Introduction

Tom Kelliher, CS 320

Feb. 2, 2000

Administrivia

Announcements

Assignment

Chapter 2.

From Last Time

Graphics system architecture.

The synthetic world.

Outline

1. Graphics: a synthetic world.

2. The programmer's interface.

3. A pictorial history of graphics.

4. Modeling-Rendering Paradigm.

5. Graphics Architectures.

Coming Up

OpenGL programming.

A Synthetic World

Previously: object/viewer, light, ray tracing, human vision.

The Pinhole Camera

Properties of a pinhole camera:

1. A perfect pinhole: only one ray from each point passes through the pinhole. Yields infinite depth of field.

2. Depth.

3. Width and height of the film.

Compute the projection of a point onto the film.

Compute the field of view.

What are the advantages/disadvantages of a lens?

The Synthetic Camera Model

Based on our pinhole camera, the model used in computer graphics.

Properties:

1. Lens

2. Bellows, allowing depth adjustment.

3. Width and height of the film.

Development:

1. Projection on the film plane may be ``reflected out front.'' This is called the projection plane.

2. Center of projection: center of lens.

3. Image on the projection plane consists of all points visible and in field of view of center of projection.

4. A clipping rectangle may be used to reduce the field of view.

The Programmer's Interface

An API must allow us to specify:

1. Objects --- collections of polygons. API primitives:
   1. Points.

   2. Line segments.

   3. Polygon outlines.

   4. Filled polygons.

   5. Curved surfaces (not supported by all APIs).

      OpenGL example:

      glBegin(GL_POLYGON);
         glVertex3f(0.0, 0.0, 0.0);
         glVertex3f(0.0, 1.0, 0.0);
         glVertex3f(0.0, 0.0, 1.0);
      glEnd();
      

      specifies a triangle.

      Adding additional vertices allows arbitrary polygons.

      Other type parameters: GL_LINE_STRIP, GL_POINTS .

2. The viewer --- through the synthetic camera:
   1. Position.

   2. Orientation.

   3. Focal Length.

   4. Size of film (projection) plane.

3. Light sources:
   1. Location.

   2. Direction.

   3. Strength.

   4. Color.

4. Material properties --- transparency, etc.

A Pictorial History of Graphics

Refer to Plates 1 through 8 in the text.

1. Wireframe images. Complex 3-D objects.

2. Hidden surface removal and flat shading. Lost 3-D. No light source.

3. Constant shading. Found 3-D. Light source. Still see the polygons.

4. Smooth shading. Polygons less obtrusive.

5. Texture mapping.

6. Fractal mountains.

7. Fog.

Modeling-Rendering Paradigm

Alternative to synthetic camera.

Modeler, interface file, renderer.

1. Modeler:
   1. Concerned with design and placement of objects.

   2. Highly interactive process.

   3. Possibly more than one modeler, tailored to each application.

2. Interface file:
   1. Output from modeler, input to renderer.

   2. Basic description of scene.

3. Renderer:
   1. Adds lighting, viewer perspective, material properties (additional components of interface file).

   2. Computationally intensive.

   3. PIXAR's Renderman.

Graphics Architectures

Early Systems

Host, DA converter, CRT.

• Eliminating flicker.

• Re-displaying a scene: computationally expensive.

Display Processors

Host, display processor and display list, CRT.

1. Separate processor.

2. Host sends primitives along, stored in display list.

3. Display processor deals with refresh, re-generation.

Graphics Pipelines

Consider a pipeline for computing large numbers of .

Properties of pipelines:

1. Depth --- latency.

2. Throughput.

Steps in a graphics pipeline:

1. Transformation --- between series of coordinate systems. Matrix multiplications.

2. Clipping --- cut scene down to field of view. Example: Excel.

3. Projection --- Project 3-D objects into 2-D. Matrix multiplication.

4. Rasterization --- Convert 2-D primitives to pixels.