About 3D Graphics
Three-D graphics require much less storage than their 2D counterparts because the 3D pictures are not pre-drawn (with the exception of texture maps).
The "recipe" for the 3D picture (meshes and animation) is stored as a mass of formulas and called up when needed.
Because the pictures are being drawn onscreen by the program as they are being seen, and not before, the computer must "think" much more and much faster than it does with 2D images.
Imagine the difference between someone pulling nicely arranged pictures from a stack, and someone else trying to accurately draw, at the same speed as the person who is pulling pictures, a collection of objects that yet another person is moving around.
Imagining such a scenario should help you easily grasp the difference in what is demanded of a machine running a real-time application.
Only the current high-speed processors are capable of meeting these extreme demands.
Even then, the geometry being drawn must be simple and have a low polygon count to make the process fast enough to meet acceptable display speeds.
Real-Time 3D Basics: Modeling for real-time graphics is a delicate process.
You must have an accurate picture of what the result will be after the object being modeled is exported into the real-time engine.
The more you know about how the average real-time engine thinks, the better your initial efforts will be, and the more time and frustration you will save yourself.
Real-time 3D and high-end, pre-rendered 3D graphics have many elements in common.
To achieve the speed necessary for presentable game play, however, real-time must use only the most necessary elements-namely, the geometry, the transform, and the surface properties of the mesh.
Most of the time, these elements are created by the export program (a third party application that converts the source model into a language the game engine can read) and put into some kind of text file (or a "c" file, before compilation into binary code) so that they can be manually edited, if necessary.
Sometimes these elements can be parceled out to a number of separate files (one for geometry, one for surface properties, and one for the transform) that are combined when the file is compiled for the game engine.
The "recipe" for the 3D picture (meshes and animation) is stored as a mass of formulas and called up when needed.
Because the pictures are being drawn onscreen by the program as they are being seen, and not before, the computer must "think" much more and much faster than it does with 2D images.
Imagine the difference between someone pulling nicely arranged pictures from a stack, and someone else trying to accurately draw, at the same speed as the person who is pulling pictures, a collection of objects that yet another person is moving around.
Imagining such a scenario should help you easily grasp the difference in what is demanded of a machine running a real-time application.
Only the current high-speed processors are capable of meeting these extreme demands.
Even then, the geometry being drawn must be simple and have a low polygon count to make the process fast enough to meet acceptable display speeds.
Real-Time 3D Basics: Modeling for real-time graphics is a delicate process.
You must have an accurate picture of what the result will be after the object being modeled is exported into the real-time engine.
The more you know about how the average real-time engine thinks, the better your initial efforts will be, and the more time and frustration you will save yourself.
Real-time 3D and high-end, pre-rendered 3D graphics have many elements in common.
To achieve the speed necessary for presentable game play, however, real-time must use only the most necessary elements-namely, the geometry, the transform, and the surface properties of the mesh.
Most of the time, these elements are created by the export program (a third party application that converts the source model into a language the game engine can read) and put into some kind of text file (or a "c" file, before compilation into binary code) so that they can be manually edited, if necessary.
Sometimes these elements can be parceled out to a number of separate files (one for geometry, one for surface properties, and one for the transform) that are combined when the file is compiled for the game engine.
