If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

1. What is ray tracing?

Susan Fong from Pixar explains how a virtual camera and image plane create a scene. For each pixel in a scene, a ray (think of it like a laser beam) is sent from a virtual camera through that pixel into the scene. The ray then checks out what it hits first in the scene. The color of the first object it hits determines the color of the pixel.

Want to join the conversation?

Video transcript

Hi, my name is Susan Fong and I'm a rendering technical director here at Pixar Animation Studios my job is to make sure that the images you see in the theater are created as quickly and efficiently as possible now that's a big job because of Christophe described in the previous video, rendering is an extremely complicated process requiring thousands of computers all working together to render our films. The render farm is where we keep all these computers, but what are these computers doing exactly? That's the topic of this lesson. In the first part we'll take a geometric look at a renderer works. In the second part of of the lesson, we'll look at some of the equations and algebra that are needed to write a render as a computer program. So let's get started! Let's start with the drastically simplified scene from Up. This object lives in three dimensions, as we can see as we navigate around it. Just like live-action filmmaking, we need to place a camera inside the scene. In computer graphics we call this a virtual camera. We first pick a location for our virtual camera they'll be used to create the image. It's shown here in white. We also need to decide which direction our camera is pointed. That's shown here in red. Next we erect a plane upon which we'll form the image. It's called the image plane, and represents what the virtual camera will see. Notice it is perpendicular to the viewing direction. The image plane is broken up into a grid of pixels that will make up the image. In this interactive the first two sliders allow you to rotate your camera around the scene. The job of the renderer sounds simple. It must determine a color for each of these pixels. This is kind of like paint by numbers where the image plane is serving as the canvas on which the renderer is painting. Now I'm only showing you a few pixels in this example. Our real images are made up of more than two million pixels. Over the years a number of different rendering algorithms have been developed. one of the most elegant and versatile algorithms is called ray tracing. Here's how it works: to determine a color, for say this pixel, a ray tracer constructs a mathematical ray that starts at the camera, crosses the center of the pixel, and then extends off into the scene. Kind of like a laser beam pointed at the pixel but from the camera. The ray tracer computes the closest intersection point between the scene objects and this ray. It then determines the color of the intersected object at that point, and paints the picture that color. So this pixel would be colored brown, this one yellow, and so forth for the other pixels. This process is done for all of the pixels in the image. Fancier versions of ray tracing can create more subtle effects, like soft shadows and variations in lighting that helped to create different moods. The ray tracing process leads to a bunch of other questions, like how are intersection points determined? And once that intersection point is determined, how do we figure out what color it is? These mysteries and more will be revealed in the remainder of the tutorial. In this first exercise, you can practice some ray tracing yourself.