Getting a Job in CG: Real Advice from Reel People, Chapter 3: What to Learn. By Sybex | 4
Getting a Job in CG: Real Advice from Reel People, Chapter 3: What to Learn.
|HOW TO GET AHEAD IN MODELING
Modeling is often the first thing people get into when they start up with CG. Because so many people get into modeling it’s a good idea to get really proficient in certain kinds of models, be they creatures, cars, humans, etc. Having a good sense of design also helps, as you’re sometimes required to design parts of your own models, or even to come up with several versions of a single idea. Here having a good drawing skill will also be helpful to you.
Modelers are usually the first group of people hired for a film project,
and quite a few get to stay on to texture or light if they have the proven
skill for it. Being a successful modeler in film, television, or games means
you have an eye for detail, a fundamental understanding of how to model
with different surface types, and the ability to quickly actualize tests
Modeling is only the first step of creating a 3D object. Shading and texturing are the steps that attach a color and surface qualities to the 3D skin. Textures are surface materials that have been painted, photographed, or otherwise pregenerated as images that are then projected or wrapped around the surface of a 3D object. Textures can be used not only for colors, but also to apply underlying surface qualities, such as specularity, transparency, bumpiness, and glow.
Shaders are the mathematical code (in Maya it also refers to the network of nodes used in texturing an object) that tells the renderer how to render a surface A shader may be as simple as a set of effects that modify the appearance of a texture—making it shiny or rough, for example—or as complicated as elaborate surface materials, such as scales, feathers, or fur, that completely replace painted textures (Figures 3.6 and 3.7).
Figure 3.6 A shading network in Maya is used to create materials and textures to map to models.
Figure 3.7 The shading network is applied to the legs and body of the elephant model.
Current games rely primarily on simple color maps, although advanced rendering engines in high-end gaming PCs and 3rd-generation consoles, like the Xbox, are allowing for the use of more sophisticated shaders and texturing effects. Already, it’s common to see transparency, specularity, and bump maps in games, and future games will also feature normal maps, which simulate very detailed geometry while still using simple underlying polygonal meshes.
To work in games, you’ll have to understand UV mapping (the assignment of sensible surface coordinates to a polygonal mesh that may bend and fold on itself in countless ways) and painting of textures and texture effects. You’ll also have to learn how to work with lighting and to bake lighting into material textures. With the increasing prevalence of global illumination and radiosity rendering in 3D packages, it’s also becoming more common to see radiosity solutions rendered and baked into texture maps.
In television and film, Mental Ray (supported by 3D packages such as Maya, SoftImage XSI, and 3ds max), and Pixar’s RenderMan (also supported by Maya and SoftImage XSI), have been industry standard renderers for a number of years. Working with these renderers requires knowledge of the texture and shading techniques that work with them. For any film work, you’ll need to understand how to create realistic-looking surfaces with color maps and a wide variety of effects maps on a very high resolution and detailed level. Television work tends to be somewhat more forgiving due to its lower frame resolution, though textures do need to look extremely detailed with photo-realistic flare.
You’ll need to master UV mapping, projection mapping, and intrinsic NURBS mapping. And since you’ll probably be dealing with different surface types, it’ll be very beneficial to understand modeling techniques for polygons and subdivisions as well as NURBS.
Lighting requires a keen eye for color and tone, as well as an understanding of the technicalities and limitations of 3D lighting systems and renderers. If you’re aspiring to be a 3D lighting artist, you’ll want to study photography, which is primarily a study of light and how it interacts with surfaces and film, and cinematography, which deals with lights in motion. Lighting is also an important component in the study of architecture.
KNOW YOUR TOOLS
A texture artist, more than any other CG artist, needs to know a handful of applications outside their primary 3D package. You need to know image editing techniques and programs like Fractal Painter, Photoshop, and plug-ins that allow you to paint directly on a 3D object like Deep Paint, and be good and damned fast at them too. You’ll be switching back and forth between painting images and your 3D model all the time, so it’s a good idea to get to know your image editing applications pretty well. The keyword here is Photoshop.
Lighting in 3D requires a mastery of the basic light types—directional, point, ambient, area, and spot lights—as well as an understanding of how lights interact with surfaces and the creation of shadows in 3D. Lighters should also have a firm grasp of how textures will be affected by their lights. Some lighting artists in some film studios are also asked to fix or create textures for their scenes or even to composite their shots into the live action backgrounds. Knowing color and how to get the right look for a frame is imperative to a lighting artist.
If you want to master lighting in 3D, start by turning off all ambient lights and relying on point, spot, and directional lights to illuminate your scene. It’s almost impossible to create realistic or interesting lighting with any significant 3D ambient light in a scene because ambient light evenly increases the brightness of every diffuse surface, as if they were all glowing with subtle radioactivity. Once you’ve lit your scene with point and directional shadow casting lights, adjust the color and opacity of the shadows, and only then consider adding an ambient light to uniformly raise the values in the shadow areas, or leave out ambient lights entirely to better mimic real world light. There is no such thing as ambient light in reality.
CG AND LIGHTING
In film, lighting in CG is frequently regarded in photo-realism work as the most important part of the CG pipeline. It is lighting that will make or break a CG scene when it has to fit into a live-action background or plate. If the lighting is even slightly off from matching the real lighting of the plate, the shot is thrown because the CG is obvious. The job of a photo-real lighting artist is to make it look as if CG doesn’t exist in that scene. In other words, the shot should look so realistic that you’ll have a hard time convincing people that a CG worked on it at all.
In real-time 3D games, lighting effects are severely limited by game engine performance budgets, so lighting requires many creative workarounds to achieve artistically pleasing results. Lighting is often relegated to environment artists, who compensate for the limited number of live lights in any scene by prerendering scenery to bake lights and shadows into texture maps. As Emmanuel Shiu points out, knowing lighting is beneficial to games but critical to film work:
So it’s always best in the game demo, or game portfolio, to show that you’re multifaceted. Maybe you’re not the best at lighting, but it’s good to show that you can do some lighting, so maybe they might be able to train you.
But in film, that does not apply. They want only the best in that field. The sharpest knife. They don’t care if you can do a little bit of lighting because they don’t need a little bit of lighting, they need really good lighters.
Because of the considerable complexities involved in lighting film projects, lighting is often the domain of technical directors with acute color sensibilities. When you’re lighting a CG set or props, they often have to blend with live plates, and achieving lighting that matches the scene requires balancing not only the direction, color, and diffuse values of light, but the play of shadows and highlights on surfaces. Lighting is tightly intertwined with the complexities of rendering systems, and as these become more advanced, with capabilities like radiosity and global illumination, lighting also becomes more complex. For example, High Dynamic-Range Radiance Imaging (HDRI) rendering extracts incredibly realistic lighting information automatically from a scene’s environment maps, but it requires painstaking preparation of an environment to achieve useful results.
Created: March 27, 2003
Revised: April 9, 2004
URL: URL: http://webreference.com/3d/cg/1