Shake Two

 

One of the most fundamental of all rhythms used to establish one of the additive rhythms that make up a given polyrhythm set is the 8-beat pattern:

or 1-2 1-2 1-2-3.

Obviously, there are a couple of variations on this. It can be inverted, turned inside out, etc. It is obligatory that every trainee 'master drummer' or 'master dancer' learn to this so they can play it and its variations in their sleep.

Although this is an 8-beat bell pattern-type cycle, since the 12-beat cycle is so much more versatile, an extra half-cycle is often added to it to make. The 12-beat cycle is then made up of two 3-beat patterns (i.e. two sets of 1-2-3) and three 2-beat ones (1-2 comes three times across the cycle). An example would be the Event Sequencer laballed as Pattern 6 in the attached patch:

or 1-2 * 1-2-3 * 1-2 * 1-2-3 * 1-2.

Obviously, the important thing that a drummer or dancer must do (indeed, also a vocalist) is keep a timeline of this nature going rock solidly and indefinitely by being able to pick out the designated accents repeatedly. They must also, however, be able to play anything else they like over the top of it and so accent the on- and off- beats at will.

Apprentice master-drummers can often be seen 'learning their chops' by playing with two shakers or rattles.

With the right hand they choose one of the 10 possible ways in which the above 12-beat pattern can be played. They would gently accent the beats that announce a 3-set. The timeline having been established, they would then produce various kinds of resultants by bringing the minor hand into play. They cycle through the whole set of 10 other patterns at will and at random so that they would create changing sets of resultants as all the while the one hand maintained th classic timeline whilst the other enjoyed the freedom of picking whichever one of the others was desired at any time.

In the attached patch I have selected Pattern 6 as the ground or timeline. There is no particular reason for this. Any one of them would have done. (I would add, by the way, that there is no significance in these numbers. They are entirely arbitrary).

In order to illustrate the kind of effect I am trying to achieve, I have connected up Pattern 1. It is not a very interesting sound, but that is not really the point. I am trying to resolve a general issue.

Where I have got completely stuck (and I have tried a number of strategies including using key quantizers, sample and holds, logic compares to act as note detectors that would then drive selected event sequencers and all sorts of other things) is in determining a strategy for getting the Red Goddess to randomly select one of the 12 patterns exhibited in the two columns of event sequencers to the right. The net result should be an ever-changing set of resultants as the Nord Modular cycles through the set by forever picking one or another pattern to accompany the basic timeline. (There are 12 event sequencers available because silence is as much a part of music as noise, and so every so often one would just let the ground/timeline sound out without any accompaniment).

I hope that what I am trying to achieve is by now obvious: that a clock generator, random generator, LFO (and I have tried all of them) or some other device can randomly select which particular event sequencer it is going to drive. A different pattern should be randomly selected at the beginning of each cycle -- just like a real human being practising away would do who was trying to learn their way around these little rhythms and thus learn the total freedom of expression that they eventually bring.

The most frustrating problem I had was that if I did manage to select event sequencers randomly, then only the first beat would play because it had no clock; and that if I did manage to keep the clock going then I could not randomly select just one event sequencer.

If anyone would like to know the strategies I tried -- and that failed -- I am willing to tell them. However, it seemed to me that since I might have overlooked something obvious in what I did, it was best not to 'contaminate' anybody by giving them the details of what I did so that anyone interested could try their own fresh approach. I am sure that I am just doing something very silly, but cannot at present see what it is.

Thanks to anyone who is even prepared to look at this ... and even more thanks to anyone who manages to solve it.



Jan Punter wrote:

Something like this maybe?

As the cost of using of sequencers is pretty high (zero page memory!) I decided to leave out the ones that generate silence. The possibility of silence still exists though as the note quantize module will select more values than are decoded by the keysplitters, all undecoded values will result into silence. Er, silence meaning that only the timeline plays, of course.

Anyway, leaving out as many sequencers as possible will result in more room to generate interesting sounds.

The logic inverter used to reset the pattern sequencers seems to be needed as otherwise the sequencers will sometimes stop at step 2 instead of 12. The Distraction version is just a try to make it a little more interesting, for me - anyway it helps to forget a about gray skies & rain

 

ShakeTwoJPunter02.pch


Distraction02.pch

Distraction03.pch

Distraction03-1.pch

Distraction04.pch

Distraction05.pch

 

Kofi Busia wrote:

Have now looked at your patches. I quite like the distractions patches ... and I am presuming they are just personal developments -- ideas you have had for patches -- based on the patch I first submitted, issues I first raised.

Jan:

Indeed, they are. I wrote more about that in a previous mail that I sent before I read yours. I always build machines and your mail triggered some new ideas to try on that.

Kofi:

The specific patch you sent in, and on which the distractions patches seem to be based, does not, really resolve my problem. There is, for example, now a dzi on every beat -- whereas in the actualy godzi polyrhythm some beats are not played at all.

Jan:

Sorry to hear that, to not have heard that, I mean ... distraction can be a problem then ? I'll re-read what you wrote ... maybe I could use some text book with what if's and answers... ;-) I don't have much musical education, so I don't really hear the things you write about. All I know of is whether it's good what I hear, or not - for my ears, that's to say. But never for sure, as it always changes.

Kofi:

Nevertheless, I am still going by the sound and the effect. I have yet to finish actually working my way through all the various logical ands/ors that you have cobbled together to see if -- particularly in conjunction with Rob's answer to Mountain man about resets the other day -- I can find some way of creating an ABC song form that loops through B indefinitely many times under user control, and that then always jumps to and plays through C pattern when finished, while always beginning from A when reset.

Jan:

I think you should look into the construction I used in the 'free' patches then, it's less modules, so easier. You said that the example patch I sent did not play a dzi on every beat, which I do not quite understand yet - but I did notice some strange behaviour that the distracted patches do not seem to have, but as I said I'll have to reread what you wrote before.