Throughout the evolution of the PC during the 1980s soundcards and sampling technology were introduced and the dedicated sound chips were replaced by digital audio handled by the main CPU. It was sampled at incredibly poor bit depths and rates like 8 bit, 11KHz. To begin with only one or two channels of sound were available, but as computing power grew software mixers capable of delivering 4, then 8, 16 and 32 channels of sound emerged. Through the late 1980s and early 1990s computer game music and sound effects mirrored the developments in pro studio production, but lagging behind state of the art samplers and synthesisers by a few years because of the difficulty in running audio software on an ordinary CPU while also delivering graphics. Commercial synthesisers and samplers employed dedicated DSP units and were designed with direct memory access and data pipelines unlike a general purpose computer. Finding it's own optimal form, computer game music became sequenced samples using quite sophisticated tracker formats, a file that contains sample data and sequencing information. In some ways this represents a high point in game music because of the flexible real-time control the game logic had over the music sequences.
Andy Farnell
http://obiwannabe.co.uk/