IN HIS TIRELESS SEARCH TO RECONCILE MUSICIANS AND MACHINES, MT'S DRUM PROGRAMMER EXTRAORDINAIRE TAKES ON THE MASSED DRUMS OF AFRICA.

WHILST IT WAS always my intention to include Africa in our rhythmic world tour (I could hardly lay claim to any kind of completeness), it has, nevertheless, been the one area in which I've harboured reservations about what could be achieved through the medium of the drum machine. This isn't through any belief that African rhythms are in some way sacrosanct and should be spared the hi-tech iconoclasm of machine programming. Rather, it reflects the concern that without the unique instruments that lie at the heart of all ethnic music (and particularly that of Africa), any attempt to recreate an authentic feel is likely to meet with little success.

In many ways, it's a magnification of a problem we've come up against on a number of occasions in this series. Whenever we move away from those rhythms which we might call "mainstream", patterns often start to become instrument-specific. And whilst not necessarily redundant for those without access to the same (or similar) sounds for which they were written, many rhythms are bound to lose something in their journey from my programming system - via MT - to yours. This is one of the reasons I've continually stressed the need for experimentation.

In our fairly exhaustive examination of South American rhythms this problem was offset considerably by the fact that many Latin instruments have become part of the currency of modern beat boxes. After equipping machines with a dozen or more bass, snare, tom and cymbal sounds, manufacturers invariably look to Latin voices as a way of increasing their sonic capability. The same is not true of African instruments, consequently this is one area in which a sampler comes into its own. We're not talking about the kind of instruments most people have lying around the house here, and I can think of few, if any, recorded sources of sounds (without sampling from album tracks). And even if samples are available, the number of different tonal variations required to achieve anything like a credible result is quite prohibitive.

This was just one of my reservations. Others concerned the difficulty of trying to isolate representative examples from a continent as large and culturally diverse as Africa, and the problem of taking what must, by necessity, be "snapshots" of rhythm tracks which traditionally progress throughout the length of a piece of music. Fortunately, I found the solution to all these problems in but a single approach. Instead of attempting to accurately notate specific African rhythms from a variety of individual countries, I've presented a series of patterns, which, though unequivocally African in origin, represent a rather more eclectic approach to recreating the rhythmic styles of that continent. Of course, I realise that by opting for this method I have laid myself open to claims of having taken the easy way out. However, I belive that it's the success (or otherwise) of the patterns I've included which should be judged. And in the final analysis, a handful of patterns which do work, have to be regarded as preferable to a handful that don't - no matter how authentic and accurately notated.

My only other concession to the problem was that of dividing the music of the continent into two areas - North Africa, and Central and Southern Africa. And whilst I accept the arbitrary nature of such a division, it does, as anyone familiar with the region will be aware, represent a dividing line between the predominantly Islamic states of the north (with their own, quite disparate musical development), and the rest of the continent. In a rhythmic context this is, I believe, a valid distinction to draw.

It also provides us with a useful split in our two-part coverage of African - though with room for only half a dozen patterns in each, you'll appreciate the limitations. I hope that these articles will prove that programming African rhythm tracks is possible, and that you will go on to adapt and eventually write your own.

As you will see, this month's collection of North African rhythms have been assembled from fairly standard beat box voices, but without the usual snare, hi-hat and cymbal sounds. The backbone to these patterns is provided for the most part by series of individually tuned tom-toms, congas and bass drums, which, though quite different from the indigenous instruments of the region, can provide a credible means of recreating its traditional rhythms.

Their success in this role relies to a considerable extent on their effectiveness when played - or when programmed - in concert. Confronted by a battery of different instruments, the ear is far less discriminating about individual sounds and concerns itself more with the overall feel of a rhythm. It is for this reason that the patterns leave little space for you to make qualitative judgements on the authenticity of specific instruments.

Within this framework, we can help matters considerably by using instruments outside their normal tuning range in preference to those played back at their original pitch. This is especially true of tom-toms, which tend to exhibit some of the more unpredictable nuances of acoustic drums when tuned outside their sampled pitch. Full use should also be made of whatever tom-tom voices are available on your machine; try as many different sounds as possible for each part and where a combination of instruments is called for, try to maintain a broad mix of single- and double-headed drums of differing sizes.

As you might imagine, the bass drum doesn't fulfil the same role in these patterns as in Western rhythmic form, and should be regarded as a lower-pitched version of the other drums. To this end, a boomy-sounding voice is more appropriate, and given the almost subsonic thud of most modern instruments, you might find this easier to achieve by raising the pitch slightly.

Instruments such as the tambourine and cowbell, though frequently encountered in African music, are rather too "refined" in the form included in most drum machines. Again, a little detuning or perhaps combining two or more instruments at different pitches can work wonders.

Obviously, the lowest common denominator when choosing instruments for these patterns was to make them as accessible as possible to the greatest number of people, but if you have a range of more authentic instruments at your disposal, you should make the appropriate substitutions. You might find that patterns can then be thinned out significantly - for the opposite reasons to those I outlined earlier.

Alternatively, if you're using software-based programming systems (where it's generally easier to manipulate data) you might try moving parts around between different patterns. One of the reasons I've "squared off" the patterns into 4/4 grids is to make parts interchangeable for those prepared to experiment. The other reflects my belief that programmers should be encouraged to adopt different styles and rhythmic forms, and to make this more feasible I've tried to present patterns in a way which will make them usable within the context of Western popular music. Needless to say, in their countries of origin, such convenient structures are rarely encountered.

To most African percussionists, the concept of keeping rock-steady time would also be quite foreign. Within the constraints of playing alongside other musicians, rhythmists would see no contradiction in the idea of keeping time by adjusting it to match the mood of the music. Neither would they pay the same attention to synchronising note for note with other percussionists playing a similar part; adopting the right feel and preserving the flow of a rhythm would be valued much more highly.

This vaguely anarchic approach to playing is one of the things which makes African rhythm so compelling - particularly the music of the central States such as Zaire and Burundi, where the sound of massed drum rhythms are so powerful. Were the drummers to adopt stricter control of their timing to the point where they perfectly synced to each other, the effect would be all but lost.

This, I'm sure, would be cited by purists in their objections to programmed African rhythms. I don't share their reservations. For those of us who aren't Paul Simon and whose bedrooms aren't large enough to accommodate the Drummers of Burundi, the drum machine represents our best means of tapping into this rich vein of rhythmic invention. With the advent of real-time programming systems and high internal resolutions, we can even capture the spontaneity of human performance. Of course, this means that parts should be programmed "by hand" wherever possible. If necessary, you can reduce the tempo until you're comfortable with the programming speed, and there's no reason you shouldn't enter parts a few notes at a time in a series of overdubs. If you have trouble sight-reading (translating the notes on the grids to the keys on your machine) there's nothing to stop you entering the notes in step time, listening how each part is formed and then re-entering them in real time after you've had sufficient practice.

For those dead against the idea of real-time programming, you could try shuffling step-time entered notes as a means of creating a more random feel, and those with more sophisticated machines or software may like to try achieving the right feel through the use of programmable quantisation. Whatever method you chose, the need to preserve the human component in these patterns cannot be overstated and will be an important determining factor in whether the rhythm tracks you produce are identifiable as African in influence.

In writing these patterns, I drew principally on the influence of music from Algeria, Mali, Ethiopia and Morocco. How well the different elements coalesce is for you to judge, but there should be no mistaking which region of the world these rhythms come from.

Because of the importance of choosing the right instruments, I've included a table giving details of the voices I used to get you started. Programming is quite straightforward, though I should point out that the tom-tom notes in Pattern 3 which appear fractionally before and after the beat division lines are intended to represent the human programming of toms 1-6. How much each deviates from its fully-quantised position will depend on the resolution of your machine and just how ragged you want the drums to sound.

Also, it's probably worth making it clear that although Patterns 4 and 6 appear in triplet time, they are straight 4/4 patterns written that way to accommodate the triplet beats in the conga and "dry" tom parts. If your machine allows you to program individual instruments in triplet time, program the rest of the patterns in straight 4/4.

(Click image for higher resolution version)