PC 3D Graphics Glossary of Terms
A
B C D E F G H I J K L
M N O P Q R
S T U V W X Y Z
- A
Accelerated Graphics Port (A.G.P.)
- A high-bandwidth 32-bit PC bus architecture to be introduced in 1997.
It provides up to 528 MBytes/sec to a graphics controller, yielding the
bandwidth necessary for texturing directly
from system memory. A.G.P. is physically different from the (132 MByte/s) PCI
bus, but does not replace PCI. A.G.P. uses a combination of frame buffer memory
local to the graphics controller, as well as system memory, for graphics data storage.
- Alpha
- Additional color component in some representations of pixels,
along with red, green, and blue (RGB). The alpha channel denotes transparency or
opacity, often as a fractional value, used in blending and antialiasing.
- Anti-aliasing
- An algorithm designed to reduce the stair-stepping artifacts (sometimes called jaggies) that result from drawing graphic primitives on a raster grid. The solution usually relies on the multi-bit rasters ability to display a number of pixel intensities. If the intensities of the neighboring pixels lie between the background and line intensities, the line becomes slightly blurred, and the jagged appearance is thereby diffused.
- B
Bitmap- A flat 2D picture, or a texture map. Most 3D systems are capable of combining 2D bitmaps and
3D objects onscreen.
- Back buffer
- A hidden drawing buffer used in double-buffering.
Graphics are drawn into the back buffer so that the rendering process
cannot be seen by the user. When the drawing is complete, the front and
back buffers are swapped. See double-buffering.
BLT (Bit-aLigned BLock Transfer)
- The process of copying pixels or other data from one place in memory
to another.
- C
- Camera
- In 3D graphics, the viewpoint through which a scene
is viewed. Flythroughs of scenes are conceptually a moving camera.
- Clipping
- Removing, from the processing pipeline points and surfaces which
are outside the field of view (known as the "viewing frustrum").
- Culling
- Removing, from the processing pipeline to spare unneeded work,
complete objects and surfaces which are completely hidden by other objects,
or are facing away from the viewer (i.e. Backface culling)
- D
Depth Cueing- Reducing an objects color and intensity as a
function of its distance from the observer. For instance, a bright, shiny red
ball may look duller and darker the farther away it is from the observer.
- Dithering
- The process of intentionally mixing colors of adjacent pixels.
Dithering is usually needed for 8-bit color, and sometimes for
16-bit. It allows a limited color set to approximate a broader range, by mixing
groups of varying-color pixels in a semi-random pattern. Without dithering, color
gradients like sky or sunset tend to show "banding" artifacts.
- Display List
- Data structure, in very compact form, describing objects
and scenes to display. Usually consists of floating-point coordinates (X,Y,Z) and
colors (R,G,B) for points on objects.
- Double Buffering
- The process of using two frame buffers for smooth animation.
Graphical contents of one buffer are displayed while updates occur on the
other buffer. When the updates are complete, the buffers are switched. Only complete
images are displayed, and the process of drawing is not shown. The result is the
appearance of smooth animation. Double buffering may be implemented by hardware
which simply changes the pointer that the DAC (Digital-Analog-Converter) circuitry
is using to access pixels, or by complete copies of the entire buffer via software.
- F
Fogging- Also known as Haze. The alteration of the visibility or
clarity of an object, depending on how far the object is from the camera. Usually
implemented by adding a fixed color (fog color) to each pixel.
- Frame Buffer
- Display memory that temporarily stores (buffers) a full
frame of picture data at one time. Frame buffers are composed of arrays of bit
values that correspond to the displays pixels. The number of bits per pixel
in the frame buffer determines the complexity of images that can be displayed.
- G
Gouraud shading- Also known as smooth shading. Applies a lighting
calculation to each vertex of a polygon face, and linearly interpolates the results
across the face to achieve a smooth lighting effect with gradual color transitions.
- Graphics library
- A tool set for application programmers,
interfaced with an application programmers interface, or API. The graphics
library usually includes a defined set of primitives and function calls that enable
the programmer to bypass many low-level programming tasks.
- J
Jaggies- Also known as Aliasing. A term for the jagged visual appearance of lines
and shapes in raster pictures that results from producing graphics on a grid
format. This effect can be reduced by increasing the sample rate in scan
conversion.
- L
Lighting- A mathematical formula for approximating the physical effect of light from various sources striking objects. Typical lighting models use light sources, an objects position & orientation and surface type.
- LFB
- See Linear Frame Buffer.
- Linear Frame Buffer
- A buffer organized in a linear
fashion, so that a single address increment can be used to step from one pixel
to the pixel below it in the next scan line in the frame buffer. The entire
LFB can be addressed using a single 32-bit pointer.
- M
Mapping- The transformation of one coordinate system into another.
In the 3-D viewing pipeline, for instance, an object is defined by the
application developer in model coordinates; these are mapped to world
coordinates; the world coordinates are mapped to normalized device c
oordinates (NDCs); the NDCs are mapped into device coordinates; and the
final picture is displayed. Also, an application for computer graphics
systems.
- Matrix Transform
- Multiplying a 4x4 matrix times a 4x1
(X,Y,Z,1) vertex. This single operation is adequate to accomplish
any rotation, translation, and projection of the vertex.
- Mesh model
- A graphical model with a mesh surface constructed
from polygons. The polygons in a mesh are described by the graphics
system as solid faces, rather than as hollow polygons, as is the
case with wireframe models. Separate portions of mesh that make up
the model are called polygon mesh and quadrilateral mesh.
- Mip Mapping
- A texture-mapping scheme, which uses multiple
resolutions of each texture, pre-filtered, at typically 100%, 50%, 25%,
and 12% resolution. The rasterizer
either chooses the resolution (the map-level) that most closely matches
the instantaneous resolution of the image being textured, or it interpolates
between the two nearest mipmap levels. The resolutions are, for example,
128x128, 64x64, 32x32, and 16x16. Each is one-half the dimension of its
larger neighbor. They are pre-filtered for best appearance at their
particular magnification.
- P
Perspective Divide- The computationally expensive
division of the (X,Y) coordinates of a point by a scaling factor
proportional to the Z coordinate, to give the appearance that
objects near the viewer are larger, and that parallel lines
(like railroad tracks) converge in the far distance.
- Phosphor
- A luminescent substance on the inside of the cathode-ray tube display that is illuminated by the electron gun in the pattern of graphical images as the display is scanned.
- Phosphor triad
- One red, one green and one blue phosphor that composes a pixel.
- Pixel
-
Short for PIcture ELement. In a raster grid, the pixel is
the smallest unit that can be addressed and given a color or intensity.
The pixel is represented by some number of bits (usually 8, 16 or 24)
in the frame buffer,
and is illuminated by a collection of phosphor dots in the CRT that are
struck by the beams of the electron gun.
- Poly
- Abbreviation for Polygon
- Polygons
- Three-sided (triangle) or 4-sided (Quadrilateral) or occasionally more-sided
"simple" geometric objects.
Complex objects like teapots are decomposed, or "tessellated",
into many (like 700) polygons to allow regular processing of the data,
and hardware acceleration of that processing.
- Primitives
- Smallest units in the 3D database. Usually points, lines, and
polygons. Some 3D hardware and software schemes also employ curves,
known as "splines":
- R
Radiosity, Ray-Tracing
- Complex methods of drawing 3D scenes, which result in photorealistic
images, but cannot be done in real time with current hardware. Essentially,
they calculate the
path that light rays follow from objects to the viewer, and all the
accompanying reflections.
- Raster
- A rectangular grid of picture elements, or pixels, on
a CRT (Cathode Ray Tube) screen.
The graphical data to be displayed on the raster is stored by the
frame buffer. Raster operations (ROPs) can be performed on some portion
or all of the raster. Such operations aid in the efficient handling
of blocks of pixel data.
- Rasterization
- The process of creating every visible pixel's
value, from the transformed display list.
- Rendering
- The process of translating high-level database
descriptions to pixels on the screen.
- RGB
- A color model in which colors are specified as intensities
(between 0.0 and 1.0, if floating point representation is used, or between
00 and 0xFF for integer hex) of the
three CRT monitor primary colors: red, green, and blue. For RGB24 (8 bits
each of red, green, and blue), a value of 0xFFFFFF hex is white.
Other common RGB formats are 16-bit per pixel, in 555 or 565 bit
partitions.
- S
Shading(Flat, Gouraud, Phong)- The process of creating pixel
colors. Gouraud is a constant increment of color from one pixel to the next,
while Phong is much more complex and higher quality. Flat shading means
no smooth blending of colors. In flat shading, each polygon is a single color.
- Setup
- Stage of a 3D rendering pipeline, where polygons are put into
a data format interpretable by a rasterizer (SW or HW). Setup typically
involves calculation of slopes of polygon edges, and incremental values
for pixel interpolation across scanlines.
- T
Tessellate- To divide an object or surface into geometric
primitives (triangles, quadrilaterals, or other polygons) for simplified
processing and rendering.
- Texel
- A colored dot in a texture map; texture element.
- Texture
- A (2 dimensional) bitmap pasted onto objects or
polygons, to add realism
- Texture filtering
- Bilinear or trilinear filtering.
Also known as sub-texel positioning. If a pixel is in between texels,
the program colors the pixel with an average of the texels colors
instead of assigning it the exact color of one single texel. If this is
not done, the texture gets very blocky up close as multiple pixels
get the exact same texel coloring, while the texture shimmers
at a distance because small position changes keep producing
large texel changes.
- Texture mapping
- The calculations required to
place each point of the texture bitmap to the correct place in the
3D world, accounting for the rotation/scaling/translation/perspective
of the world's objects. This mapping takes a large number of memory
references and computation, often accounting for more than half the
total CPU time in the 3D pipeline. May therefore be hardware accelerated
in the graphics chip.
- Three Dimensional Graphics
- The display of objects and
scenes with height, width, and depth information. The information is
calculated in a coordinate system that represents three dimensions
via x, y, and z axes.
- V
Vertex- A point in space, described by (X,Y,Z) or
(X,Y,Z,W), where "W" is a fudge factor that keeps track
of perspective effects in a string of computations.
- VRML
- Virtual Reality Modeling Language. A database
description language applied to create 3D worlds. A standard for
internet-based 3D modeling, but still evolving - current Web
VRML browsers implement VRML 1.0 or 2.0.
Does not have as much description flexibility (yet, anyway)
as Autodesk 3DS (3D Studio) format.
- Z
Z-buffer- A common method of eliminating hidden
surface removal (which the viewer should not see) in the rasterization
process. An integer value is stored in RAM, the "z-buffer"
for each point on the screen. Before each pixel is drawn, the
existing z-buffer value is compared to the z (depth) of the
object at that point; if the existing value is less, it is
nearer, so the new pixel is discarded.
However, if the new Z value is less, then it is written to
replace the old value, and the corresponding pixel is updated.