Graphics Reference
In-Depth Information
How the Traditional View Is Implemented
In the OpenGL system, the actual processes that implement the pipeline are
grouped into different kinds of functionality. A block diagram of these func-
tional groups in a generic graphics system is shown in Figure 1.5.
The first functional group handles the vertex processing that is shown in
Figure 1.1. The input to this group includes vertices, normals, primitive defini-
tions, colors, lights (and their parameters), materials, and texture coordinates.
The output is a set of vertices as pixels with their color, depth, and texture
coordinates, and perhaps as revised primitives.
The next step is rasterizing. This implements the Vertices-to-Fragments
step in the rendering pipeline of Figure 1.3. The input to the rasterizer is the set
of vertices in screen coordinates with their depth, color, and texture coordinates,
along with how the vertices are to be connected. The rasterization process inter-
polates the vertices to create fragments, and the same interpolation is applied to
determine the depth, color, and texture coordinates for each fragment.
The second functional group is fragment processing. The input to this
group is a fragment rasterized from
a graphics primitive. The frag-
ment's color is determined by pro-
cessing its color, depth, and texture
coordinate information. The out-
put of fragment processing is a set
of completed pixels, the “RGBAZ
Pixels” of Figure 1.5, with color
(RGB), blending (A, for alpha), and
depth (Z) values, ready to be inte-
grated into the color buffer.
The final step is this integration
of pixels into the color buffer. This
corresponds to the Fragments-to-
Pixels section of the rendering pipe-
line of Figure 1.3. The pixels from
the fragment processor are inte-
grated into the color buffer by ras-
ter operations that merge the frag-
ment with the pixels in the frame-
buffer. This is the same for both
fixed-function and shader-based
graphics.
Figure 1.5.
The OpenGL pipeline in graphics hardware.
