The Evolution of Digital Video


The digitisation of TV can be traced back a quarter of a century to the first digital video effects box, which scaled and rotated live images. This was followed by many expensive digital video devices: stills stores, caption generators, paint systems, vision-mixers, and so on. These were all digital islands, taking analogue video in one end and putting it out at the other. Analogue video is dirty: as each process is applied, and every time it’s recorded to tape, noise is added. This is hard to detect after just a “pass” or two, but the material becomes corrupted after any more so-called “generations”. Producing special effects for pop videos and commercials was a nightmare. It involved hooking as many open-reel analogue video recorders and digital video devices as possible to a huge vision mixer, so every element could be combined and recorded in a single pass.

In 1985, Sony created the first digital VCR using the D1 format. By providing an all-digital environment, this eliminated generation losses since it enabled material to be recorded, copied and treated with no loss in quality. D1 revolutionised video effects, but the process still relied on video tape and so remained linear. Productions were assembled slowly, laying everything in sequence with recorders, taking up to a minute to sync for each edit, in a process called pre-roll. If something needed changing, everything after that point had to be repeated. Creating a composite or multilayer shot meant recording to and from D1 VCRs, adding a layer at a time. The process was laborious, even with a basic computer to keep track of time-codes for each editing event.

The late 1980s saw the first video disk recorders. They held only 30 seconds of video, but they offered a taste of non-linear editing, reducing the time it took to assemble effects shots. More radical was a device called Harry from UK manufacturer Quantel. It combined a few minutes of digital-disk storage with a 2D graphics system and a crude but elegant means to assemble video clips. It was the only digital non-linear editing workstation capable of producing broadcast-quality material for almost a decade. Quantel’s solutions are proprietary, however. Not even the most powerful computer workstations available at the time supported the 180 Mbit/s throughput needed for live video. The only computers involved in video (typically for 3D graphics) worked on a frame at a time, recording each to tape in sequence. Disks couldn’t support high data rates, buses weren’t wide enough and processors couldn’t have done much as the data whizzed past anyway. Video was too big. After all, desktop PCs had only just managed to handle stereo audio at under 2 Mbit/s.

Two factors created a revolution. The advance of computer technology pushed PC bandwidth up while video compression techniques reduced the bandwidth needed. By the end of the 1980s, the two crossed and began to have an impact. Now they’re having an equally big effect on video distribution and its use at both home and in the workplace, with MPEG-2 and DV having emerged as the predominant standards