The Reconfigured Eye: Visual Truth in the Post-Photographic Era. - book reviews

Art in America, Jan, 1994 by Fred Ritchin

After lazily believing for 150 years that there is nearly objective truth in photography, our society is now entering a digital age in which photography's hegemony as reliable witness is ending. Photographs will soon be as easily simulated as "taken." Yet we never quite got around to discussing adequately what photography was all about or determining what it actually meant for us. Was it ever a reliable witness? In which ways was its mechanical approximation (or artistic subversion) of human seeing helpful? There is still insufficient public appreciation of the complexity of the photograph, of its ambiguity, efficacy and subjectivity. The most mechanistic and universal of the media, photography is burdened with the myth that "it never lies," and its verisimilitude is widely relied upon. But at the same time, the medium is often attacked (by American visual artists, among others) for its unmitigated mendacity. Without discussion of such issues, we--the alleged beneficiaries of the information age--are as unprepared for the digital revolution as previous generations were for the horseless carriage.

Into this confused situation comes a logical, clearly written, amply illustrated book, The Reconfigured Eye: Visual Truth in the Post-Photographic Era. its author, William J. Mitchell, dean of the School of Architecture and Planning at MIT, attempts to prepare us for the next paradigm shift by focusing on the subject of human seeing. His method is to compare the ways in which things seen have been realistically depicted by painting and photography with the depictions produced by their successor/competitor, digital imaging. The Reconfigured Eye, which is itself configured as a sophisticated textbook, contains chapters titled "Electronic Tools," "Digital Brush Strokes," "Virtual Cameras," "Synthetic Shading" and "Computer Collage." These provide technical and esthetic explications of processes that both simulate and amplify photography. For the most part, Mitchell discusses tactics that produce a realistic image while modifying some of the givens of the photographic process.

There are two major strategies that imitate and expand upon the "photographic"--one is to digitally combine extant imagery, primarily photographs, to create new images (generally with the aid of a "paint" program that, among other things, allows the user to become a seamless retoucher), and the second is to generate new photo-realistic imagery mathematically. In the first case, one or more photographs are scanned, digitally encoded and presented on a screen; then, using appropriate software, it is possible to make changes in those images by altering anywhere from one to several million pixels (the picture elements which are the building blocks of the image). Retouching devices that simulate conventional tools--e.g., an eraser or a brush--are available, as well as devices that permit the use of newer techniques, such as copying a color from one section of an image onto another area. It is possible to "paint" the background, or any other part of the image, as if one were painting a simulated surface. (Mitchell mentions this last technique as a separate strategy.)

In the second major approach, which begins with algorithms rather than a photograph, the computer is given a set of instructions about the size and look of an object, or a set of objects, and can then generate an image which is "lit" by a simulated light source. Thus the image will appear to be seen from a specific point of view--that is, by a "virtual camera" (and by a "virtual observer"). This technique is obviously very useful for filmmakers, who can, for example, simulate a spaceship, shoot it up into orbit and blow it up--all without incurring the cost of physically building and destroying the vehicle. The more natural-looking parts of a scene are the most difficult to simulate--a human face, a landscape--while Coke cans and spaceships are the simplest, reflecting some of the biases of contemporary mathematics.

Mitchell's strength is his systematic, logical analysis of the factors that divide previous and future media. For instance, he provides an illuminating discussion of the differences between the analog photograph and the digital image by comparing the characteristics of grain versus those of pixel. As he does throughout much of his book, Mitchell liberally quotes major figures in the history of photography to make his point. On the issue of grain, he cites the great modernist photographer Edward Weston, who found that two characteristics constituted the trademark of the photograph: "First there is the amazing precision of definition, especially in the recording of fine detail; and second, there is the unbroken sequence of infinitely subtle gradations from black to white." Both, according to Weston, "pertain to the mechanics of the process and cannot be duplicated by any work of the human hand."

As Mitchell observes, a critical difference between the photograph and the digital image occurs at the spatial building-block stage; images are encoded digitally by dividing the picture plane into a uniform "Cartesian grid of cells," or pixels. The intensity or color of each pixel is specificed "by means of an integer number drawn from some limited range. The resulting two-dimensional array of integers (the raster grid) can be stored in computer memory, transmitted electronically, and interpreted by various devices to produce displays and printed images." In digital images, "fine details and smooth curves are approximated to the grid, and continuous tonal gradients are broken up into discrete steps."