«Computer art' has become a meaningless term, because soon virtually all art will be computerized in some way or another. The author introduces the concept of proceduralism as a label to represent a special class of art, one that constructs images using abstracts qualitative and quantitative parameters, rather than simulate classical drawing and painting.  This approach to making art differs radically from drawing and painting approaches because the picture-making process is detached from the picture. The net result is that an entirely new area of creativity has been unveiled for the artist.  As such, proceduralism  is a logical successor to conceptual/process art; it is a major art movement and a new medium.»
Abstract of The Proceduralist Manifesto, by Judson Rosebush, published in LEONARDO, Computer Art in Context, Supplemental Issue, pp. 55-56, 1989.


Visual representations of mathematical objects are often beautiful, as if our brain perceived, thru beauty, the parametric equilibrium which gave its birth.  Artists of the Antiquity and Renaissance did search the mathematical laws of beauty. Today the process can be inverted: using a computer to make visible mathematical constructs.

In 1979 I used a programmable calculator TI54 to precisely plot the ellipse for a copper plate I had to cut. Suddenly, there was a new graphic tool. I had already studied perspective as it was invented in the Renaissance, but this method has limitations when working with objects that are not defined by right angles, like dodecahedrons.  So I conceived a 3D projection system on a more powerful hand held calculator, the HP-41CV, I tested it with a dodecahedron, then transposed the application on an Apple IIe. I could then test 3D projection algorithms different from the canonical projection modes used by 3D apps.

Apple IIe
Apple IIe

I had been impressed by the work of a Canadian, Gerald Bull, who worked a few miles from my place at the Valcartier military base, in research on high velocity guns.  He quit the base and started a business to build orbital guns.  His work reminded me the Jules Vernes' novel I had read in my childhood, Voyage de la Terre à la Lune, and I imagined that a larger gun than the one built by Bull would send in orbit stacks of pentagonal prefab walls, that, once assembled in space, would make orbital stations. The dodecahedron is probably one of the best structure for low gravity environments, and stacked pentagons could make compact loads able to stand the enormous acceleration required to reach an orbit.  So my dreams merged with my code and I created several etchings on that theme.

Village orbital
Village orbital

Also: 37DDCAnneauContrepointAutoportraitDodécahèdresHyperespaceStation OortVillage orbital

 Once I could purchase a Mac in 1994, I tried to pursue the same programming experiments I had done on an Apple IIe, but it was radically changed.  Before I had two books to program the Apple IIe, but then a complete bookshelf for my new Mac.  And with plenty of already written software to use, I begun to to play with Adobe apps.  Inspired by my passion for reading, I created imaginary spaces in two or three dimensions filled with text objects.

Certes... is inspired by the reading of À la recherche du temps perdu, Marcel Proust. 
Click image to enlarge

The conception of these images was analog, I was working with the mouse as if it was a brush whose color were letters, or as if I was making collages with Lettraset or 3D type, assembled by hand.  It was not the fruit of a logical process defined in a formula or a program.  So I returned to programming, to study the possibilities of compositions elaborated by accumulation of simple shapes progressively transformed in their shape, size, location and orientation.
Here is a collection of these images, arranged in a tubular composition
Click image to enlarge

All the previous synthetic images were in black and white, so I decided to experiment with color, trying to find harmony by an analogy with music.  I studied the graphs formed by the addition of the sinusoidal curves of the notes composing an harmonic chord .  I choose a complex image of the previous experiment, and made 1200 colorization experiments, but the principle of construction of this image, the accumulation of successive transforms of shapes, created very complex topologies inappropriate to study the relationship of colors with one another.  Some of the maths I used:

Click to browse

  • 0679-705
  • 0706-732
  • 0733-759
  • 0760-860a
  • 0760-860b
  • 0760-860c
  • 0760-860d
  • 0760-860e
  • 0897-924
  • 0924-950
  • 0951-977
  • 0978-1004
  • 1005-1031
  • 1032-1058
  • 1112-1138

 So I built an hexagon with 15769 small hexagons regularly tiled, colorized according to 1288th color experiment of the previous study.  Only a dozen colors were used twice, the remaining 15,750 colors were used only once, unique in the picture. This is the very opposite of the musical theory were only a small set of predefined sound frequencies are used and repeated.    HexaHexa1288  

The small hexagons that made the large one were in fact like a very long string of hexagons, circling around the centre like a spiral,  their color relations were sequential, not really spatial.  The first experiment was topologically complex, this one was simple, despise the number of objects.  I was working then with simple programming tools, mostly long scripts driving other applications, they were slow, the data structures I could use were limited.  So I went back to native programming this time with  Xcode 3 and Cocoa, and built another application, the base for other experiments, where I can address the color problem on a truly spatial basis.  

Trou de mémoire is one of theses results.

 Click to enlarge 


 More results of the ongoing experiences are available at the chronological gallery.