RENDERING PROCESS:

• Introduction
• Principles
• World structure
• Projection
• Finding walls
• Horizontal intersections
• Vertical intersections

MISCELLANEOUS:

• Glossary
  
• Downloads
  
• Links
  

AN INTRODUCTION TO RAYCASTING:
Raycasting is the term given to the technique we'll be using to render a 3D scene. To give you an idea of the visual appearance of a raycasting engine, I'll show you this Wolfenstein 3D screenshot in fig1. The Wolfenstein engine was one of the first real-time 3D engines, and id's lead programmer John Carmack may well have been the person to start the raycasting craze.

fig1: Screenshot from Wolfenstein 3D

As you can see by the image, the output is rather blocky. The visual effects aren't as realistic as those produced by games like DOOM, Quake, Unreal etc., and the engine's not as fast as theirs. But raycasting is a simple technique that's simple to code -- that's what makes it a good starting point.

Raycasting is a variation of another rendering technique called raytracing. Raytracing is unsuitable for real-time applications because it's just too slow. It involves tracing a ray from each pixel of the screen out until it hits part of an object. The part of the object is then drawn according to the distance the ray travelled and the point of the object that the ray hit. Afterwards, the next pixel is taken as the new starting point and the process repeats itself. If you had a screen in a resolution as low as 320x200, this would mean (320 * 200 = 64000) rays to be traced ! As tracing a ray involves extensive calculations, it's an extremely slow process. I wonder how long a 1024x768 screen would take...

However, the effects of raytracing are stunning and virtually any shape can be rendered. Take a look at fig2, which shows a screenshot from The 7th Guest. Players of this game must stick to a pre-determined path because the images needed to be pre-rendered to allow for the speed and smoothness. This is the exact opposite of Wolfenstein, where you may go anywhere (within reason) and the images are generated dynamically (or 'on the fly').

fig2: Screenshot from The 7th Guest

Raycasting, as I've already said, is a variation of the raytracing method. It gets its speed from various 'geometrical constraints'. That's just a fancy way of saying that the world a raycaster is expected to render follows some basic rules, such as all walls are fully vertical (no slopes are allowed; the walls are all perpendicular to the floor). Because all walls are at 90 degrees to the floor, rays can be cast for every screen column rather than for every pixel as done in raytracing. This means that, for a 320x200 screen, instead of casting 64000 rays, we'd only cast 320, which is a significant improvement. It may surprise you (or not) to learn that raycasting is an entirely 2D process (as will be proven later), which allows everything to be simplified even further.

Although raycasting is an old, and greatly obsolete, technique, it gives a good starting point for 3D because it provides an introduction to simple algorithmic programming. And because raycasting uses simple trigonometry, the algorithms are fairly straightforward to implement in code.

CONCLUSION:
This section was a rather brief introduction to raycasting, but it was simply to give you a vague outline of what the technique entails and what can be accomplished, without delving deeply into the topic (there's plenty of time for that later :). Let's move on to the first step -- Understanding the Principles Behind Raycasting. The first few steps are basically introducing you to raycasting, so no formulae for a little while.

NEXT PAGE