Rendering Engines

Jean Nouvel’s former project architect once critiqued my work. She spat in frustration, “Do not show an architect a perspective rendering. They are lies and thus should only be shown to clients,”
Rendering is a tool designers must use, not for themselves but to sell their idea. The goal is to make a creation beautiful in a way humans perceive as real (or virtual reality). The challenge is making it truthful.

Through this exploration of rendering engines, we chose a tactic of attempting to achieve the same ends with different means. The goal was to create a simple scene with a calculated source of light and two objects; a sphere of reflective aluminum and a cube of wood. There is a multitude of engines available to us, mostly falling under the categories of gaming rendering and graphic rendering. We studied the graphic side with ray tracing systems, testing Autodesk Revit, 3DS Max, Rhino, and Maya.

3DS Max in particular is the software I chose to explore, mainly because it was the program of which I was most ignorant. It is frequently used by video game developers, TV commercial studios and architectural visualization studios as well as for movie effects and movie pre-visualization. This Mental Ray engine can be used for 3D modeling, animation, and rendering with unique features such as customized particle animation and perspective matching. Objects can be created through polygon modeling, adding detail through tools such as “bevel” and “extrude”. NURB surfaces are also an option in this engine. This makes the program more compatible with Rhino and AutoCAD programs, which function in a similar way.

Material scripts and varying light instances, fixtures, and apertures are easily controlled in generated scenes. Perhaps the most important of these are light, owing to its innate ability to the change color, texture, accessibility, and atmosphere of an entire site so drastically. The type and quality of light can be chosen and added to a scene. Materials are applied to geometries through drag and drop assignment. Exploring properties can control an object’s tiling, mirroring, decals, angle, rotate, blur, UV stretching, and relaxation.

We found, through very elementary exploration, that many factors can be controlled and the smallest details can change a rendering drastically. Unless one is wholly proficient in a single program, there will always be a mixture of programs used to create a rendering. We found that we wouldn’t use one of our renderings “raw”. Every rendering we found elements that were out of our control and would probably need to be fixed in Photoshop. It is extremely difficult to determine how we view the exterior world and apply it to a computer screen. Can a scripted program truly replicate the color of light on a pure white surface? Can it display the layers of the color of a sunrise on a landscape?

Even if it can create something indistinguishable from a photograph, how much truth is in it? It may be more honest to show a rendering done by hand, to be taken as an art piece and an interpretation of the human experience rather than painstakingly attempting to recreate reality.

Comments are closed.