GA2: a Programming Environment for Abstract Generative Fine Art

Philip Galanter

Interactive Telecommunications Program: New York University

e-mail: galanter@nyu.edu

Fine artists looking to use computers to create generative works, especially those artists inclined towards abstraction, often face an uncomfortable choice in the selection of software tools. On the one hand there are a number of commercial and shareware programs available which implement a few techniques in an easy to use GUI environment. Unfortunately such programs often impose a certain look or style and are not terribly versatile or expressive. The other choice seems to be writing code from scratch, in a language such as c or Java. This can be very time consuming as every new work seems to demand a new program, and the artist's ability to write code can seldom keep pace with his ability to imagine new visual ideas.

This paper describes a software system created by the author called GA2 which has been implemented in the Matlab software environment. By layering GA2 over Matlab the artist can take advantage of a very mature programming environment which includes extensive mathematical libraries, simple graphics routines, GUI construction tools, built-in help facilities, and command line, batch mode, and GUI modes of interaction. In addition, GA2 is very portable and can run on Macintosh, Windows, and Unix systems with almost no incremental effort for multi-platform support.

GA2 is a work in progress and an extension of the completed GA1 environment. It is medium independent, and can be used for all manner of image, animation, and sound production. GA1 includes a complete set of genetic algorithm operations for breeding families of graphical marks, a database function for managing and recalling various genes, a set of statistical operations for creating various distributions of marks on a canvas or animation frame, a unique Markov-chain-like operator for generating families of visually similar lines or paths, and a complete L-system implementation. GA2 extends GA1 by adding more generative techniques such as tiling and symmetry operations, Thom's cusp catastrophe, and mechanisms inspired by complexity science notions such as cellular automata, fractals, artificial life, and chaos. All of these techniques are encapulated in genetic representations.

This paper is supplemented with examples from the authors art work, and comments on the philosophy behind this method of working, and its relation towards the reinvigoration of abstraction after post-modernism.        

Foundations of Generative Art: a case study in teaching art oriented programming

Philip Galanter

Interactive Telecommunications Program: New York University

e-mail: galanter@nyu.edu

The Interactive Telecommunications Program is a well known professional master's program for artists interested in new media, and is part of the Tisch School of the Arts, often called the "NYU Film School". ITP graduates are very much in demand, and can be found in creative leadership positions throughout the multimedia industry.

This year a new class, "Foundations of Generative Art", has been added to the curriculum. In this paper the instructor reports on the challenges and rewards in teaching topics from science, mathematics, and computer science to a non-technical audience of creative students. Our experiences with 2 unique programming environments, Design by Numbers and StarLogoT2001, as well as samples of student work will be included.