Don't hold your breath for online TV

There is a Doonesbury cartoon that dryly observes that a sudden increase in YAP!.com’s site traffic meant that the "demand for tiny, jerky videos that never play has nearly doubled".

Sad to say, this is an accurate reflection of the state of video content on the Web – it simply is not an attractive proposition and is really only watched for its novelty value.

The only exception to this, unsurprisingly, seems to be pornographic videos, for which resolution is not such an important issue as suggestion is. Even in these, however, the dominant (transmission) model seems to be to provide short mpeg excerpts for download and subsequent replay than to provide access to live streaming.

So what happened to the much-hyped promise of live TV via the Internet? Where is the broadband access to live footage of sporting events? And could the lack of this capability explain the recent retreat of Rupert Murdoch and his cohort away from Web publication to the relative safety of interactive broadcast TV?

Allow me to suggest very strongly that it does and, further, that there will be very little consumable activity in the direction of Internet-delivered TV programming or movies for quite a few years yet.

A short time ago there was a flurry of stories crowing that newspaper-based Web sites seemed to be far more popular than their TV-based equivalents.

These stories derided TV sites for their poor preparation, lack of commitment and resources, and general level of involvement and enthusiasm for the new medium. They ignored the almost total absence of profitability among newspaper-based Web sites.

Since then, the headlines have been dominated by stories of layoffs, staff reductions and a supposed drive among on-line publishers towards profitability. Lachlan Murdoch was reported as attributing these down-sizings to a lack of faith "in the economic model that’s driving these stand-alone, advertising web sites".

But the writers who wrote off the TV-based web sites missed the most important point: the Internet is not yet ready for video. What all of these "driving towards profitability" endeavours have in common is that their sites rely almost exclusively on text-based content. Not a single video-streaming or live broadcast site to be seen.

So, what is this utopia of Internet-delivered TV programming? Why couldn’t it work and why isn’t it working? What’s not to like about this idea?

The vision is simple and eminently desirable to consumers:

I turn on my TV-style watching device and, in about the same time it currently takes for my TV to warm up (preferably less time), I have access to all the Internet, including a back catalog of (depending on subscription) every movie, TV show, video clip or CD ever made. Using, say, an infra-red mouse device, I can relax on the sofa and click on one of them, say episode 42 of "I Love Lucy" and it appears – full size or whatever proportion of the screen I deem appropriate – on my viewing device. While watching it, I can pause at a certain spot and call up the biography of a guest actor who appeared on that show, because I’ve never heard of her before but apparently she was highly regarded in that period and she’s very pretty. The biography contains links to a catalogue of her career. I can use the same device to call up my friend and invite her over to watch an all-night replay of the actress’s entire career this weekend. And so on.

Sports fans can substitute the latest event of your choice, complete with player profiles, historical data and records, etc. Current-events freaks and political animals substitute the news bulletin, debate or party conference of your calling. It would be great if everyone in the household could have their own such device, so that dad can watch the footy, mum a girl movie and junior can find out how to make an A-bomb in between on-line classes.

Unfortunately, the best current version of this for most of us, even using the broadest-band connections, is a small, poor-resolution screen that appears surrounded by logos on a device that takes an eternity to start up – if you can get satisfactory access to the Web regularly.

So, what is the problem with on-line TV and when can we expect our services to improve?

Notwithstanding the argument that people don’t want to use their TVs for surfing the Web, the answer is one of simple maths: the best Internet delivery services available are simply nowhere near fast enough to cope with the services described above.

A cable-TV line and modem can cope with a maximum of about 2M/second, which has to be split between the number of users attached to any cable. So if you have three devices in the house and all want to use them all once, you get 666K/second each. DSL lines can handle about 1.5M/sec (9M/sec if you include the asynchronous download only).

That sounds pretty damn fast until you consider that a single stream of TV-quality digital video footage requires a transmission of about 50M/sec – more than five times faster than the fastest option available. Delivery of these services over a wire network is simply not feasible.

Of course, there is always Gilder’s Law, which states that available bandwidth will treble every 18 months. At that rate, if we assume that a current broadband user can expect a 1M connection, then in 6 years (4 x 18 months), that same user should be able to use the Internet to access a single broadcast-quality video signal with enough spare bandwidth to request accompanying text information or other views. It will be a lot longer before everyone on their block can access the same quality of the signal of their choice.

In the meantime, the ‘net will remain the province of text-based sites, and Murdoch and Co will seek their fortunes via satellite-borne Interactive-TV loops with PVR devices. Unsatisfactory as these loops are, I, like other content creators, publishers and consumers, just have to wait patiently for the day when "I Love Lucy" will replace tiny, jerky videos that never play.