Tuesday, October 17, 2006

Brush or brain ?

One often compare the fractal program to the "brush" of the painter. I don't agree. It's much more (at least for me).

Look at the image above... There is a lot of small details that would be difficult to paint (on a canvas or in photoshop, for instance). And even if one manage to paint them actually, one would be stuck at a fixed resolution, while the "parameter" encoding in UF and other fractal generators allow to render images virtually at any resolution. This is what is usually considered as the advantages of fractal programs as "brushes" over more traditional brushes. They allow to produce big images with arbitrary precision.

But in fact there is much more than a "brush"... Look again at the image above... I didn't have the slightest idea about what would come out as finished image when I started. (Actually I even often start from a previous image and mutate it.) I could never have imagined this complicated pattern, nor could I have planned such a complex mix of colors. So the program is not just a convenient tool to render images, like a brush is a convenient tool to put paint on a canvas, it has a crucial role to play in the creation process. Speaking just for me to avoid hurting people, I'd even say that it's at the heart of the creation process, providing me with patterns and shapes that I try to combine in a tasteful way. And I've a whole folder of images with great patterns, that (probably because of insufficient artistic abilities) I was unable to cook properly. I think cooking is a good analogy, indeed.

Of course, one may say that these patterns do not come from nothing, someone wrote the code behind them. As in this case this someone was me, I can tell you what I had in mind when I wrote it. The image uses a big coloring, but only a small fraction of the code is actually used. This fraction corresponds to what is needed to draw a regular tiling with equilateral triangles, to superimpose such tilings of different sizes, and tweak them a bit (in some top secret way, of course). So more or less, here is what I had in mind when I wrote the code :

A little bit dull, isn't it ? With this plus some very basic operations, repeated a few times, combined by the computation power of a computer, you get this pattern on the final image. The computer somehow magnifies artistic intuition. It starts from a very stupid pattern and it turns it into something complex, unexpected, unpredictable.

So I can sum up the creative process in a caricatural way :
- Very simple and stupid pattern idea from me (and when I say "from me"... anybody could have thought to this...).
- Incredible complexification by the computer to produce something interesting and new (let me emphasize this "new"... it can produce patterns and shapes no one ever imagined, just because they are too complex or irregular).
- Cooking of the patterns and shapes by the artist. This step can fail easily...
- Rendering with arbitrary precision and details by the computer.

So really it's more than a brush... it's most of the artist's brain. :-D

(Ok, after I wrote all this, I uploaded my image... and noticed that all these "details" and "patterns" weren't so apparent at such a low resolution... so you'll have to believe me.)

7 Comments:

Blogger cruelanimal said...

Sam,

This is a fascinating image. I've really enjoyed your posts examining your process. I agree the brain (vision) usually trumps brush (tools) in the most successful art.

I also refer to certain steps in the art-making endeavor as "cooking," although sometimes I think my creations come out of the fractal kitchen closer to a stew or a buffet rather than a well planned meal.

10/18/2006 1:28 PM

 
Blogger Tim said...

Cooking is a great analogy. Great cooks make wonderful food to eat. But it is impossible for a great cook to do anything without ingredients. There is no cook who can make meat or vegetables or spices. On the other hand, these ingredients without a cook, are much less appetizing. Although sometimes, even a raw carrot can be good on it's own.

The unpredictability of fractal formulas has always surprised me. I've used a couple of simple formula parsers and amazed at how small variations in the string of z's, c's , operators and numbers can produce completely different images.

I wonder how many fractal formulas could have been worked out on graph paper, by hand, to the extent that their basic shape could be seen? For instance, did Professor Mandelbrot have any idea what the mandelbrot formula would have looked like before the arrival of computerized fractal program? I'm sure someone must have tried to start graphing one, just out of curiousity, even if it entailed several weeks of work.

10/18/2006 1:46 PM

 
Blogger Unknown said...

I wonder how many fractal formulas could have been worked out on graph paper, by hand, to the extent that their basic shape could be seen? For instance, did Professor Mandelbrot have any idea what the mandelbrot formula would have looked like before the arrival of computerized fractal program? I'm sure someone must have tried to start graphing one, just out of curiousity, even if it entailed several weeks of work.

I don't recall if Mandelbrot did any hand sketches of his set, but he was working at IBM at the time and had access to the mainframes, so he probably didn't need to. Gaston Julia did publish some rudimentary drawings of some of his sets, as did David Hilbert and Bubba (?) Koch (Hilbert and Koch curves, respectively). I suspect that the amount of calculation a) wasn't as much, as analysis was/is more important than calculation, and 2) wasn't regarded as so much of an issue, since there weren't mechanical computating machines (or cable tv) back in the day.

10/18/2006 4:07 PM

 
Blogger Philip said...

The computer somehow magnifies artistic intuition.

An interesting idea. I'll have to give that some thought.

even if it entailed several weeks of work.

I think I read somewhere that they suspected points weren't escaping, don't know if they proved it or gave up after a few hundred iterations.

10/19/2006 9:12 AM

 
Blogger Sam said...

To Terry : Thanks ! Happy you enjoyed the image and the posts. About cooking, I've usually no real plan either... and when I find a clear recipe which works well and would allow me to make plans, I start creating boring images. :-) Or maybe the images themselves aren't boring, but I'm not surprised anymore and find it boring...

To Tim : About formulas... There are different types. Sometimes you see some pattern, or read some paper about some math subject which can have interesting graphical applications, and decide to make a formula reproducing them. Then the result is completely predictable (well... if you program it correctly... :-D).

But in other ones, like the one I used for this image, you imagine a global concept which should produce interesting things, and then add a lot of features by trial and error just because they produce nice and unexpected results. (So again, very far from "pure science"...)

Actually, if I remember well, my original goal when I started writing this coloring was to reproduce some drawing I was making by hand. Take any self intersecting curve, and color the regions it separates with two colors (it's always possible). You get patterns that have a "Miro" feeling. But then it deviated completely, and it was a very good thing, because now it's infinitely richer than what I planned.

10/19/2006 3:06 PM

 
Blogger Paul said...

Hey Sam,

I like the 'cooking' analogy too. When I prepare a meal, sometimes it turns out good, other times (Most other times, actually), not so great. With me, the same is true for fractal creation.

A look at my site reveals that many of the 'ingrediants' that go into my own fractal recipes come from your formulas and colorings.

So, Thanks for your 'ingrediants'!

10/24/2006 12:49 AM

 
Blogger Sam said...

:-) You're welcome, Paul !

10/29/2006 5:17 AM

 

Post a Comment

<< Home