
Pages
Top:
Polygonal City 15, the finished print
Middle:
Documentation photo of the site
Bottom:
A screenshot of the scene in wireframe



Although these prints can be appreciated without any knowledge of the techniques used to make them, it is worthwhile, as with any other art form, to have a basic understanding of their production. This is also useful to clear up any misconception about the status of computer generated art and the prejudice that some people hold that it is somehow inferior because it is not handmade. As with any newly introduced fine art medium it takes time for the acceptance of the craftsmanship required and the recognition of its link with artistic merit.
One problem associated with explaining a digital medium is the opacity of jargon laden language. This is particularly true with such a nerdy process as 3D modelling which has its origins in the study of how light affects three dimensional objects.
The production of the Polygonal Cities series involves two software programmes, a bitmap editor such as Photoshop which is used to prepare the textures that are applied to the 3D models and a programme where the actual building of the models takes place. This software usually manages the addition of lights and cameras used to render the final image or animation. The process can be compared to a film set and is indeed the same used in the production of cinema special affects.
The models are made out of polygons - hence the series title Polygon Cities. Usually these are triangles that define the surface of the object and the software calculates how the light strikes these surfaces and where shadows will fall. The screenshot below shows the x-ray like quality of the meshes, that are the points and lines that define the structure of the triangles without their skin. It appears complex but the objects are commonly made by manipulating primitives which are simple readymade forms such as a box, cylinder, cone or globe. They can be scaled, distorted, added to and subtracted from to skilfully build quite elaborate objects.
This is basically how the Whopper Chopper, the children's plaything that dominates the left half of the print, was made. However, this method is inadequate to generate the curved tubes. This requires a process where a shape, in this case a circle, is swept along a path that defines the desired curvature. Once complete, all the individually made components need to be scaled to match each other and are then carefully aligned in three dimensions and grouped to form the finished object.
The freshly created objects are by default a dull grey colour and await materials which will bring them to life. The Whopper Chopper and the Kentucky Fried building are given straightforward materials that apart from a colour value only require specifying specularity which controls the size and colour of the highlights. Therefore the plastic appearance of the Whopper Chopper is determined by a moderately sized highlight and the absence of any specularity at all gives the paint work on the restaurant a dull chalky finish.
The concrete apron that anchors the composition would be quite lifeless if it were given such a basic material. A useful solution is to employ a procedural material which uses mathematical code to generate pattern. The Krushbar, Drive Thru sign and the Colonel Sanders sign use a third technique where a bitmap image is projected onto the form. An example is the giant tub which uses a jpeg image downloaded off the net and processed to snugly fit the form in a shrink wrapped fashion.
The technique of projecting images onto shaped surfaces, much like painting onto a cardboard cutout, is used to draw the woman, the slice of suburbia in the background and the sky. The advantages of working this way are compelling; the tediousness of building complex models is avoided but more importantly, and provided there is access to a large archive of images, is flexibility. Virtually all human figures and skies in this series make use of this procedure.
Although the work flow is not determined by a strict sequence, the final procedure once the models are textured, is lighting and composing. This print uses just a single sunlight source which emits rays in all directions and is configured to simulate the hard Australian midday sun. The composing is done through a virtual camera which allows the user to select the amount of telephoto or wide angle zoom - much like a real world camera.
Inevitably there is much final tweaking when test renderings reveal problems or a last minute insight prompts improvements. During the final rendering the computer traces the light rays from its source to the objects and paints a high resolution image which is saved and sent to a skilled printer who will massage it to bring out the widest possible range of tones and colours.
For those people who want to know the nuts and bolts, all work is done on an Apple iMac using Adobe Photoshop Elements 6 to edit images, Cheetah 3D to make and render the 3D models and Coda to write these web pages. The print editions are printed by Pharos Editions.