[PD] Structures without random

ben at ekran.org ben at ekran.org
Sun Aug 10 01:47:16 CEST 2003


Hi brandon,

In fact I was thinking about chaotic systems as I was reading the first
part of your message. chaos is an interesting subject, and we can have
systems that go all the way from chaotic to periodic, to even a fixed
point by the change of a single variable.

I suppose the question is, what do you really want it to do? Do you want
structure or randomness? Chaos is not random because its not diffuse,
there is structure in that some numbers are more likely to occur than
others. I just made some fractal plots today actually, to give you an idea
about how complex the structure is... (attached)

The current release of chaos has 4 attractors, 2 3D and 2 2D. lorenz,
rossler, ikeda, and henon. lorenz and rossler come from fuild dynamics
models. henon from astrophysics and I'm not sure about ikeda. Right now
you have to tweak the variables yourself to generate something from chaos
to periodicity. The next release of chaos (coming soon) has (I think 15
new attractors) and allows you to search for variable sets with a certain
amount of chaos your looking for (Thanks to Michael McGonagle).

Anyhow the current version is in CVS or you can get it from
www.ekran.org/ben/research/chaos.tgz (includes windows and OSX binaries +
source)

Anyhow if you do a search on chaotic and strange attractors you'll find
lots of info.

Oh and a last note:

According to Benoit Mandlebrot, the fractal patterns we observe in nature
are due to differential equations. These equations are the foundation of
what is known as chaos mathematics.

chaotic systems are models of complex natural systems.

Some other techniques for generating a number of controls from one would be:

the PSO object I will soon release. It creates a population of random
particles that try and imitate the input. You can control as many
particles as you want, in as many dimentions.

Onother similar option is to chain a bunch of my x-y controllers together,
then each controller tries to follow the previous, so you have a train of
controllers following your movement. The more (slave) controllers you
have, the more abstracted their movement will be in relation to your own
gesture.

Anyhow play around with a few things and find out what does what you need.

good luck.

Ben


> Hello,
>
> I am curious as to what type of structures everyone
> uses for controling data.  There are probably millions
> of ways of answering this and there will always be new
> ways of thinking about it.  As much as I try to avoid
> it I inevitably find myself putting one or two types
> of random components in a patch to control a variable
> paramater somewhere.  This is because I want to get
> some numbers running through the system.  This seems
> to be the quickest way to do this.  With the patch I
> usually only interface with a mouse, and so that means
> if I plan on having multiple things happen, I need a
> way to automate them.  I have buit a step sequencer,
> and a data recorder that will record atoms to a table
> and play it back in the same order at variable speeds.
>   But I had a someone suggest that I look into chaotic
> systems as opposed to random systems.  I am not sure
> what a chaotic system means though?  It seems as
> though the way I am looking at it is more black and
> white, like there is either complete structure, or
> complete randomness.  Can someone help me get out of
> this box I have put myself in????
>
> Bradon~
>
> __________________________________
> Do you Yahoo!?
> Yahoo! SiteBuilder - Free, easy-to-use web site design software
> http://sitebuilder.yahoo.com
>
> _______________________________________________
> PD-list mailing list
> PD-list at iem.at
> http://iem.at/cgi-bin/mailman/listinfo/pd-list


-------------- next part --------------
A non-text attachment was scrubbed...
Name: chaos.jpg
Type: image/jpeg
Size: 40061 bytes
Desc: not available
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20030809/53ff7a1e/attachment.jpg>


More information about the Pd-list mailing list