Video on Demand
Video-on-Demand 'The Killer Application' ? December 2001 - January 2002 By Phil Sabin EVP Operations and Production Europe Online SA

For Video-on-Demand (VOD) to be the killer application that drives new entertainment platforms it must be just that - 'On Demand'.

Delivering near VOD (nVOD) services is a bit like shooting wide of the post - a miss is as good as a mile. On demand means I get it now, when I want it, not when it suits your server capability. I am sure lots of nVOD suppliers have discussions centred on 'how long do we make the customer wait.' The truth is it really does not matter - if it is 10, five or two minutes, it is too long. On Demand requires instant results, this is only possible with local storage on set top boxes.

The Kagan Broadband summit seemed to have great bias toward server-based solutions to VOD. I believe this is to do with tradition, not the best solution to the problem. If you need to get material from point A to point B, and have it available instantly at point B whenever required, it makes sense to store it at point B. Content suppliers are used to holding the content and deciding when to pass it onto the customer, the control is with them. The new world is more customer focused, more like the Internet - give the customer a wide choice of content from which to make a selection, pass the control onto them. As with anything we feel in control of, we feel more comfortable using it.

Giving customers more control doesn't mean they have to put more effort into their entertainment choice, it must be organised in such a way that is accessible and relevant to them. This is simply not possible if you are trying to serve the material centrally. The customisation and breadth of content will put an impossible strain on any delivery mechanism. This makes set top boxes the only realistic method of delivering true VOD.

Two common arguments against set top box storage are: -
It makes the box too expensive, and
You can't store enough content.

Currently set top boxes store about 100 hours of content, which amounts to a storage cost of about 100 dollars. We predict that within the next two years the cost will have significantly reduced and the capacity will have quadrupled giving 400 hours of content for people to watch truly on demand. That is about three months viewing for the average person without having a major impact on the price of the set top box (at least no more significant than that caused by MHP [Multimedia Home Platform]).

The great thing about this solution is that it is truly scalable. With any cable system you will run into a multiplication of content hours multiplied by the number of viewers. That means whatever system you design, if it is successful eventually the network will not cope.

So how do we deliver to the set top storage? The answer is that it does not really matter. Whether content comes via satellite, cable or even DSL, set top storage will make the best of whatever bandwidth available on whichever medium. Obviously speed gives more flexibility and can deliver a greater variety of content so technologies with higher transfer rates are better but even a lower speed constant trickle has great potential to update stored data.

Currently satellite companies offer the best delivery method. They are already delivering content directly to people's homes, doing away with the need for an expensive, terrestrial 'last-mile' connection. Through hybrid terrestrial/satellite technology, they can deliver 1.2 Mbps streaming as well as 2 Mbps background downloads to set top boxes across Europe. This gives users a playback image quality that is better than SuperVHS and the capability for VOD and 'Entertainment-on-Demand.'

The service can be delivered to everyone within the satellite footprint with no extra cost implications and no degrading of quality or speed. The vast amount of data that can be delivered via satellite means that a set top box can contain a constantly updated entertainment archive allowing true EOD.

For cable networks satellites can deliver to a cable headend anywhere in the footprint, be it analogue or digital, where a simple transmodulator will use one cable 'channel' to deliver an enhanced digital EOD service to the end user. There is no need for costly network or server upgrade and this option gives a quick way to sell an upgraded service and collect more monthly revenue; furthermore it is totally scalable.

VOD is the killer app now but as peoples' awareness of the technology changes EOD will dominate. VOD is something people understand and slips easily into their views on entertainment and technology, as they get more used to the service they will adopt more and more of the on-demand functionality.

Eventually we will expect to watch all our visual entertainment 'on demand' be it sports, news, serials, soaps. After all it is what we've come to expect - how popular would the telephone be if you had to schedule calls?