Andy Finney: Choosing the hardware for BBC Domesday Project
Article written by Andy Finney
Producer on the original BBC Domesday Project
When the Domesday Project was devised back in 1984, one key question we had to resolve was how to deliver it.
There were two strong reasons for using the BBC micro: it was widely available in schools and its video output was compatible with 'real' television. Similarly, the Philips LaserVision system was suitable because it was a tried and tested tool for interactive video in the UK.
In fact, there was no other real option. (In the USA the Apple II occupied this niche and the laser videodisc players came from Magnavox.)
BBC Video had already demonstrated that a videodisc with a data channel could be produced commercially. They had published the Videobook of British Garden Birds (presented by David Attenborough) in 1982.
This disc was designed to be a little more interactive than most consumer videodiscs by being 'active play', which meant you could still-frame the video on the disc, carefully indexing the sections on each bird, having two soundtracks (with and without commentary) and adding a teletext magazine. It was apparent from the technical specification of LaserVision that it had enough bandwidth and was stable enough to carry broadcast standard teletext. This was something that was not possible on video cassette.
The BBC Natural History Unit, who produced the disc, also provided information on the birds and the Ceefax department produced the teletext data. The oddest part of the whole process was that, in order to insert the teletext data into the TV signal BBC2 had to be taken off-air while its network equipment was used to make the videodisc master. As far as we know, this disc was the first consumer product anywhere in the world to use an optical disc to store data, rather than digital audio. I was later told by Philips that this BBC disc helped them decide that optical data discs could be viable. If only we had patented it.
Originally, the Domesday team thought it might be possible to store the data for the project as television frames full of teletext. The BBC Micro teletext adapter, another piece of readily available kit at the time, could gather teletext from any line of the TV picture. I had written a demonstration interactive application (in BBC Basic) which provided a user interface to the Birdbook while also taking background data from the teletext on the disc and blending it into a single application.
A team at the University of London had produced a demonstration disc which included some full-frame teletext. The problem was one of error correction in a mass-produced disc, since teletext refreshes regularly when transmitted and so relies partly on this repetition to handle occasional errors. This would not work with a videodisc.
As it happened, the question of teletext was sidestepped by Philips Research who told us that they could add a rugged data channel to the videodisc format by replacing the two analogue sound channels with a channel carrying the data. This was LV-ROM. In the final system, the player also included a system to lock the disc video to the computer video and combine them in various ways under control of the software. So the computer was plugged into the player and the player was plugged into the monitor.
The computer was almost a standard BBC Master computer, complete with a second processor (yes, Domesday used two cores) and a special filing system to control the player. The only new piece of hardware was an interface between the computer and player which transferred the data and controlled the player. This was the then brand new SCSI interface, beloved of high performance hard discs to this day, and it slotted into the micro in a new slim circuit board.
The firming up of our delivery platform and the production of the content took place side-by-side and we were a long way into the project when Philips were able to produce a real LV-ROM disc for us. But it all became a little more real when we were able to hold that first disc in our hands.