Rendering consists of taking the digital attributes of the model, the lighting and the camera and creating an image. Rendering is a complex process and requires much more computer power than the modeling stage. Before expending the time and money to render an entire shot, the computer artist may wish to render single key frames of an animation sequence to check that the simulated image is the desired one. The artist may also render wire frame or low-resolution approximations of the shot to get a feel of the look of the animation before fully rendering the scene.
Because the objects in the computer-generated scene are only simulations, they act quite differently from real-world objects that must obey the rules of physics. If not properly animated in three dimensions, computer objects may interpenetrate one another, destroying the illusion of solid, real objects. If not properly constructed, seams may show between supposedly seamless parts. The artist may discover unwanted artifacts created by the size and shape of the pixels, the scan lines of the monitor, or errors in texture mapping or surface generation for the first time in the rendering process. The modeling and rendering cycle is often an iterative and interactive one, with the CGI designer returning to the modeling stage to correct problems that can only be detected after rendering.
Final rendered images can range from simple wireframe approximations of objects, to highly faceted objects, to realistic smooth shaded objects. The style in which an artist renders an image is often a factor of aesthetics tempered by the pragmatism of meeting a production deadline or budget constraints.
During the rendering process, the computer may also control a seamier to digitize film frames and to composite them with the computer-generated images.
Was this article helpful?