The computers used for restoration
I started on the 30-line recordings over 30 years ago and the systems I built or bought to do the restoration trace a history through personal computing.
The first 30-line recording I found, was an extract on an audio documentary on LP – John Bird narrating ‘We Seem to Have Lost the Picture’. This turned out to be a few seconds from a BBC transcription disc which further turned out to be a copy of the Major Radiovision disc. At the time, I was using the second computer system that I had built. This was a Z80A 8-bit microprocessor system with 48kBytes of RAM, with the remaining 16kBytes (of the 64kBytes maximum addressable space)available as multiple software-switchable banks of EPROM and extra RAM. The software could directly select the extension memory dynamically in a process called paging. Burned in EPROM were an excellent Pascal compiler and a 8k Basic interpreter. To make it simple on myself, I based the design on the NASCOM computer that had been launched in 1977, and added modularity with plug in boards on a rack-mounted system providing analogue input, dedicated FM-decoding for amateur slow-scan TV, a 24 bit parallel hardware interface to a stand-alone 128 x 128 pixel image store that I had built in 1977-8 with the absolutely brand new £50/chip Texas Instruments 16kbit DRAMs. Each pixel was 12bits (4 bits or 16 levels per colour). The video clips of my 30-line conversions use the direct output from that image store. Eventually this system would grow in the early 1980s to a have a 256 x 256 pixel display and dual Shugart SA400 5.25inch floppy drives, for which I had extended the operating system to arrange storage and retrieval.
1982-83 saw the tracking down and discovery of five of the Phonovision discs with the sixth and last (RWT620-6) being discovered in a filing cabinet in the Science Museum in Kensington, some years later. By 1985, I had decided to keep up to date and bought an Atari 1040STF, a 16-bit machine based on the Motorola 68000; a beautiful machine to program. This was much simpler to code than the 8080 microprocessor family of which the Z80 was a special version. The challenge with the Atari was extending the device to control my custom-built image stores and the like. There was a small proprietary socket for cartridge memory which I used to build an interface that allowed me to capture audio, and drive the image stores. Even though the image stores were now getting old, they were superior to the graphics of the Atari ST – 320 x 200 pixels with 16 colours from a palette.
The Atari ST complete with software language packages and 20Mbyte hard disc lasted me for quite a few years until I went for the IBM PC architecture and bought a Gateway 80486-based machine in 1993. This started me on the upgrade path of the IBM PC architecture, initially using Turbo C for DOS.