Not everyone learned about MIDI when it first appeared. To many musicians it remains a mystery — especially guitarists and home recordists, who may be facing it now for the first time. Even keyboardists can find it baffling. I asked Julian Colbeck to explain the subject.

You know what MIDI stands for, yes? 'Musical Instrument Digital Interface'; but thereafter you're somewhat lost? Well, this article is for you.

Much has been, is, and will be written about MIDI. But it seems that many people continue to be confused, because they never grasped the basic facts of the matter (if, indeed, the basic facts were ever explained clearly in the first place), and it's all a bit like trying to understand Smiley's People when all you watched was the last episode. So, to those who do have a reasonable knowledge of MIDI and its applications, no apologies - just flip over the page and read something else. The following is intended as a 'Hello to MIDI' — no more, no less.

What is MIDI - the plain English facts

MIDI is simply a digital code of conduct; a set of electronic rules and regulations for sending information (via 5-pin Din leads) back and forth between pieces of musical equipment. Information such as what? Well, MIDI is all about control; allowing one instrument to control certain aspects of another's performance. This can take the form of simply 'slaving' two keyboards together so that when you play one you hear both, or it can allow, say, a keyboard to trigger drum machine voices, or it can simply ensure that rhythmic devices like drum machines and sequencers keep in time with each other. Further down the line, MIDI can allow you to manipulate (by 'remote' control) certain more specific aspects of a linked instrument's performance, like patch changes, whether sounds are to respond to touch sensitivity, etc. This covers about 0.5% of its potential!

Unfortunately, but necessarily, precisely what MIDI can accomplish on any given instrument or group of instruments is not a static affair. An instrument that 'supports' MIDI, as they say, doesn't have to implement every little scrap of inter-communicatory gobbledygook on offer; it can pick and choose. All MIDI dictates is that, if you want to be able to communicate a certain piece of information, 'this is how you do it'. Accordingly, some MIDI-blessed instruments allow you to get up to all manner of sophisticated interplay (Disgusting! — Ed.) and others simply permit a level of communication no more complex or all-encompassing than a two into one 'Y' connector. Precisely what each MIDI-blessed instrument has chosen to implement is normally contained in its MIDI Implementation Chart — a nasty, complicated-looking wodge of ★★★s, Xs and Os to be found stuffed up the back of the instrument's Owner's Manual. More of this anon. So, to recap: MIDI is simply an agreed-upon, common language allowing all manner of seemingly unrelated instruments to control or be controlled by each other. But like all languages, some choose to exploit it more than others.

Why do we need MIDI?

We don't. MIDI is a luxury item. Amazingly enough, life did exist before MIDI, and some people actually managed to produce good music without it. Beethoven and the Beatles spring to mind.

Though it's often easy to forget, MIDI was brought into this world to make life easier, as in: 'Wouldn't it be nice if I could connect up all my synths and hear them all at once', or 'I wish there was some way to change patches on all my synths without having to physically press buttons on each in turn', or 'Wouldn't it be great if I could control stage lighting effects from my guitar'. Well, none of the above are exactly essential to the furtherance of playing good music, but all are undeniably neat tricks to have up your sleeve, and all (and of course much more) are now possible using suitably accomplished pieces of MIDI equipment.

How MIDI came about

You don't really need to know how MIDI came about, but it's worth giving you this potted history since it illustrates some of the problems involved in creating a communication system that strides across all manufactural barriers — problems that have, inevitably, manifested themselves in the final specification of MIDI.

In the late 1970s and early 1980s many synth players, frustrated no doubt by the ghastly limitations inherent in available equipment, spent a lot of time mucking about with instruments in order to link them up and control one from another. Several manufacturers made life easy for the inveterate fiddler by providing communication sockets to allow their instruments to 'talk' to each other; Yamaha had their Key Code interface, Roland their DCB, etc. etc. And of course many a monophonic synth sported CV and Gate sockets - CV being Control Voltage which took care of pitch, and Gate covering timings; the length of a note. The trouble was that although many instruments worked to a one volt per octave standard, several did not. The result on connecting up certain different brands of synthesiser? Cacophony!

During the early 1980s several far-reaching concepts were being developed. Sequential took the world by storm using recently available low-cost microprocessors in the design of their first synth, the Prophet-5, and soon everyone was doing it: Oberheim responded with the OB-X, Roland with the Jupiter-8. Polyphonic synths became affordable, stable, and fun. At about this time sequencers were making a comeback — again, thanks to affordable microprocessors (Roland's CSQ-100 and 600 came out in November 1980) and drum machines, too, were being made better and cheaper than ever.

With all this new gear floating about, the question on incompatibility (so and so's sequencer won't work with so and so's synth etc.) was getting to be a real drag. Roland President Ikutaro Kakehashi moaned about it to Tom Oberheim in the summer of 1981. In turn Oberheim mentioned it to Sequential's President Dave Smith, who was fired by the idea of a universal interfacing system and resolved to do some work on such a project in the autumn. After a Heads of Company meeting in October that year, a loose form of common ground was agreed upon. There were sceptics, of course. This system couldn't be expensive: 'Why should I spend more money on a concept that allows people to use other equipment than mine' was a common gripe (hence the 5-pin Din as opposed to an XLR connector, for instance). 'I'm not going to tell everyone all our trade secrets' was another. You see, if this common language was going to to work then manufacturers had, to an extent, to spill their technological beans and be honest with other. And then there was the question of verbal communication between Japanese and American companies. 'Mono? Oh, I see; I thought you meant Mono as in monophonic synth'. Daft maybe, but true.

But a final agreement was made in August 1983 in Japan, and published as MIDI Specification 1.0. This document, now widely available (and in fact reprinted in a worthy tome entitled MIDI for Musicians by Craig Anderton), sets out the rules and regulations for entry into the MIDI club. Taking the above into account, you can then understand why so many aspects and applications of MIDI have been left open-ended. To tie up all the loose ends would severely restrict manufacturers' freedom to experiment, go their own way, and choose which facilities to include on their products and which to leave out. MIDI could only ever have been a 'use it as little or as much as you like' type of compromise. Dictatorial, and it would never have got off the ground.

A bit of nitty gritty

The driving force behind Dave Smith's initial work on what became MIDI was the prospect of solving the age-old problem of simply trying to hook up two different manufacturers' synths so that they work in tandem. Accordingly, he wanted to devise a system that permitted such a thing in as simple a way as possible. But as it quickly became apparent that MIDI's applications could extend far further than this, a number of levels or modes of operations had to be formulated. Logically enough, these are now known as MIDI modes. There are four or them.

The first mode is 'Omni', also known as Mode 1 or Omni On, Poly. This was designed as a power up mode, whereby when two instruments are first switched on and connected via MIDI cables all common ground information channels would be open. All available information could be discussed. 'All' — Omni, as in the Latin omnis, meaning 'all'. OK? But then what if you connected up more than one instrument, or a sequencer perhaps, or a drum machine? Things could go really crazy if there was no provision for specifying precisely who you wanted to be connected to at any given moment. Enter Mode 3 — Omni Off Poly. Omni off, goodbye 'all', hello someone in particular. In this mode you can have at your disposal a sort of 16-channel messenger service. By matching up channel numbers you can then communicate with a specific piece (or pieces) of equipment in a multi MIDI setup. Think of it like CB radio, whereby you can communicate with a specific person (or persons) on a chosen frequency band.

So what happened to Mode 2? As I mentioned earlier, the English-Japanese language barrier was responsible for a number of notable early cock-ups, and the need for the eventual appearance of Mode 2 appears to be one of them. When a 'mono' mode was first discussed, some people took it to mean mono as in monophonic synth — one voice only; whereas others took it to mean mono as in one voice at a time, separately controllable. The distinction will become clearer in a moment. To satisfy both parties, Mode 2 Omni On Mono, though similar to Mode 1 in that all available information is communicated on all channels, can only respond monophonically. Play a chord on your controlling synth, and in this mode a connected instrument will 'play' one note only. This note can be the highest, lowest or last note in the chord, depending on the note priority of the said instrument. Personally I find this mode debatably useful, but there we are.

On the other hand, the similarly named Omni Off Mono Mode 4 is extremely handy. This mode only really applies to instruments than can split up their voices, allowing them to pump out a number of different patches or sounds simultaneously: a glorified split keyboard effect in a way. Instruments that have this capability include most recent Sequential offerings (from the Six-Trak onwards), the Casio CZ range (four voices at a time), the Oberheim Xpander and, most recently, the Yamaha FB01 module.

What happens here is this. Each individual voice can now be assigned its own MIDI channel number — a feature that becomes vitally important when it comes to using sequencers. You know, or must have known, the old problem; you have a nice new synth, you buy a nice new sequencer which can record and playback umpteen different tracks, but... damn!... all the tracks come whizzing back to you using the same patch! The choice then becomes umpteen different tracks of bass or umpteen different tracks of brass or whatever. Now in Mode 4 you can make track 1 a bass sound, track 2 a string sound and so on, so creating a full and reasonable complete-sounding arrangement, still using just the one instrument. But there are limitations. The MIDI spec has yet to tackle the thorny problem of who, how and which detailed pieces of MIDI information, like who gets to be manipulated by the controlling keyboard's pitch wheel etc. This question becomes more complex still when MIDI guitar setups are used, but let's leave this until 'MIDI for the Reasonably Knowledgeable' (planned for a subsequent issue, Ed).

As I've mentioned already, manufacturers can pick and choose which aspects of MIDy they bung on their instruments, and this applies even to these basic modes. Some older models only include Mode 1, which means that even a fairly simply channel-assignable setup isn't on. Most instruments don't implement Mode 4 because they don't have this multi-timbral capability anyhow. Fortunately, you can see at a glance precisely what an instrument has implemented by looking up Mode in its MIDI Implementation Chart — normally column no. 2.

And then...

If I were writing a book on the subject, now would be the time to tell you about 'Channel Voice Messages' and 'System Common Messages' and the like. But I'm not, and it isn't! Suffice it to say that MIDI gets, or can get, extremely specific and detailed. You can decide to allow after-touch control over volume to be sent from your controlling keyboard, you can transfer programme information from one synth to another, use personal computers, use MIDI to synchronise sequencers and drum machines... but tha purpose of this article has been simply to introduce and explain only the most basic whys and wherefores of the subject — an appetite whetter. But it's worth mentioning here one of the most common misconceptions about MIDI; namely that it can persuade your instrument to do something which, intrinsically, it isn't capable of. In other words if you hook up two synths, only one of which can respond to touch sensitivity, the non-touch sensitive one will always remain so. MIDI works on the 'common ground' principle. What might appear confusing, however, is the fact that several recent synths (Yamaha's DX21 for instance), although not blessed with their own touch sensitive keyboards, do have this capability built into their circuitry. In this case you can make the instrument operate with player-controlled dynamics when a regular touch sensitive instrument is being used as a 'master' keyboard. Once again, this type of information can be found in ao instrument's MIDI Implementation Chart.

Types of MIDI product

Due to its origin and original application, it has been all too easy to think of MIDI as 'keyboard player's business'. Wrong! MIDI is simply a communication system and one that, increasingly, has thrown open the doors of creative potential for all instrumentalists. Here are just a few of the types of MIDI instruments and devices on offer.

KEYBOARDS

1: The Master Keyboard

This can also be known as a Mother keyboard or MIDI Controller keyboard. A true master keyboard contains no sound-producing circuitry of its own. Its job is to manipulate a number of MIDI-linked instruments and devices — how each instrument appears' over the keyboard, sorting out split or layer points, remembering combinations, assigning different real- time controllers like pitch and mod wheels, and generally masterminding an entire MIDI operation. Typical master keyboards include the Roland MKB-200, Yamaha KX-88 and Akai MX-73.

2: The Remote Keyboard

This is similar to a master keyboard in that it cannot produce sounds on its own, but its main function is to operate as a roving or 'sling-on' keyboard that can be strapped on like a guitar. Generally, remote keyboard don't offer quite the same level of MIDI management as master keyboards, but you should still be able to change patches, have separate access to more than one MIDI instrument, and be able to select and vary certain real-time controllers. Typical models include the Roland Axis, Yamaha KX5, Korg RK-100 and the new Casio AZ-1.

3: The MIDI Module

Nothing complicated going on here: a MIDI module is simply a keyboard-less instrument which can be controlled either from a master keyboard or any other MIDI keyboard, MIDI drum pads, MIDI guitars... The beauty of modules is their lack of size and their cost-effectiveness. Since most of us only possess one pair of hands, to have umpteen keyboards littered about the stage or living room can simply be a waste of space and money. All types of instruments are now appearing in modular form: synths, pianos, samplers etc.

DRUMS

1: Drum Machines

MIDI is now standard on all drum machines. The reasons you'll need MIDI controllability mainly concerns sequencers — any make of which can run in synch with any make of drum machine via the now always compatible MIDI clock. MIDI is also used when you want to 'play' drum machine sounds courtesy of an external controller. A controller can be a keyboard whereby certain notes on the keyboard can be assigned to certain sounds on the drum machine, or it can be MIDI drum pads (see below) — any form of MIDI controlling device will do!

2: Drum Pads

A set of MIDI drum pads is the drummer's equivalent of a master keyboard. They make no sounds and are simply a way of triggering sounds contained in an external sound source, be it a MIDI synth, piano, sampler, whatever. By defining Note Numbers you can 'tune' a series of pads to trigger specific notes on a MIDI instrument. This was ably demonstrated at last August's BMF by American Roland demonstrator Tommy Snyder, who played a breathtaking 'marimba' solo on a pair of Roland Octapads which were triggering sounds on, God, I've forgotten which one, but a new Roland synth!

MIDI pads can also trigger sounds from a drum machine of course, or effects' such as Tony Thompson's snare drum sound captured on a sampler! There are also several devices on the market that allow MIDI control from acoustic drums, via contact mikes and a box which converts the signal into MIDI pulses.

GUITARS

Roland have long championed guitar synths. Just as everyone thought that it was time to go home and give up, the MIDI guitar seems, finally, to be taking off. MIDI guitar setups can either be complete units in themselves, as in Roland's still burgeoning range, or comprise add-on units that can be fitted on to your own guitar, along with a 'box' that converts the signal into controllable MIDI data (Roland allows this also). Newcomers include the Shadow GTM6 Guitar to MIDI system (reviewed in Issue 11). One intriguing prospect on the horizon is the Stepp DG-1, about which I know very little, I'll admit (see our news feature also in Issue 11 — Ed.).

SIGNAL PROCESSORS

Why do we need MIDI on a DDL, eh? Once again I suppose we don't, but there are advantages, you know. By introducing a DDL or digital reverb into your MIDI setup you can match individual settings to individual patches. You can set up a situation whereby, on changes patches on your synth, a linked DDL will automatically change its patch to a predetermined complementary setting. For live performances, this capability should need no further justification.

AMPLIFIERS

As with the above, a MIDI amp (which boasts a series of memories, such as Peavey's Programax) allows you to change control settings without having to physically twiddle knobs all the time.

And Finally...

The above are just some of the types of instruments or devices which are currently benefitting from MIDI. There are, of course, many more. Personal computers such as the Atari 520ST are now being produced with MIDI built in, but for a couple of years now one has been able to gain access to computer control via MIDI-computer interfaces. Also making great strides are pitch to MIDI devices that convert audio signals into MIDI data. To an extent they are similar to drum triggering devices or MIDI guitar converters, but items such as the Fairlight Voice Tracker or the IVL Pitchrider are (at least theoretically) suitable for controlling MIDI instruments (synths, pianos, drum machines) from, say, your voice or a woodwind instrument through a microphone.

On re-reading this article I feel I must reiterate that my brief has been to introduce MIDI, not explain its every little manifestation. Although there is much more to talk about, I hope I have at least presented the basic facts in a clear light — one that will enable you to delve a little deeper into this fascinating and almost limitless subject as and when you want to. More will follow, I promise!