‘Japan is signing up mobile TV customers at a rate of 2 million a month,’ said Alan Moskowitz, a senior MobiTV executive in a catch up with the company this week. Alan is director, Office of the CTO, at MobiTV, having previously been in charge of Alliances at the leading mobile video platform player.
But his mind, and the mind of the company, are clearly no longer solely considering the 3 million customers MobiTV has managed to carve out in partnership with customers like AT&T and Sprint, offering an aggregation of TV channels, using cellular streaming technology.
Instead it is firmly on the problem with how to make the same thing happen in the US and the rest of the world, as has happened in Japan, and we spent the next 40 minutes talking about little else.
The Japan Electronics and Information Technology Industries Association (JEITA), published data in July that said that the month of July alone had witnessed the shipment of 1.6 million devices, but quarterly average is closer to around 1.3 million. The total back in July was 11.7 million 1 Seg ISDB-T handsets, and at that rate (assuming no acceleration and no inventory build up ‘ either of which is possible) Japan will have over 15 million devices in circulation today and perhaps 18 million by year’s end.
In our own report published at the beginning of the year when there were just 3.4 million 1 Seg devices out there, we were projecting that Japan would be on just 6 million devices by now, under half the actual figure.
That’s more handsets shipped in 3 months than MobiTV has signed in subscriptions since it first offered a Java downloaded video player, and began signing up mobile video customers back in 2003.
We can see MobiTV is excited because it hopes to be an arms dealer both in terms of technology and the content, and the Free to air approach that has been pioneered in Japan and Korea, would mean many more mobile TV units shipped, and more fees for Moskowitz’s company.
‘The ATSC M/H standard will be the US equivalent of 1 seg, and it sits on top of the 8VSB (Vestigial Side Band) modulation method used in ATSC. The standard should be in place in time for technology suppliers to build products and ship them to use with the 700 MHz channels that the US will auction next year, for use in 2009,’ said Moskowitz, but he admitted there was still a lot of work to get through.
‘Japan has proved the business model. The war is over. It’s now up to companies like us to start convincing operators to go down that route. We know that Nokia has 6 of its top guys on the ATSC M/H (Mobile handheld) standards committee, and they have pushed for the intermediate stack to be DVB-H compliant. And why wouldn’t it be? That way the same software client could manage a DVB-H network, ATSC M/H channels and DVB-SH satellite channels.’
His view is that paid TV services just cannot prop up mobile TV. There needs to be a free component, as there was in broadcast free to air TV, to drag the audiences to the concept, and then 10% to 20% of services will go paid. ‘If any handset you bought already had mobile TV on it, and you didn’t have to pay to use it, you’d use it. Then the cellular operators would have a base to upsell premium with services.’
But how will any of the broadcasters convince US cellular operators that it is in their interests to go along with this when there is no ARPU in it, especially as both Verizon and AT&T have already chosen to back MediaFLO as their mobile TV standard? That’s 50% of the handset population of the US already spoken for isn’t it?
What the idea needs is a pebble, or better still a rolling boulder, to start the avalanche. ‘We think that will come from the MVNO community in the US. Virgin Mobile has 5 million subscribers, and if suddenly they were all walking around watching TV it would be enough to start a landslide. Or the 2.7 million at Leap Wireless might do the trick. Virgin is certainly media oriented and unlike the big US cellcos, they aren’t going to build anything of their own, so they might jump at a chance to get into mobile TV through a different route.’
Moskowitz pictures 300 million Americans virtually overnight having access to existing broadcast networks, and then cellular operators providing the equivalent of cable TV on top with another technology. But is it really that simple? The same experience was gone through in Korea, where the government mandated a free to air business model for T-DMB for new spectrum that it offered. Meanwhile a private consortium went and built a satellite technology (called S-DMB), which went and got almost a half a million subscriber head start and which today boasts over 1 million customers. But because it was free to air the T-DMB service has rocketed to 5 million customers by mid 2007.
However this is all about government mandates. Japan and Korea’s regulators made a ruling, and their control over their cellcos is perhaps stronger than the FCCs control over US operators. We could not imagine the FCC on the one hand selling 700 MHz spectrum with open access clauses, and at the same time dictating not just a technology, but a business model for some of the other spectrum, and getting away with it.
If MobiTV and companies like Harris, LG Electronics and Samsung convince broadcasting operators to make the relatively conservative investment of a few $100,000 to upgrade their transmitters to the new standard (a figure we got from Harris) then they are in a classic chicken and egg argument. If they put out a mobile TV signal within their market area, then who will make the device to receive that signal? Or an MVNO like Virgin might decide to invest in some devices, and find that their customers can only use them in 20 or 30 of the 210 US Designated Market Areas (areas where the same signal is broadcast). This is somewhat like the roll out of HD radio in the US, which has been a painful, lengthy process. Perhaps such an outcome could be avoided if a company like Nokia underwrote the upgrades for a key number of US TV markets, or promised to make handsets available. It might be just the type of move that Nokia needs to build its market share in handsets in the US.
Perhaps a major like Sprint might merge the idea with its forthcoming WiMAX network and offer two levels of TV, a broadcast version using ATSC M/H and a WiMAX unicast, on the same devices.
But no-one in the WIMAX community has given much thought to how best to offer video across a WiMAX network, and most of them feel that for the first 18 months of any network, a unicast TV signal, one which is competing with voice for spectrum, is all that is needed. Sprint has certainly not said how it plans to go about offering mobile TV, only that it will. MobiTV must know those plans since it is the software and content supplier for Sprint, but Moskowitz chose not to share those plans with us.
When we were speaking at TelcoTV a few weeks back our own Faultline principal analyst had been asked to answer the question ‘Is mobile TV a sprint or a marathon,’ and we pointed out that no-one has yet seen the ATSC mobile technologies because they are being built especially for the US, nor the DVB-T2 specification, expected later this year, which is likely to be adopted in some 70 or so countries in the next three to four years. Both should accommodate a second mobile TV signal although the DVB-T2 technology has not yet been selected.
If either of those technologies take off, we are in not only in for a marathon battle between mobile TV standards, but we are likely to be seeing the birth of one of the most convoluted, and intricate entertainment markets ever conceived, with multiple layers of competing technology all on the same device.
First off we must understand that mobile TV has around 13 technologies currently chasing the space, not counting cellular streaming, which is the basis of the MobiTV products to date, and streaming over an open internet model, which is where everyone hopes this business may one day end up. Both of those approaches detract from voice usage, and require too much cellular bandwidth and therefore cannot dominate until the creation of 4G networks and perhaps not even then.
All the other technologies fall into four main areas, they either posit that 1) someone has to build an entirely new radio network and put an extra radio in any receiver, or 2) they aim to take some limited part of the cellular spectrum and devote it to delivering video in a broadcast or multicast sent from existing base stations; or 3) they take an existing network like digital radio or satellite radio networks and piggy back a video signal on the existing network or 4) they anticipate existing TV broadcasters will add a mobile signal in with their existing Standard Definition TV broadcasts.
DVB-H and MediaFLO (and DVB-SH a satellite version of DVB-H) are all examples of the first category and all of the news focus has been on these to date, but the price of entry into this type of market is high and US build out will be in the order of $500 million for MediaFLO on top of around $300 million getting the technology to market. The DVB-H effort is substantially behind this in the US, though launched elsewhere, with similar costs and only AT&T now has spectrum for another thrust at DVB-H, since it bought two 6MHz channels from Aloha Partners for $2.5 billion a few weeks back.
Against a backdrop of apathy towards mobile TV in the US (though clearly not elsewhere), with unconfirmed reports that the Verizon MediaFLO service, VCast TV, has only shipped in the low tens of thousands, this all seems to invite the other approaches to the party.
So instead of seeing the whitewash by Qualcomm’s MediaFLO right across the US, we may see a shift to a free model and a free model is perhaps best supported by a different technology, though MediaFLO could just as easily go free to air, but it would need a whole new advertising platform and a way of extracting advertising revenues in order to do that.
With option 2) Cellcos in the US don’t have a spare 5 or 6 MHz of spectrum lying around to commit to mobile TV from their existing cellular spectrum, unlike some European and Asian operators that may bring a multicast to the market using their allocated one way TDD spectrum.
To try 3) using an existing network, you must be the owner of it, and so we know that once XM and Sirius merge, they will be ready with advanced video services, but that these will need a large antenna to be received and this is really an in-car market in the main.
Which leaves the last approach 4) that of bundling a mobile TV signal in with the SD signal.
And this is exactly what ISDB-T is, breaking a 6 MHz channel into 13 segments, with enough data throughput to cram a mobile signal in a single segment, with the terrestrial TV channels taking the rest.
But the key thing is perhaps NOT the technology, it is the business model, and in the end if the pay TV approach isn’t working, Verizon and AT&T will have hands on experience of it and find themselves back in the market for a new approach.
If Nokia indeed wins the fight to blend the upper software layers of DVB-H into the ATSC standard, it won’t necessarily help unless a DVB-H network actually gets built in the US, and if those MediaFLO reports of slow uptake are true, then who is going to risk another $500 million building a network for DVB-H. We have speculated that AT&T will, based on the Aloha spectrum, but if MediaFLO is not successful that may undermine that decision.
Perhaps the better marriage here is with DVB-SH. Alcatel is working with McCaw company ICO Communications to deliver satellite mobile TV from the sky, but most radio engineers turn their noses up and suggest that this signal won’t get through walls in when using 2 GHz. But they get through walls when in a cellular configuration with ground based base stations close together, and Alcatel has always suggested that with some spectral diversity (multiple antennas) devices can be built that need a repeater on just 50% of the base station sites in cities, and that the satellite is fine outside cities.
Another scenario, with a common software stack with ATSC M/H, would be taking input from two radios, one in 700 MHz and another in 2 GHz and routing them to the same application thereafter.
We have long suggested that mobile TV will become a blended service with perhaps a bulk free to air delivery, some premium channels, some VoD cellular streaming and even a handset DVR, all blended through a single EPG, and a single service front end, emulating the existing complexity that exists in the layers of fixed TV services.
But the US market could be far more convoluted than this, with handset marriages between MediaFLO and ATSC, between DVB-H and ATSC, between EV-DO delivered unicasts and a WiMAX multicast, between a cellular multicast using MBMS, some unicast and perhaps DVB-H ‘ all of these combinations and any others you can think of, working together more or less off the same chips, with multiple antenna and radios ‘ but appearing to the customer as either a single TV service or as competing services.
That latter likelihood is much worse. Multiple interfaces, from multiple suppliers, needing different handsets to watch different qualities of TV channels, with different levels of advertising ‘ in other words a mess.
Maybe it would be good for Cellcos to avoid all of that and simply abdicate the mobile TV space after all and let broadcasters make the first move and go back the Moskowitz vision.
Meanwhile Harris and LG Electronics and its US subsidiary Zenith, which invented much of the ATSC technology in the first place are pushing their own version of ATSC MH called MPH in-band mobile digital television at the ATSC standards setting process as a Candidate Standard but then relying on other subsystem candidates for compressed audio and surround-sound from AMR-WB+ and HE-AAC-v2, and video using H.264 MPEG4 and offers different transport layer options of either the MPEG Transport Stream or pure IP.
Back in June the ATSC received 8 replies to its RFP for the standard, from Coding Technologies, Coherent Logix, DTS, the Mobile DTV Alliance, Micronas Semiconductor, Nokia, Thomson and Qualcomm. So far we haven’t heard anything about what the Qualcomm submission was suggesting, perhaps an alliance between the ATSC standard and the software stack on MediaFLO?