Digital fragmentation using 68 columns (i=2)

December 2015

The rise of the computer introduces data as a new language, as Boolean logic is the foundation of the ways computers “think.” While computers are not living creatures, we can “speak” to computers through user interfaces or through various levels of programming (though the definition of “life” itself has become more and more debatable with improved technology). We can also use computers to translate directly from one sensory language to another, as is the case with audio visualization.

Neo-Neo-Impressionism is a concrete realization of this concept. A picture on a computer is understood visually by a human user, but to a computer, this image is just a set of numbers. I wanted to represent the image as a computer understands it through data, but also in such a way that the viewer can understand what the computer “sees”. The viewer will see image as form and color, but also the actual numbers the computer uses to interpret it.

Digital fragmentation at 17 columns (i=0) and 544 columns (i=5), and Georges Seurat's original painting (Courtesy of Wikipedia Commons)

The Processing program I developed takes in the image, slices it into a grid of cells, and calculates the average color of each cell. In different iterations, the number of columns, and therefore adapted resolution, is increased exponentially by two; and so the generated grid more closely resembles the original image. Each cell is represented both in actual text and the color of said text, and is "stamped" on the blank canvas. The background color of the canvas is the average color of the entire sampled image.

Comparison of 17 * ( 2 ^ i ) columns, where i = [ 0, 1, 2, 3, 4, 5 ] The final print is a digital recreation of Georges Seurat’s Circus Sideshow. I chose this image when I realized that my process was similar to that of the Pointillists. Here, as the Pointillists would argue, I am not just starting with a painting, but a set of dots that make up a painting. These dots are then refragmented digitally. This image can be fragmented perfectly into 17 columns (i=0) and 172 rows on a 2000px wide canvas, where each cell is just barely touching its neighbors. Iterations of this image increase the number of columns by a power of two. Rows are not increased to preserve legibility. The final print contains 544 columns (i=5), or , and measures 44” x 29” at 150dpi.