Seamingly Invisible

Seamingly Invisible

Are we aware of our disaster?

'Seam-carving' is a term which refers to an algorithm for content-aware image resizing originally developed in 2007 by Shai Avidan and Ariel Shamir. It was intended to make large images viewable on small screens. Prior to this technique, the only ways to make an image smaller in one dimension were by scaling, which would distort the image, or cropping, which may remove important information from the picture. In contrast, seam-carving works by finding paths of least importance – vertical or horizontal sequences of pixels which are visually less important than other pixels in the image – and removing them. Visual importance is established by the relative contrast of a pixel to its neighbors. A dark pixel neighboring bright pixels is considered important, and so is preserved, while a darker pixel among other dark pixels doesn't add as much visual information to the scene, and is thus more likely to be removed. This approach is referred to as 'content-aware' because the algorithm and its output are dependent on the particular content of the image being processed.

What does this algorithm have to do with disaster? Are we content-aware of the earth on which we stand, the water in which we swim, the air in which we sigh (sein)? The Cabot-Koppers Superfund site in Gainesville, Florida is a disaster in process. Over decades, starting in 1912 it was polluted by creosote, dioxins, benzene, napthalene and chromated copper arsenate, chemical byproducts of the production of pine-tar, an essential component of industrial wood treating and preservation techniques. It became a Superfund Site in 1983, and ever since then has been the center of a protracted legal, social, cultural and political battle between those who have been affected by the environmental damage to the site, and those who are responsible for its cleaning and remediation.

Where and when does this dramatic, deadly struggle enter the visual system? There have been protests, marches, rallies, newspaper articles, fliers, lectures and more, all temporary and ephemeral in the temporal scale of the conflict. Victims have been born, lived and died in the slow-motion wreck of this disaster. When I visited the site with some fellow artists and activists in 2015, I was struck by the invisibility of the catastrophe. The already crushing anxiety of the 'uselessness' of art, especially in the face of such a tragedy, was compounded by the event's apparent non-presence. You can't take a picture of the site – any visual evidence, removed from the context of the site, decayed into a flat image of an anonymous field. Green grass, a pond, or a dirt road. How do you represent the invisible?

Koppers is an intersection point in an array of massive systems in constant often imperceptible motion; environmental, geological, chemical, economic, societal, cultural, and political. Creosote and arsenic slowly trickle through the layers of earth to the Floridan Aquifer. Awareness is raised, and parties reach out. Deals are struck. Topsoil is removed, spread, and contained. Samples are taken, results disseminated. Although the site resists capture and representation, its phantom is everywhere present.

Seamingly Invisible is a computational artwork for content-aware, abstract imaging of invisible systems. On September 16th 2015 a drone piloted by Steve Rowell flew above the city of Gainesville and captured video of an empty field, the Cabot-Koppers site. I took excerpts from this video and processed it through a seam-carving algorithm, making the process of erasure of each pixel visible. As the seams are removed, evaporating into the air, they reappear in a neighboring image, persisting, trickling down the screen. Over a duration of hours, the image dissolves, developing into a tracing, a map of the visuality of the site. Objects recede and grounds emerge over time, according to the judgment of what is important, what is valued and what is not. Information decays and is lost, buried, replaced by a phantom. The work is a processual, computational metaphor and rendering of what it is to be with disaster.