutsche Telekom’s TV environments, and given the activity level of Ericsson on the account, which is handling the GVP (Global Video Project) for Telefonica, it is unlikely that Deutsche Telekom is looking any further afield than there.
We had always assumed that the plan was to use the Ericsson Multiscreen Solution immediately, alongside Mediaroom, as a first port of call. And that still must be the case for the most part, since it would take Ericsson a few years to integrate its multiscreen system with Mediaroom and perhaps longer and cost more for a simple replacement approach. That strategy would be a kind of surround and slowly replace strategy ‘ freeze the existing clients, and add and configure the entire Ericsson IPTV solution, including Multiscreen, and then sell all new clients the new system and systematically replace Mediaroom, perhaps over three years, replacing set tops as you go.
That strategy may now be in doubt purely because Ericsson will shortly complete the purchase of Mediaroom, sometime in the second half of 2013 and until it sees the code and assesses it, it can’t really make conclusive statements about its future roadmap ‘ for instance it might keep parts of the Mediaroom system alive, who knows ‘ although engineering NIH (Not Invented Here) will probably make that unlikely.
There are two or three new scraps of data that have emerged this week which gives us some hints as to where and when this will all be resolved ‘ and our spies out there in the pay TV community tell us that surprisingly that the company most likely to break the deadlock first is AT&T, due to make hard decisions by the end of this year, whether or not it throws Mediaroom out, or works around it.
That means that the decisions that Deutsche Telekom had hinted it would make in 2012, with an OTT system expected to be launched then, will have been overtaken. We made a great fuss about the Liberty Global Horizon project being two years late, but Mediaroom customers were closer to 3 years late then they adopted it and may be later still, as they move away from the technology.
What happened is that Deutsche came up with a plan that it was testing, and then Ericsson bought Mediaroom and it has had to wait until the dust settles ‘ both in terms of a new possible roadmap that this might give Ericsson, not where it just surrounds Mediaroom, but where it makes radical changes to it. It could make it embrace new set tops and servers, not just CE and Windows based devices, and it could complete and open up the nascent TV Everywhere strategy to Apple devices.
We would argue that the problem is larger than even this. Recently Cisco stood up and said at a recent conference, in the form of Nick Thexton, VP and Chief Technology Officer, of the Service Provider Video Technology group, that ‘Cloud TV is more or less ready,’ suggesting that it would guide companies like AT&T, were it to ask it, to a vision of future TV that is cloud based. We understand that Cisco has just licensed the UI technology of ActiveVideo.
Talking about ActiveVideo, in one of its recent white papers it claimed that its Cloud TV solution was more than 80% cheaper in total cost of ownership terms, than traditional pay TV technologies. Basically it makes the savings by either leaving in the set tops you already have, or throwing them out entirely.
This then should be the next generation of pay TV for both AT&T and Deutsche given that the long term benefits of a cloud strategy would give them huge price elasticity when up against cable, and a chance to accelerate their TV customer base through pricing advantages ‘ if this white paper was true. But it comes perhaps a tiny bit too early, at a time when cynicism about Cloud approaches is still at its height, and no major installations have been built out to millions, rather than thousands of customers.
We have reported before that Deutsche had always planned to go with a new system based on Huawei network system upgrades and an Ericsson CDN and middleware, initially this was seen as a strategy of surrounding Mediaroom or bringing a ‘lite’ version of IPTV to new regions, with the extra-ordinary step of bringing any Apps to the screen. This part of the strategy was actually announced, using a special version of Accedo’s App Store software, adapted for the Microsoft Mediaroom server driven architecture. This was done back in September 2012 by writing apps in Visual Studio and delivering them wrapped in Mediaroom Presentation Framework, and using them to control how the CE set top presents it on the TV screen.
Presumably the trialed version of Ericsson’s roadmap at DT would have involved all of its existing partners such as Edgeware for its VoD appliances around the CDN, Verimatrix for security of the IP portion of the traffic, but as we say, perhaps some revamped set top options so it is not tied purely to Arris/Motorola.
But it is worth noting that Deutsche is not just a 2 million home German installation for Mediaroom. It has patches of Mediaroom in its other territories and one of the best ways of testing the water is to put the proposed system to work in a less sophisticated, smaller market.
In total Deutsche Telekom has about out of 4.6 million pay TV customers in Europe, a number that we have in the past projected to grow to around 8.3 million by 2017, which by then would make it the 4th or 5th largest pay TV operator in Europe. It has systems in Romania, Croatia, Hungary, Macedonia, Montenegro and Greece as well as Slovakia, using every possible pay TV technology, so it has a key requirement to also embrace OTT.
Three weeks ago security player Verimatrix announced that it would provide its VCAS conditional access system to Magyar Telekom, a subsidiary of Deutsche Telekom, and that this was to secure multi-network and multi-screen TV services.
Magyar is one of the Deutsche Tekeom subsidiaries along with incumbents in Montenegro, Macedonia and Slovakia, which have installed Mediaroom and it is possible that Deutsche Telekom wanted to use this as a ‘trial’ solution for the rest of its European base.
We understand however that here the systems integrator was Huawei, not Ericsson, and that this is very much a surround system ‘ add multiscreen over the internet without it being very much to do with the core TV service. In fact the announcement suggests that this is as much a partner to the DTH service, rather than specifically for the IPTV service. We’re still unsure if this system can allow the service to be sold over the top of other cable-based broadband lines, although we know the programming can be delivered out of home, over both 3G and WiFi hotspots, so theoretically it could be used as a Trojan horse to get into rival pay TV environments.
Deutsche has about an even number of clients on each of IPTV, Cable and satellite in Hungary, adding up to a total of 844,000. The system is based on HLS streaming and can be viewed on PCs, Macs, iPhones, iPads, Smart TVs and Android-based phones and tablets.
Verimatrix, which issued the announcement, is one of the many companies that has a ‘framework’ agreement with Deutsche Telekom, a contract that sets pricing and call down procedures, without actually stating that it will be chosen for any given installation. If it continues to make incursions into the Deutsche territories, this could end up as one of Verimatrix’s largest clients. As we understand it the cable and satellite systems mostly use Conax, and so this should be seen as a design win against the Norwegian provider, and against Microsoft, which was the conditional access provider for the IPTV service.
The other deal announced this week perhaps has nothing to do with Deutsche’s IPTV plans, but it has taken on board some brand new technology from Danish firm Octoshape and will offer a new service called LiveStream Perform, powered by this technology. It will offer this service through its ‘Business Development and Innovation’ arm, which doesn’t suggest that it is core to the main business.
Instead of using TCP/IP this system uses UDP as the core communications protocol and Octoshape also offers some patented ‘throughput optimization’ technologies, which we assume are a combination of the use of robust UDP, allied to something like Forward Error Correction, so that when any packets are lost the end point re-calculates what they should have been instead of having the TCP protocol back off and send data slower, requiring a change to a slower bit rate version of the video. This way it claims that broadcast quality video can get through a normal every day DSL broadband line in consistently high quality, and without using adaptive bit rates at all.
Octoshape is swimming upstream with this technology, but it doesn’t mean it is wrong, and it says that adaptive bitrate technologies are unable to sustain video quality, due to packet loss, congestion and other factors that reduce the available throughput. If at any time the available bandwidth drops, adaptive bitrate technologies lower the video quality to avoid buffering. As a result the video quality and playback viewing experience continuously changes, something that viewers are known to hate.
We think that this is really being pushed as a new service that Deutsche Telekom can offer to the German broadcasters ‘ a reliable, and scalable way to send video of broadcast quality ‘ we assume that at the initialization of the video there is some process to set the right picture resolution for the particular device, and it remains set at precisely that. Many German broadcasters have already gone on record saying that it is too expensive to put their TV channels on radio broadcasting networks and they are planning to send them instead over the internet.
The two companies say that the LiveStream Perform system offers ‘dial-tone’ reliability of broadcast TV today through best-effort broadband networks and this can reach global audiences. Octoshape also has its own multicast technologies which are advanced versions of the original multicast services that are invoked by the IGMP (Internet Group Management Protocol) which are the basis of managed IPTV. There has suddenly been a new burst of activity around Multicast, adding new protocols so that if a particular router or group of routers is not multicast enabled, this no longer blocks video sends. Octoshape (as well as the majors like Cisco) are pioneers of this new IP multicast market.
If this is genuinely a revolutionary video experience that German broadcasters, are happy to use, then perhaps Deutsche Telekom might ‘eat its own dog food’ and offer its own multiscreen services in what is a completely different manner from the rest of the global telco community, dropping adaptive bit rates entirely.
One reason that this is not such a stupid idea is that this flat bit rate approach is far easier for cellular networks, like T-Mobile, to accommodate when it arrives at the edge of their networks. ABR services notoriously inflate data rates to take up all available cellular bandwidth, which is why video optimization techniques have to be used to flatten the streams, costing cellular operators cash and heartache in the process and leaving them to fall foul of net neutrality accusations.