The PC In Music

Brian Heywood traces the development of the most popular computer platform in the world, and examines its applications in music.

The last decade has seen a revolution in the use of personal computers in all aspects of society. The impact of any kind of computing on people's daily life a decade ago was almost invariably behind the scenes and at work; the home was a blissful haven from the chattering of dot‑matrix printers, the chugging of floppy disk drives and the whine of VDUs. Today the personal computer has insinuated itself into all aspects of our daily lives, to the point where Microsoft can cheerfully spend millions of dollars hyping up an operating system upgrade to the general public. And, of course, these technological developments have also changed the way we make music.

Today's big three personal computers for music, the PC, Apple Mac and the Atari ST, don't get a mention in either the editorial or the adverts of the first ever issue of SOS. This is not to say that SOS ignored the use of computers in music, since a BBC Model B sequencing package — the UMI‑2B — received a major review. This undoubtedly reflected the relative expense of the American computer systems at the time.

When SOS first hit the news stands in late 1985, the IBM PC had been around for four years, while the PC‑AT and the Apple Mac had been available for about a year. The Atari ST was the new kid on the block, and was meant to steal the Macintosh's growing market by being cheaper and faster. In terms of music support, the PC was actually doing quite well, at least in North America. In fact, one of the first professional sequencers available on any platform — Sequencer Plus from Octave Plateau (who later merged with Voyetra) — was developed on, and for, the PC.

The way we use computers today in the business of making music mirrors how the 'state of the art' has developed. Computer technology now allows us to approach our music making in a variety of ways that were only available to professional studios a mere decade ago. The development of the IBM PC and its clones as arguably the most popular personal computing platform has driven this change to a great extent. By reducing prices and making the technology do more, the PC has given the competing platforms something to strive for, and generally 'improved the breed' of all personal computers. But computers are complicated beasts, and the PC — possibly due to its very adaptability — is especially so.

Putting together a music system using a PC means making decisions that ST or Mac users wouldn't need to make until they needed a far more advanced system. The glittering prize is that the PC musician has far greater range of possible music system configurations. The 'wooden spoon' is that he/she can get lost amongst the poor documentation, misleading sales pitches and disingenuous specifications — lies, damn lies, and statistics. There's nothing inherently difficult about the process — it's just a matter of getting the right information.

The PC is designed to have an 'open' architecture, which means that there's a lot of competition in the marketplace, and you're not tied to a particular vendor. These factors make the PC a very flexible and relatively cheap computer to use for music applications, since fairly small companies can address specialised niche markets — such as MIDI interface cards, hard disk recording, and so on. The downside of the PC's flexibility is that you need to have more of an idea what's going on under the bonnet to choose the PC most suitable for your needs.

PC History

The PC is actually a family of computers, which started in 1981 with the introduction of the original IBM PC. This machine was based around an Intel 8088 microprocessor running at 4.77MHz and was designed to compete with the business computers of the time, namely 8‑bit CP/M computers with 64Kbytes of RAM running at 4MHz. Much to IBM's chagrin, the design was hijacked by other computer manufacturers, who produced PC 'clones' which were not only cheaper, but usually better than the IBM original.

Since then, Intel have produced a series of processors derived from the original 8086 chip, but progressively faster and using larger data word sizes. In 1984 IBM introduced the PC‑AT (PC‑Advanced Technology), which is the basis of the Industry Standard Architecture (or ISA) used by most of today's PCs. The current head of the Intel family is the Pentium, a 32‑bit processor running at 60 or 66MHz. Incidentally, the only reason it's not called a 80586 is that Intel are not able to register a number as a trademark.

Thanks For The Memory

One of the most confusing aspects of the PC is the number of different ways of looking at the RAM — again, historical reasons are responsible for the confusion. The original PC had a fairly conventional design for its memory map: its program memory started from zero, with the operating system placed at the top of the available RAM. Since part of the operating system was stored in ROM, which was placed at the top of the memory map (just below the 1Mb mark), this left 640Kb of usable RAM. Time and technology moved on, but the massive popularity of the PC and the requirement that old DOS software should still be able to run has meant that this basic memory map is still used. In the days before Windows, two competing systems — expanded and extended memory — were developed to allow DOS programs to use the memory above the 1Mb mark. In Windows you shouldn't need to worry about the type of memory being used, since Windows provides memory management for its applications. So the three types of memory are:

• Conventional memory.
• Upper memory blocks (UMBs).
• Extended and/or expanded memory.

The conventional memory is the 640Kb of RAM that can be used directly by DOS programs, the upper memory blocks are chunks of free RAM located between the top of DOS memory and the 1Mb point, and expanded and extended memory is located above the 1Mb point and is used by DOS programs. (DOS 6 provides a nifty utility called MEMMAKER that will automatically optimise your PC's configuration for the RAM you have installed.) It's worth noting that a PC's memory figure includes the 640K and the UMBs, so that a 4Mb PC will only have around 3Mb of extended memory.

Musical Uses Of The PC

The PC has three major areas of use in the musical arena: notation, MIDI sequencing, and digital recording. While software can address more than one of these areas, the requirements for each are somewhat different.

• NOTATION
  In some ways, the production of musical scores is the most obvious use of the PC in music. Most scoring programs also add MIDI auditioning and import/export functions so that you can audition what's on the page and transfer the music to and from the real world. However, requirements of a listenable performance and a good‑looking page are entirely different, so beware of applications that purport to offer both. There are some good combined notation and sequencing packages around, but they tend to be pretty expensive.

Since all decent notation packages on the PC run under Windows, you need at least a 486 PC, the faster the better. This is because the scoring application needs to format and display a lot of graphics, and a faster machine will reduce the amount of time spent waiting for the screen to update. If you plan to use the scores in other DTP packages, you need to check that you can produce the finished score as a Postscript file, since this is the only reliable way to transfer the images.

• SEQUENCING AND MIDI INTERFACES
  If you want to make music rather than produce a musical score, one of the simplest ways to produce high quality sound with the PC is to use it to control a MIDI synthesizer. This method of making music is attractive for a number of reasons, the major one being that there is a wealth of different sound generation technologies available at quite a reasonable price. It also makes sense to use a dedicated sound synthesizer, as the amount of work the computer needs to do to produce reasonable sound quality prevents it from doing anything much else. Of course, the synthesizer doesn't have to be external to the PC — a growing number of PC expansion card‑based synthesizers is becoming available.

The processing requirements of a MIDI‑only system are really quite modest in terms of the amount of data involved, and for a DOS‑based system, a 286 or fairly modest 386 is really all you need. However, under Windows, the operating system can slow things down somewhat, so you should really consider at least a 486SX/33 to get reasonable performance. One of the major advantages of using a Windows sequencer is the range of sophisticated MIDI cards that can be handled. Some modern synthesizer modules can even be directly interfaced to the PC via the serial port.

• DIGITAL RECORDING

The most exciting recent application for the PC is directly recording digital audio onto the PC's hard disk. There are two basic routes for this:

• Use the facilities built directly into Windows as part of the MPC (Multimedia PC) standard;
• Use dedicated add‑on audio hardware.

The second option will give superior facilities and performance, but at greater cost. The only advantage of going the MPC route is that you can incrementally 'scale‑up' your system as your requirements increase, and prices come down.

Soundcard Anatomy

A Windows MPC soundcard comprises three basic elements:

• A sampler section for recording/playback of sound;
• A computer‑controlled mixer, for combining audio from various sources;
• An internal music synthesizer.

Most soundcards also have a joystick port, the capability of adding a MIDI port and quite often a CD‑ROM interface.

Traditionally, the synth section found on most of the 'standard' (or games‑derived) soundcards was based on a Yamaha design known as the OPL3 chipset. This technology is based on FM synthesis, which now sounds rather dated. Quite a number of recent soundcards have augmented FM (or replaced it) with a wavetable syntheziser, which uses instrument samples stored in ROM (Read Only Memory) or RAM (Random Access Memory). Currently the better soundcards for music use wavetable synthesis, but the quality of the sound depends on both the amount of memory and the quality of the audio data recorded onto them. The cheaper wavetable cards skimp on both. While ROM‑based wavetable cards are fine for reproducing music that has been designed for MPC or General MIDI (GM) players, if you want to get more creative, you really need a RAM‑based card. These cards allow you to design your own sounds (often based on Windows .WAV files) using software supplied with the cards.

The future seems to be with DSP‑based synthesis: Wave Guide, a type of physical modelling from Media Vision and Yamaha, creates a software model of an instrument; and its sound output is a side effect of the calculations, so you design the instrument rather than the sound it makes. Both wavetable and Wave Guide technology is available on daughter boards that can be added onto an existing soundcard fitted with a 'Wave Blaster'‑style connector, so it may be possible to improve your soundcard's MIDI sounds. For serious musical use, check out cards from Turtle Beach, Roland and the new daughter board from Yamaha.

Future Tense

The PC has changed a lot since it was introduced in 1981, and in the process has become the most popular computer ever made. A couple of years ago, Microsoft were saying that there were over 30 million Windows users, and since it has been estimated that there is a four‑to‑one piracy rate, this means that the figure could have been as high as 120 million. As for the future of the PC, the Intel architecture shows no sign of running out of steam, so there's probably a new processor just around the corner to handle those applications which need even more power than is available today. So there's life in the old workhorse yet.