[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [microsound] AI & rhythm perception - 'groove' heuristics?



David Powers wrote:

To clarify, I mean you can break that 16th note pulse in many ways -
but the final result of a dense groove is that the different parts
blend into a single stream of 16ths. Then in more sparse sections,
parts are stripped off to reveal some simpler sub-rhythm, which
typically hits on between 2 and 4 points within the space of two
beats.

I think that's true, but it doesn't capture what I find interesting about the notion of groove.


http://damian.fusion.net.nz/frey/nice-loops/richard-davis-breathe.mp3

Yes, there is a continous 16th pulse in there, but what makes it groovy is the hihat + bassline with its sense of collection and rising between beats 1 and 3, the sense of sudden drop at beat 3, pause to recollect between beats 3 and 4 and then a little mini scramble upwards between beat 4 and beat 1 of the next cycle, to drop again at beat 1. The hihat skips 16ths over the first half, which lends energy to the bassline, but then when the bassline is lying prone collapsed over beats 3-4 the hihats bang out all 16ths as though providing a commentary on (or a roof over) the action.

Then in the second iteration you have the addition of the open hi-hat which adds a sense of backward and forward, kind of a direction reversal every 1/4 note.

The record crackle is window dressing that nevertheless adds a little 'squeeze' to the 1, and the little synth lines serve to outline the larger formal structure. I see a bouncy ball in a gravity well with springy floor and walls, and the whole thing is kind of breathing over the space of a measure, expanding over beats 1-3 and collapsing over beats 3-1.

But what makes the grooves interesting is the constant building -
reducing - building  - reducing to various combinations over time, not
really any particular rhythm. You'll hear most all combinations within
an hour or two set, typically, as we are talking about a rather
limited set of permutations.

I'm not so interested in whether a groove maintains interest over time, just whether it's 'groovy'. Over any given four bar stretch, a groove can be more or less 'groovy' in my books, and the larger scale interest doesn't have any influence here.


Long-term interest is my goal though. In theory, if I can write a 'groove' based on some kind of inversion of the heuristics (ie start from the numbers and work back to the groove), then vary the numbers continuously while holding the relationships between them constant, I can generate grooves that continuously change but stay groovy in the same kind of way. Or perhaps reveal some other strange musical relationship... I have no idea whether this would even work but the possibility is kind of exciting, don't you think?

Some of what really makes great grooves in this style, for me, are not
just rhythms, but actually timbral and pitch modulations that push and
pull in certain ways.

Exactly! It's that pushing and pulling that I'm most interested in, because I believe that it hints at a kinematic basis for grooviness. There's a sense amongst the pushing and pulling of /balance/, and for me a groovy beat is one that pushes some kind of centre of balance around in a way that makes it almost topple over, but not quite. (The best I can think of is Sutekh, check his album Incest Live or this sample: http://damian.fusion.net.nz/frey/nice-loops/sutekh-incest-06.mp3 , it's an amazing example of how to build dynamically balanced but unstable beats.)


--
Damian Stewart
+64 27 305 4107

f r e y
live music with machines
http://www.frey.co.nz
http://www.myspace.com/freyed

---------------------------------------------------------------------
To unsubscribe, e-mail: microsound-unsubscribe@xxxxxxxxxxxxx
For additional commands, e-mail: microsound-help@xxxxxxxxxxxxx
website: http://www.microsound.org