EVOGENIO® - Evolutionäre Kunst

BasicLevel - EvoArt - EvoArt1 -  EvoArt2 

Dr. Günter Bachelier - My Evolutionary Art processes

img12The second level of my artistic work is the methodical level, where I transfer evolutionary concepts and methods to develop evolutionary art processes that enable me to create non-representative works of art that fit my aesthetic preferences. I view an art process as a sequence of actions by an artist or group of artists whose outcome is a work of art (image, sculpture, ...) or the process itself is a work of art (performance, dance, ...).

In this context, an evolutionary art process can be defined as an artistic process where the three evolutionary concepts population, variation and selection are used. This does not necessarily mean that the whole process is computerized, or even that a specific evolutionary algorithm is incorporated, because evolutionary art can be done with paper, pencil and a dice. For instance, William Latham was doing evolutionary art on paper, long before his ideas were converted into a computational approach {Todd1992}. Such a definition includes cases where an implemented evolutionary algorithm is only a part of the whole process, specially if the outcome of the algorithm  can be optimized by the artist. If the desired output of an art process is a complex image, it is natural to use the aid of computers to generate image individuals, otherwise the change of generation would take too long. The resulting evolutionary art process might be style and material independent.

The background for this development dates back to around 1995, when my scientific interests shifted from Artificial Neural Networks {RitterMartinetzSchulten1992} and specially self-organizing maps {Kohonen1995,Bachelier1998b} to Evolutionary Algorithms and particularly Evolution Strategies {Rechenberg1994,Bachelier1998a, Bachelier1999}. I  first became aware of Evolutionary Art when reading the exhibition catalog of Ars Electronica, 1993 about Artificial Life and Genetic Algorithms {AE1993}.  I realized immediately that this could be a crossroads, a junction, bringing together my research and artistic interests.

In the first stage I developed an integration of my conventional art process, (self-organizing painting), and an evolutionary art process that uses the basic concepts: population, variation and selection. I used the digital  images in my archives for the initialization of the whole of my early evolutionary art runs and I evolved these images over some years (see fig. 1 for example).

Fig. 1: Image individual from the end of my first evolutionary art  phase: E-2001-15
E-2001-15


I also used the results for post-processing operations -- such as the  creation of abstract panoramas (see fig. 2) which were developed  until 2003. They can be viewed on the web (http://www.vi-anec.de/Trance-Art/Gesamt-Panoramen.html}) with a specific panorama viewer. A physical  transformation of this kind of art is difficult because panoramas often have a height:width  proportion of 1:10.


Fig. 2: Example of an abstract panorama: P-2000-017
P-2000-017

 
Around 2002, I realized that this kind of evolutionary art process was ending. So, in the context of my contrast aesthetics, I began to play with the idea of how to combine contradicting approaches (such as geometric,  constructive and concrete art) with the images derived from the non-representational art of self-organizing painting. The combination of geometric shape and non-representational content opened a new dimension to my work and the combinatorial properties of evolutionary art were suitable for exploring this new dimension.

Geometric forms naturally lead to questions about symmetry, so I combined this development with the wish to work with biomorphic shapes by applying the supershape formula {Gielis2003} in many of my new works. Fig.3 shows some examples from the repository of masks generated from supershapes.

Fig. 3: Examples of supershape masks
SSHAPES



References


BasicLevel - EvoArt - EvoArt1 -  EvoArt2