NBC Teletext

~ the obit ~

Every analog television picture has a gap in it. The electron beam finishes the bottom of the frame and has to travel back up to the top, and during that trip it transmits nothing you can see. Broadcasters worked out in the 1970s that you could put data in that gap, and Britain had been doing it since 1974 with Ceefax. NBC started trials in Los Angeles in 1981, got FCC approval, and began regular transmissions of NBC Teletext on 16 may 1983, on the same day as a similar service from CBS.

It was a magazine of about fifty pages. News and weather, with sports and whatever was on that evening, all of it cycling past continuously so that a page you asked for arrived within a few seconds. NBC's version was built on NABTS, the North American Broadcast Teletext Specification, developed by AT&T with Canada's Communications Research Centre out of the Telidon system. NABTS was the sophisticated option and it looked it, capable of proper weather maps and recognizable movie posters where the European standard drew blocks.

The problem was on the other end of the signal. NABTS decoders cost hundreds of dollars and there was no consumer box on the market when the service launched, because NBC was waiting for television manufacturers to build the capability into sets and the manufacturers were waiting for the audience to exist. Broadcast executives had talked about twenty-five dollar decoders. Nothing near that ever appeared. In 1984 NBC was still projecting profitability by 1987 and ten percent household penetration by 1990. It shut the service down in january 1985.

I was a year old when this ended and it was a different continent, so there is nothing personal in it at all. The number I keep looking at is the decoder. When I got online in 1998, the receiving end cost me a modem, because the machine doing the decoding was a computer already in the house for other reasons and the format was text that any program could parse. NABTS required a dedicated box that did one job for one broadcaster and nothing else. A data service is only ever as cheap as the hardware at the far end of it, and this one took twenty months to prove that.

~ the rap sheet ~

~ how it landed ~

The economics were lopsided from the start. AT&T quoted roughly two million dollars to equip a station to originate NABTS. A station in Salt Lake City ran a World System Teletext setup, the British standard, for about forty thousand. The cheaper standard was less capable and it worked, which in broadcasting is usually the end of the discussion, but the American networks had picked the other one and had a national rollout attached to it.

Meanwhile Zenith was making WST decoders for a rival service in Cincinnati and selling almost none of them, to the point where it considered dropping the product line entirely. So the pattern held across both standards. It was never that Americans did not want the pages. It was that no one could buy the thing that displayed them, and no manufacturer would tool up for a market of a few thousand.

~ the cracks ~

There is a second story sitting underneath this one and it did real damage. Line 21 of that same gap in the picture carried closed captioning for deaf and hard-of-hearing viewers, a service with a small audience that depended on it absolutely. Some broadcasters running NABTS dropped captioning because the equipment would not do both, and advocacy groups fought them over it in public. Interactive television in America arrived, in part, by taking something away from people who could not afford to lose it.

NBC lasted twenty months. CBS kept ExtraVision going into 1986 and then quit as well. Neither network ever published a number for how many households had a decoder, which is its own kind of answer.

~ what got buried with it ~

The standard outlived the service by more than a decade, which almost never happens. NABTS was still a perfectly good way to push data through a television signal, and once the receiving hardware stopped being a dedicated box the objection went away. Intel launched Intercast in 1996, where a TV tuner card in a Windows PC pulled HTML pages out of the vertical blanking interval while the picture played in a window alongside them. WaveTop used the same transport, and Microsoft built the whole thing into WebTV for Windows under Windows 98.

All of those are dead too, and they died of the internet arriving on a wire that could answer back. But the run is worth noting: a technology that failed in 1985 because the decoder cost too much came back in 1996 because the decoder was a card in a machine people had already bought for something else. The idea was never wrong. It was waiting for somebody else to pay for the hardware.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »