Original URL: http://www.theregister.co.uk/2005/04/08/ant_set_top_browser/

ANT - trying not to be tiny in a big IPTV world

ANT music

By Faultline

Posted in Developer, 8th April 2005 10:08 GMT

analysis Any company calling itself ANT is bound to have to put up with comments about its size, but despite its dominance of the IP set top browser market, the company cite very small.

This is best illustrated by the fact that last year the market for IPTV was, according to sales and marketing director, Stephen Reeder, only 1m devices.

At maybe $2 on average per browser installed, that’s just $2m in browser sales, and given that ANT wracked up 720,000 of these sales, a neat 72 per cent of the market, it must have made under $1.5m in revenue on browser sales.

In fact ANT’s entire revenues for 2004 were only revealed a few days after it went public last month, at £2.2m ($4.1m) and growing at roughly 25 per cent per annum since last year. Perhaps ANT is aptly named since it is usually other companies that get the credit for the simple, but important job it carries out in the IPTV food chain.

ANT began some 11 years ago and was really a lifestyle business for a few UK technologists for many years, but when a browser project for Bush Internet TV came up, in 1999, the company took a turn into its current direction of writing TV browsers.

After that a Spanish TV project Querio opened up the secret to the marketplace. Reeder told us, “Querio had out to tender a multisourced set top supply and selected a license straight from us. It told Thomson and some other manufacturers that they must take their browser license from us.”

Al fresco

Today that browser focus still brings in most of ANT’s revenue. It is the application in an IP TV set top that captures every keystroke (from the remote control usually) and interprets it for action either locally or for remote action. It talks to the head end and takes instructions from it.

ANT’s original browser is called Fresco and comes in two versions, one for the IPTV market and Fresco Lite, for other types of digital TV, such as cable and terrestrial.

Around 2m units Fresco have been deployed since the beginning of 1998, with over 700,000 shipped in 2004 It’s new browser, Galio was launched at the end of 2004 and is currently in testing. This will deliver what ANT calls a next generation user interface, which includes dynamic, animated and blended information overlays, suitable for a range of TV programs and for interactive services.

ANT says Galio is suitable for set tops and personal video recorders and was developed to take advantage of the latest W3C Internet standards and to reduce server load and bandwidth requirements. Galio will let operators develop more visually sophisticated services and already 11 set top manufacturers have taken the technology onboard. ANT also offers digital media management software that lets con-sumers to view, manage and share their personal media including photographs, videos and music on the set top.

“Our browser is written in C,” says Reeder, “and today it works pretty much on any operating system and on any silicon.” “Most set tops run some variant of Linux today, I’d say about 80 per cent of them,” says Reeder, “but there are other operating systems to deal with, Nucleus, PSOS, VXWorks and we even adapt or browser to run directly on the Bios of a DSP chip, for instance the Texas Instruments DM642.” It takes between four – six weeks to move the ANT products to another environment.

The attraction in using a DSP chip to run the browser, is that the same chip can be used to decode the MPEG or other codec stream as well, potentially reducing costs. “If you put the work all on the DSP, you don’t need a master processing chip,” explains Reeder. In fact the entire process of building a set top is all about driving down price. For IPTV, the set top is the main piece of customer premises equipment (CPE) and in every kind of network service operator it is the CPE that drives up the service price, because it is a contingent cost that has to be paid once again, every time a new client is connected to the service.

Whereas the margins for operating say a cellular network, are so high that cellular operators can afford to subsidize their equivalent, with handsets that cost in the region of $400-$600, the operators for IPTV services are working on thinner margins and can’t always afford to do this.

The commonly targeted price point currently for set tops is around $100, and the DSP element of this can be anything from $17 to $25. If a service operator has a separate codec chip this can be cheaper, around $7 to $10 for an MPEG2 version, but then the browser needs a processing chip to run on, and that can take the cost back up. Each set top is a trade off. If an operator is worried that content is going to be easier to take in one format or another, the decision may have to include support for multiple codecs.

“It’s cheaper to buy a chip that just handles H.264 natively,” says Reeder, but the price goes up if you need it to handle both VC1 (derived from the Microsoft Video Codec) as well as H.264 (this standard was worked on by the CCITT and H.264 adheres to its nam- ing principles, but it was also created by MPEG and is also called MPEG 4 Level 10 Advanced Video Coding).

“There’s a Conexant chip coming that has the H.264 codec built into its hardware that’s will probably start at around $15 and go down towards $10,”says Reeder.

The importance of using these more modern codecs cannot be overestimated because they can halve the bandwidth required to run an IPTV service. There is not much to choose between these two codecs, because in principle they handle compression in similar ways. This is evidenced by the fact that the licensing authority, MPEG LA has had to agree that at least 12 companies have contributed essential patents required to make the VC1 codec work (a standard version of Microsoft’s codec that is being ratified by the Society of Motion Picture and Television Engineers). This means that virtually the same companies that will receive technology royalties from H.264 will also get the royalties that Microsoft had hoped to keep for itself.

This fact is only recently discovered and now Microsoft is arguing with the other 11 consumer electronics companies, thought to include Sony, Toshiba, Matsushita and Philips, over just how to license this technology. The wrong licensing terms and Microsoft will suddenly find that its own use of the codec in Windows Media Player, will suddenly mean that even Microsoft cannot afford to “give” the player away. The net result could mean that the standard is abandoned. “A market with only one new codec to worry about would make set tops easier to build,” agreed Reeder.

This affects ANT, because if more complex chips are needed in set tops this competes for the total budget of the set top. ANT’s software doesn’t handle any of this codec work but Reeder says that the browser is the toughest part of the programming within a set top.

“The other jobs are decoding, which is managed by a middleware client and the real time streaming protocol stack, which operates below the browser level and a media player of some sort. The browser is the glue that brings all of this together and it is the most complicated piece of software in the set top. We have 140 man years of development in our browser.”

But in another guise Microsoft is trying to muscle in on ANT’s market and threatens to take it out in one fell swoop.

Hidden interfaces

The approach to menuing and handling remote control commands used by Microsoft in its forthcoming IPTV will do away with the browser all together. Microsoft has been working with Alcatel, an existing licensee of ANT’s, towards IPTV. “The Microsoft approach is an entirely closed solution. They don’t use browser technology in their set top design, in direct contravention to how everybody else implements the technology. “In effect, Microsoft is building a set of integrated applications which are all proprietary and which talk to each other using .NET,” says Reeder This will mean that no external software can be written for it without Microsoft’s blessing, because all of the interfaces are hidden.

“This won’t work in IPTV, and Microsoft is a little behind the thinking here. The Telcos will not buy into a closed solution, and Alcatel is trying, with some success to soften Microsoft’s approach and embrace other people’s products. The way IPTV has built up is that it relies on open networking so that everything can be multi-sourced.

“We understand that Thomson now has two products lines that do exactly the same thing. One for $600 to $700 with the Microsoft software on it, and one for $100 to $150 that uses our technology.”

One of the biggest Telcos in the US, SBC Communications has already pledged itself to use the Microsoft IPTV software when it comes available, as it rolls out its ambitious “Project Lightspeed,” costing $4bn to $6bn in the US.

But the software is thought to be up to 18 months away from completion and although Microsoft already has a code base, it is not expected to be robust enough for production use for some time. SBC also has an interim IPTV set up using 2Wire set tops, and some analysts have suggested that it will never deliver the Microsoft version because this interim step will prove too popular.

“When operators build services they want to see what their service is going to look like on the screen and so they develop the look and feel using our browsers, and then they know it’s available on virtually any set top they later want to buy,” said Reeder. At the moment ANT’s biggest clients are coming from the Far East with Chunghwa Telecom’s IPTV pilots in Taiwan, and others in China, each expected to absorb as many as 500,000 units during 2005. “The Chinese operators are expecting to soon be installing IPTV in the low 10 millions, and that’s what we are really excited about,” says Reeder.

Even that number, as IPTV grows exponentially, will only make ANT something like a $50m revenue company, but with a cost base of just 42 staff, there’s plenty of margin in the business for it and its value could be considerably higher.

Copyright © 2005, Faultline

Faultline is published by Rethink Research, a London-based publishing and consulting firm. This weekly newsletter is an assessment of the impact of the week's events in the world of digital media. Faultline is where media meets technology. Subscription details here.