Upcoming Events
Unite 2010
11/10 - 11/12 @ Montréal, Canada

GDC China
12/5 - 12/7 @ Shanghai, China

Asia Game Show 2010
12/24 - 12/27  

GDC 2011
2/28 - 3/4 @ San Francisco, CA

More events...
Quick Stats
46 people currently visiting GDNet.
2406 articles in the reference section.

Help us fight cancer!
Join SETI Team GDNet!
Link to us Events 4 Gamers
Intel sponsors gamedev.net search:
Look Up: (916 Terms)
Browse the Dictionary
Audio (80) Business (59) Community (19)
Design (89) Design Patterns (7) File Formats (32)
Games (64) General (83) Graphics (241)
Hardware (54) Network (41) OS (26)
People (30) Programming (143)
Browse Results: [All], Graphics
A mode of the VGA video card hardware that displays the screen in 320 x 200 pixels.
2 1/2 D Graphics
Slang. Refers to games with 3D polygon sprites and backgrounds, yet which have gameplay restricted to a 2D style. Examples are Super Smash Brothers, Einhander and Kirby 64.
A buzzword used to describe raycasting engines, such as Doom or Duke Nuke Em. The maps were generally drawn as 2D bitmaps with height properties but when rendered gave the appearance of being 3D models.
The name of your third dimension for viewing things. The first two being width and height, the third being depth.
3D Accelerator Card
A type of graphics card which helps a computer to process 3D graphics very quickly.

* Computers cannot actually do 'real' 3D imagery (well, none that you can buy without three major lottery wins and a Chair of Lucasian Mathematics at Cambridge University). We haven't invented the Star Trek-style Holosuite(tm) yet, so computers have to create a fake 3D which gets shown on a monitor or TV. Unfortunately, the conversion from 3D to 2D that has to be done requires a lot of mathematics. A 3D Accelerator card has specially-designed stuff on it which does a lot of the mathematics for the computer - leaving the computer to get on with more interesting work. Most 3D Accelerator cards now also accelerate the basic 2D stuff.

6 Degrees of Freedom. This refers to the ability to move in the X, Y and Z axis and also rotate around the X, Y, and Z axises.
AABB is an acronym for Axis-Alligned Bounding Box. Effectively it is a cuboid which is not rotated. All of its edges are parallel to the axes it is alligned to. AABBs can be defined using just two points, a maxium point and a minimum point. The rest of the box can be worked out from these two points. AABBs can be used for collision detection and scene culling (among other things).
Absolute Coordinate System
A coordinate system where all objects must be shown from the position of your viewpoint. Coordinates are all static and the Viewpoint (camera) moves through the preset coordinates.
Accumulation Buffer
A surface where multiple scenes can be added to before the surface is displayed.
ACT File
.ACT files are the actor files for Genesis3D - also referred to as models.
Affine Texture Mapping
The simplest form of texture mapping. The texture coordinates of a polygon are linearly interpolated across the polygon surface. This technique does not account for perspective and therefore produces swimming texture effects.
Affine Transformation
A combination of a set of linear transformations, usually stored in a single matrix. Affine transformations include rotation, scaling, translation and shear.
Albedo Texture
An albedo is a physical measure of a material's reflectivity, generally across the visible spectrum of light. An albedo texture simulates a surface albedo as opposed to explicitly defining a colour for it.
See Also:Specular Highlighting, Texture, Texture Mapping
The effect produced when a complex image is put onto a display with a limited resolution. This effect is a product of downsampling. Examples include jagged looking lines in lower resolutions.
Alpha Blending
Assigning varying levels of translucency to graphical objects, allowing the creation of things such as glass, fog, and ghosts. This can be accomplished by using alpha channels, or other means.
See Also:Alpha Channel, Alpha Testing
Alpha buffer
A linked list of depth-sorted colors, typically representing each pixel in a z-buffer. The colors may either be added to the list in a pre-sorted order or the list may be able to sort them by itself. After all colors are added the list is used to create on resulting color by blending the containing colors from the back and forward using their alpha channel.
Alpha Channel
In 32-bit color, 24-bits are used for the color, and the extra 8-bits represent an alpha value, or alpha channel. This value is used to determine the pixel's translucency level.
Alpha Testing
A method for creating transparency by checking the alpha value of a given pixel.
See Also:Alpha Blending
The process of creating simulated motion or activity.
Anisotropic Filtering
The level past trilinear filtering, this uses samples from multiple Mip Maps to get the best approximation for a texture. Very heavy performance cost.
Antialiasing refers to the process of adding additional pixels around the border of an object in order to blend it into it's background more smoothly, and to reduce the appearance of jagged edges. Typically, the colour used is the average of the surrounding background pixels and that of the object being antialiased, or an approximation of the average. This technique was invented by MIT's Media Lab.
This is the process that gets rid of the aliasing effect. A filtering process is normally used in the process that removes "jaggedness effect" produced by pixels.
A section of a circle that is measured in degrees or radians.
In graphics, persistent portions of images resulting from improper blitting operations.
Aspect Ratio
The aspect ratio of an image is the ratio of its width to its height. Aspect_Ratio = Width / Height See Pixel Ratio
A generic term for graphics, sounds, maps, levels, models, and any other resources. Generally assets are compiled into large files. The file formats may be designed for fast loading by matching in-memory formats, or tight compressions for handheld games, or designed to otherwise help in-game use. It is often useful to have an asset tool chain. The original models may be high-density models with R8G8B8A8 images. You may have a model striper and image compresser that reduces the model for LOD, and compresses the texture to a DXT compressed image. These assets may then go through further transformations, and end up in the large resource file.
Axis Aligned Bounding Box
A form of a bounding box where the box is aligned to the axis therefore only two points in space are needed to define it. AABB's are much faster to use, and take up less memory, but are very limited in the sense that they can only be aligned to the axis.
A secondary surface where the current frame's graphics are stored before they are transferred to the primary display surface.
Backface Culling
Because the polygons facing away from the viewer are not seen, and the extra time spent drawing them would have no effect on the visual quality of the scene, these backfaces are almost always removed in some manner.
Backface Removal
See Backface Culling.
Bezier Curve
A curve that is generated from the creation of several control points.
Bezier Patch
See Bezier Surface.
Bezier Surface
A curved surface created from a mesh of control points.
Bilinear Filtering
Uses the four adjacent corners to interpolate the value of a pixel in a texture map. This decreases the jaggedness of a texture when it gets larger, but also takes more memory and gives it a blurred look.
Bilinear Interpolation / Bilerp
When viewing a texture up close the texture becomes very aliased and ugly (for an example, in Doom get as close to a wall as possible). Bilinear Interpolation is the process of smoothing out the texture so that it is blurred and looks more smoothed when viewed up close.
A 2D image rotated in 3D so that its normal follows the normal of the viewing direction. Another definition would be: A 2D image that is rendered on the same plane as the viewing camera in a 3D world.
Usually a rectangular block of information where data is stored by the picture to make an image. For instance, if you capture the screen the data you would have is a bitmap of the screen.
An open-sourced 3D modeller released under the GPL. Blender was originally a commercial product, but was open-sourced when the parent company, Not a Number, folded. The Blender Foundation was formed by one of NaN's founders, with the intention of raising enough money to purchase the rights to make Blender open source, and met their goal within a matter of weeks. Blender is a powerful modeller, featuring built-in ray-trace renderer, NURBS curves and surfaces, Beziers curves and surfaces, powerful mesh modelling tools, meta-balls, skeletal animation and inverse kinematics, non-linear animation, sub-division surfaces, particle systems, etc... A Python interface allows for the creation of custom scripts for import/export, special effects, and so on. http://www.blender.org
When an image is drawn so that images drawn before it can still be seen. This is done through blending the source colors with the destination colors at different percentages.

Also called Transparency.

Bit Block Transfer (properly abbreviate BLT). The process of transferring a bitmap to or from a display surface via the blitter.
A highly specialized processor which is designed solely for working on graphics. * Usually adept at copying rectangular chunks of graphic data around from place to place. Since Operating Systems such as Windows, MacOS often deal with large rectangular chunks of stuff on a screen - such as, er, windows - a blitter has become a standard feature in today's graphics cards.
A method of outputting sprites by only showing non-transparent colors, without any checking during run time. This is done by precompiling a bmp into a piece of code that contains a set of screen memory writes that only include non-trasparent colors.
.BMP, a bitmap file format. Used as the standard file format for Windows, BMP files can display data through a number of different color depths, and is usually uncompressed. BMP format does include a Run-Length Encoding version for 8-bit files and does not support a 16-bit format.
Boolean Geometry
Named after mathematician George Boole Boolean geometry refers to combining multiple objects. Common operations include "unions" which combine two shapes and "difference" operations. Difference operations can be used to cut one shape out of another. 3d Studio Max and the game Red Faction for the Playstation II are good examples of how Boolean geometry can be used in practical applications.
Bounding Box
A box approximation of an object used for collision detection. An axis-aligned bounding box (AABB) is defined by the minimum and maximum 'x', 'y', and 'z' values. An oriented bounding box (OBB), however, generally gives a closer fit, because it is always aligned with the object.
Bresenham's Algorithm
An algorithm to compute which cells in a grid should be drawn in order to display a line between any two cells in the grid. Used to draw arbitrary lines to the screen.
BSP Tree
Binary Space Partition Tree. This is a sorting method for sorting nonmoving polygons where polygons are either in front or behind the currently polygon. The resulting linked list gives you the proper sorting for all polygons on the screen.
Bump Mapping
A process of rendering polygons that gives them an illusion of depth.
Carmack's Reverse
Refers to a modification to Heidmann's original stenciled shadow volumes technique generally attributed to John Carmack, although others came up with the same modification at about the same time. Rather than incrementing and decrementing for the front and back faces (respectively) when the depth test passes, the method increments for back faces and decrements for front faces when the depth test fails. This prevents shadow volumes from being clipped by the near plan, but introduces the problem of them being clipped by the far plane.

You can find out more about the algorithm here.
Cartesian Coordinate System
The standard coordinate system. With three dimensions, there are three scalars, x, y, and z used to represent a point at a given distance from a reference point, the origin.
Cel Shading
A technique which causes rendered objects to look as though they are hand-drawn, cartoon images.
nVidia's (relatively) simple shader language. It's currently one of the most commonly used shader languages, and it supports OpenGL and Direct3D even on non-nVidia cards. Toolkit and information can be found at www.developer.nvidia.com
Commonly used to describe a color component that makes up an image. A 24-bit image can have red, green, and blue channels. 32-bit images have room for a fourth channel, commonly known as an alpha channel.
Clear Reduction
An optimization of Z-buffering that buffers 1/Z values rather than Z values. Traditionally, Z-buffers are cleared each frame, but clear reduction makes it so they must only be cleared far less frequently.
Clipping Pane
The Clipping Pane is a barrier to determine the loading and display points of 3d models and textures. The farther away the clipping pane from the main object, the further you can "see", but comes at a heavy performance cost.
Clipping Plane
The Clipping Plane is a barrier to determine the loading and display points of 3d models and textures. The farther away the clipping pane from the main object, the further you can "see", but comes at a heavy performance cost.
Color Look Up Table. An index of colors, for instance to hold 256 different colors in a single byte and look up a color using 3 bytes.
Color bleeding
The idea of colors blending or "bleeding" into their surroundings, creating a smooth and realistic effect.
Color Cycling
Normally used with Indexed Color, this is a method of changing the index information for an image so that the colors change in a way that makes it appear to animate or cycle colors.
Color Depth
Refers to the amount of memory used to represent a single pixel, and is most commonly measured in bits. Common values are 8-bit, 16-bit, 24-bit, and 32-bit. More bits means a wider range of colors.
Color Key
A value indicating the color to be used for transparent or translucent effects. For example, when using a hardware blitter, all the pixels of a rectangular area are blitted, except the value that was set as the color key; this creates nonrectangular sprites on a surface.
Complete Binary Tree
A binary tree in which all leaf nodes are at level n or n-1, and all leaves at level n are towards the left.
Coordinate system
(see also: Matrix) An orthogonal grid in which elements are placed. For displays, each point in the coord system is a pixel. Note: On a computer monitor, x-coords increase from left to right, but y-coords indrease from top to bottom (contrary to traditional math).
The artifacts that appear around a bright light source. Often in circular or star like shapes.

Cross Product (Vector Product)

The cross product is a little more tricky. The cross product between two vectors yields a third vector which is perpendicular to the first two.

For vectors A, B, and a resultant C :

A X B = C

Or in component form :

( A.x , A.y , A.z ) X ( B.x , B.y , B.z ) = ( A.y * B.z - A.z * B.y , A.z * B.x - A.x * B.z , A.x * B.y - A.y * B.x)


This is commonly utilized to compute something known as a surface normal. As by the cross product formula, a normal is simply a vector that is perpendicular to some plane. In graphics, this plane is usually a polygon of some type.


Please note that A X B <> B X A . Ordering is very important in cross products. What actually happens is that B X A will yield a vector which points in the opposite direction of A X B.


Cube Mapping
An alternative to sphere mapping used in environment mapping, cube mapping gets a 'screenshot' looking in 6 different directions and arranges them in a rolled out cube. When applied, the object appears to reflect the environment around it. A 'cheap' alternative to raytracing reflections, cube mapping is fast enough for realtime.
Cut Vertex
A vertex whose deletion along with incident edges breaks up the remaining graph into two or more disconnected pieces.
Depth Cueing
A scene effect to make objects that are farther away from the camera are rendered with a lighter color so that they are given the illusion of distance
Introduced in DirectX 8; DirectGraphics combines both Direct3D and DirectDraw. (DirectDraw was removed after DirectX 7)
Dirty Rectangles
A method of updating only the changed parts of the screen. The screen is divided up into rectangles and only rectangles that have changes are makred "dirty" and then are redrawn to clean them up. Increases drawing speed as less is drawn.
The process of creating an illusion of more colors than are really available in the current color depth by creatively arranging individual pixel patterns.

Dot Product (Scalar Product)

The dot product shouldn't cause you any trouble. It is simply a way to multiply vectors. Keep in mind that we would normally describe these vectors in unit vector notation.

Here is a typical representation of a dot product for vectors A and B:

A · B

Or we can describe the same equation in terms of vector components:

A · B = A.x * B.x + A.y * B.y + A.z * B.z

Using a Dot Product, we can obtain the angle between two vectors A and B as follows:

Cos q = (A · B) / (|A| · |B|)

q = arcCos ((A · B) / (|A| · |B|))

Okay, lets break down what just occurred above. What we are saying is that the cosine of theta is equal to the dot product of A and B divided by the product of A and B's magnitudes. The magnitude of a vector V is denoted by : |V|


One can calculate the magnitude of a vector simply by taking the square root of each term

in the vector squared. Ie.

|A| = sqrt (A.x * A.x + A.y * A.y + A.z * A.z)


Double Buffering
A video buffer consists of a memory allocation for the information that is drawn to the screen. The first buffer is what is actually drawn to the screen, a second and third buffer are used to create a workspace to draw to that doesn't require synchronization to the vertical retrace of the monitor.

Double buffering gives the program a buffer to draw on that is not dependent on the retrace. The second buffer can be held in video memory and then "flipped" to change places with the primary buffer, which then is used as the second buffer until the next flip.

Drop Out
A drop out occures when a few polygons vanish because there are too many polygons being displayed on screen for the computer to handle.
Environment Mapping
An effect where an object reflects its surroundings, much like chrome.
Usually a polygon that is part of an object.
Flat Shading
That which assigns only one color shade per face.
The process of putting a hazy, fog-like area near the end of the visibility distance so that objects appear to disapear more naturally.
Frames per Second.
Frames Per Second. A measure of animation display rate. Not to be confused with First Person Shooter.
Fractional Dimensions. The name comes from the process of using fractional dimension mathematics to create images or data.

The data often works on a system of patterns being repeated in larger and smaller forms with the same algorithm.

A processed scene that consists of a static view of the game moment, it includes all objects, sprites and textures. Along with a set of frames, its possible to create an animation.
Frame Skipping
When a computer cannot process the frame graphics, or the game engine fast enough, a solution is to compensate the frame rate with "skipping". Each time the game loses performance, it skips some calculations and think ahead for compensating the time, without the heavy processing.
Frustum Clipping
When a polygon is clipped against the viewing frustum.
Flexible Vertex Format.
'Genre-lizing' (or non-US 'Genre-lising') is the act of defining a game to be within a specific genre. This has mainly pertained to the RPG genre, but is not limited to it. Genre-lizing, from a development point of view, is limiting to a designer in that they are forced to include attributes of a game that are associated with that genre and to not include those attributes that are not associated with that genre.
Ghost Images
Another term for Phosphor Persistence.
The bloody pieces of meat that fly off of a character when they get shot, chopped, etc., etc. Usage: "After I used my rocket launcher on that demon there were gibs _everywhere_!" Probably derived from 'giblets.'
Graphics Interchange Format. A format for saving graphics that usually uses a 8-bit (256 color) indexed bitmap.
Gimble Lock
A problem encountered when one tries to rotate using the 3 axises where one ends up rotating in the wrong direction after several rotations.

Common fixes to this problem are to use matrix multiplication for rotations or quaternions.

The OpenGL Auxillary Library. A library including functions to load textures and do other common tasks. It is no longer supported or updated, but it is still used for many beginning OpenGL tutorials.
The OpenGL Utility Toolkit. This set of libraries provides a set of helper functions to OpenGL, including methods to abstract the windowing system (for cross-platform development), rendering "standard" 3D objects, etc. For more info, visit the GLUT homepage
gMAX is a powerful subset of the game industry-standard 3D content creation tool – 3d studio max – that will be provided to game players for no charge on the web. gMAX is narrowly targeted at the interactive computer & video games market, and will be based on the next major release of 3d studio max. It is designed to serve as a highly extensible and customizable content creation platform for both professional and consumer use.
Gouraud Shading
Also known as intensity interpolation. A shading model where the light intensity is calculated at each of the vertices, and is interpolated across the polygon.
A form of data representation. Without graphics, we'd just have numeric displays. * Graphics use imagery, such as pictures, lines, points, colors and so on to represent information. It's the difference between a spreadsheet and a pie-chart. The spreadsheet is a numeric representation; the pie-chart is a graphical representation. Since computers see all things as numbers, programmers and electronics engineers have had to design ways to convert those numbers into moving images. The electronics engineers gave us the Graphics Card for this very purpose.
An algorithm by which a color is converted into a shade of gray(where R, G, and B are all equal), usually by weighting each of the R, G, and B components by certain percentages.
Graphical User Interface.
Real world lighting contains a high range of luminance values. HDR, or High Dynamic Range lighting, is essentially a technique that exceeds the normal computer graphics color range of 0 to 255, allowing for more realistic lighting models.
Height Map
A method often used to create 3D landscapes, height maps contain a grid of points that are given height values and the landscape is rendered by building polygons out of them.
High-Level Shading Language. A C-like language developed by Microsoft for DirectX Graphics. Shader programs, such as vertex and pixel shaders, can be written and compiled in HLSL rather than the various hardware shader assembly languages. The resulting compiled can then be loaded onto modern graphics hardware that supports programmable shaders.
Hidden Surface Removal The CPU removes hidden surfaces before sending meshes to the videocard. Because the CPU can remove entire meshes at a time, this can save a lot of time when checking individual triangles.
Heads-Up-Display. This means status information which are always visible like health or ammo information.
Indexed Color
When color information is stored in a look up table that contains the colors red, green and blue information.
Indot refers to non-quad and non-tri polygons. For instance a pentegon without triangulation or quads. (This is only appreant during modeling in a 3d application that supports this feature.)
Because the electron guns that draw pictures on TV screens were initially too slow to draw the screen in one pass, the first lines at the top of the screen would be fading by the time the last lines were drawn, pictures on TVs were drawn using what is known as interlacing. First the odd lines down are drawn (line 1, 3, 5, etc.). Then the even lines are drawn (line 2, 4, 6, etc.). The image shown by one pass is known as a field, and the complete image drawn by two passes is known as a frame. Standard TV broadcasts run at 30 frames per second (fps). In an effort to boost hype for a product, sometimes ads or press releases would state that their game was "60 fields per second", instead of 30 fps. Most computer monitors are non-interlaced as well as many arcade screens.
Using a ratio to step gradually a variable from one value to another.
An approximation between two known values. In graphics, this is a process in which the software increases the resolution of an image by first filling the image with blank pixels and then coloring the blank ones based on the color values of its surrounding pixels.
Inverse Kinematics (IK)
The process of creating realistic positioning of a complex object, such as an arm, based on the positioning of a lower-level node in the skeletal heirarchy, such as the twisting of a hand. This process works in the reverse of forward kinematics.
Isometric Tiles
One way of drawing 2D tiles so that it appears that they look 3D as they are layered with depth. In computer games this is usually a misnomer as isometric means a 1-1-1 perspective. Normally in games the depth ratio is less so that the tiles do not seem as wide.
This is an ISO standard by the Joint Photographic Experts Group. JPEGs work with color resolutions of up to millions of colors (also called 24 bit color). When you save an image as a JPEG, it gets compressed, making the file smaller. However, JPEG compression is lossy, so each time you save a JPEG, you lose some data and reduce image quality. When saving a JPEG, you can choose a compression level from low to high. Low compression gives you better quality but a larger file size. High compression gives you a smaller file - but less quality.
Lambert Shading
A shading model where the light intensity is calculated at one point along the polygon and is used to shade the entire face. Also commonly refered to as flat shading.
Left Handed Coordinates
A version of the Cartesian coordinate system where positive X points right of the origin, positive Y points up from the origin, and positive Z points beyond the origin.
Lens Flare
A refraction that appears in a camera lens normally when it is directed at the sun. A favorite added effect by many computer artists.
Light Map
A special type of image map that, when applied, effects the intensity of the texture of an object.
Usually refers to a line segment -- i.e., a one dimensional span of space between two points.
Level of Detail. Often pertaining to the complexity of 3D models, such as having less or more polygons when the model is closer or farther from the camera.
A coordinate that represents a point in a light map.
Scripted (non-interactive) 3D animation comprised of pre-set camera motion, character motion and environment changes. Mostly used for cutscenes within a game in place of pre-rendered CGI animation or film. Typically uses a game's real-time 3D engine to depict a portion of the game story in a non-interactive cinematic representation. Gamasutra Article
A premium modelling program produced by Alias, it is used by several companies in the games industry as their primary modelling tool. A free version is available at alias.com
A 3d modelling term that refers to a model as a series of polygons.
Milkshape 3D
A popular 3D modelling package. It was originally made for modding games, and can export and import a wide variety of popular game formats. It supports textures and texture coordinate mapping very well, and has a nice SDK. The registered version is currently $25. http://www.swissquake.ch/chumbalum-soft/
Mip Mapping
This process takes a texture and breaks it down into smaller pieces such as 1/2 the size, then 1/2 the size of that, and so on, so that different textures can be used at different distances from the texture. This helps the texture retain what its overall image should look like at different distances.
Mode X
A tweaked VGA card graphics resolution that displayed graphics at 320x240 which created a square pixel resolution. The term Mode X was coined by Michael Abrash.
Usually referring to an object that is created out of a number of polygons, splines or NURBS which to create an object in 3D.
A 3d engine developed in the late 1990s based on DirectX. Now known as 3d State. It has generated a large following of programmers and is praised for it's ease of use and compatability with many programming languages and compilers such as Microsoft Visual C++, Borland C++, Delphi, and Visual Basics. More information can be found at thier website at www.3dState.com.
Mortal Kombat
A series of popular fighting games created by Midway. Mortal Kombat was one of the first console games to show blood. Thus, it has always been a target, for the "Violence in video games" issue.
Motion Capture
Used for creating complex movements for 3D objects, such as a persons walk, run, jump or other actions. This is currently used more on console games due to the expensive costs in owning or renting equipment.
A format for encoding video. MPEG-1 was the original format which is used extensively for computer graphics, MPEG-2 is the basis for DVD.
A "vector" that points away from the face of an object at a right (90 degree) angle.
Normal Mapping
Normal Mapping is a lighting technique used to light a low-polygon model like a high-polygon model by means of a bitmap image. Each RGB value of each individual texel within the normal map is used to represent the x,y,z components of the mesh normal at each texel. Rather than using interpolated vertex normals to calculate the lighting data, the normals from from the normal map are used.
Non-Uniform Rational B-Splines
Little electronic, pipe, or mechanical looking objects that add function/life to a space ship texture.
A collection of polygons, usually arranged to create a representation of a real-life shape.
Recursively using 3 (perpendicular) cutting planes to subdivide space. Used for image/model storage and visibility determination (i.e. terrain rendering.) (Article)

See also: Spatial-partioning, Binary Space-Partitioning Tree (BSP), and Quadtree

SDK used orginaly in SGI workstations, currently the 3D render of choice by most developers.
Painters Algorithm
The idea of drawing from back to front, like a painter would while composing a scene. Problems can arise as some objects may be partially behind and in front of other objects.
A table of colors indexed by a buffer as a color source.
A collection of indexed colors.
Palette Swap
A method of swapping colors in the color table (palette) to change the appearance of images using that color table.
Parallax Occlusion Mapping
Employs per-pixel ray-tracing for dynamic lighting of surfaces in real-time on the GPU. The method uses a high precision algorithm for approximating view-dependent surface extrusion for a given height field to simulate motion parallax and perspective-correct depth. Additionally, the method allows generation of soft shadows in real-time for surface occlusions. Alternatively, POM can be coupled with well-known bump mapping algorithms such as normal mapping if physical accuracy can be sacrificed for greater computational efficiency.
Parallax Scrolling
The effect one sees one looking at a scrolling series of "layers". Each layer represents a distance from the viewer (the foreground, the background, far in the distance, etc.) As the entire view "scrolls" or moves in a 2-dimensional direction (no depth movement), the layers scroll at a different speed. The farther away from the viewer a layer is supposed to be, the slower it will scroll. This creates the illusion of depth.
Particle System
A particle system is a collection of entities, related or unrelated, that comply with a set of logical and physical rules. The components of a basic particle system are: an emitter, particles and particle modifiers. A particle is an entity that holds the necessary info about a particular particle in the system. The particle has different properties, or attributes, such as: velocity, position, size, affecting force, color, etc. The emitter is the object (usually not visible on screen) responsible for emitting the particles into the scene and giving them initial properties. Once these properties has been set, they can later be modified by the modifiers. The particles belonging to one particle system are usually associated with one texture. The texture is often applied to a rectangle that, in each game loop, is adjusted so that it always faces the camera, a technique known as billboarding. One visual effect can be made up by many particle systems, with different textures and characteristics. A burning fire, for example, could consist of one particle system for the flames, one for the smoke, and another one for emitting sparks.
.PCX is a graphic format created by ZSoft. Was sometimes used in DOS and early Windows games as it is a fairly simple format and supports Run-Length Encoding.
Perlin Noise
A function, inveted by Ken Perlin, who invented it to generate textures for the movie Tron (1982). One of the first films to use computer graphics extensively, Tron has a distinctive visual style. He won a special Academy Award for Perlin noise in 1997. Perlin noise is widely used in computer graphics for effects like fire, smoke and clouds. It is also frequently used to generate textures when memory is extremely limited, such as in demos.
Phong Shading
A shading model where the intensity of the light is calculated at each pixel in the polygon. This is extremely processor intensive, however, so it is virtually never used in unaccelerated, speed critical applications. "Fast phong shading" was an attempted optimization of the algorithm, though it is ineffective, and can not really be called phong shading.
Phosphor Persistence
Also called "Ghost Images". The after image of an moving object that is displayed on a black or mostly black background. A result of the time that it takes for the phosphor layer of a CRT to go from an excited state to a ground state. Not to be confused with artifacts.
Photon Tracing
The most realistic, but also the slowest, 3D rendering technique where you spawn photons from all light sources and use realistic photon physics (including reflection, refraction, diffraction, dispersion, etc...) to find out which ones reach the camera. This technique may also be used to pre-render photon maps to add a more realistic touch to games.
Image which approaches photographic quality. With a large enough color palette (around 16,000 colors) it is possible to display these types of images on a TV or computer screen.
Commonly referred to a standard way to see if a network connection is available - a packet is sent to a destination, which returns the packet or additional information. The gamers definition usually refers to the time it takes for a message to get to and back from a server or another gamer. Usually going through the game's network processing (i.e. the ping is handled by the game and not computer's network hardware or operating system) The original acronym is from Packet INternet Groper.
The width of a drawable surface multiplied by the number of bytes per pixel. For instance an 800x600 screen at 16-bit color would have a pitch of 1600.
Picture Element. A pixel is the smallest object in the computer graphics world. Screen sizes are measured in pixels such as 800x600 or 640x480.
Pixel Ratio
The pixel ratio is the ratio of the pixel's width to its height. If the pixels are square, the pixel ratio is 1 and the number of pixels in X divided by the number of pixels in y equals the aspect ratio. Pixel_Ratio = Pixel_Width / Pixel_Height X_Resolution / Y_Resolution = Aspect_Ratio / Pixel_Ratio See Aspect Ratio
Portable Network Graphics. A graphics format that was brought about due to the licensing disagreements conerning GIF and JPEG.
A dimensionless representation of a discrete location in space.
Point Filtering
The simplest form of texture filtering. This method will just take the closest texel to the pixel, which will increase the jaggedness of a texture as it gets larger.
Polar Coordinate System
A coordinate system in which a point is described by the angle and distance from a point O, the pole.
A closed form made up of at least three sides. Each side is made up from a line segment and each end of a line segment is a vertex.
Any time that a 3D object isn't smooth and you can see the polygons that make the object up. The object then looks too polygonal, for example, the 3D object of a man. On his face, his nose comes to a sharp point, his ears look like decagons.
Portal Rendering
A type of rendering where the world is split up into self-enclosing sectors. Each sector is made up of multiple walls, and each wall may be a portal to another sector.

The advantage of portal rendering is the simplicity to code it and the increased speed in your game. For each portal you make a clipping frustum out of the wall polygon, then render the sector that the portal connects to. The initial frustum is the viewing frustum used by the camera. That type of clipping is 3d, the 2d counterpart is where you calculate the 2d rectangle of the portal on the screen and do 2d clipping.

Post-Perspective Homogeneous Space
Coordinate system after the perspective-projection transform.
Potential Visibility Set
Hierarchical schemes partition the environment to allow efficient reduction of the superset of polygons down to a more managable subset of candidate polygons. (PVS)
To create graphics outside of a program. The opposite would be to algorithmically, procedurally or dynamically create the graphics at run-time.
This type of 3D scene is rendered and then stored, usually as a bitmap, and are often used as backgrounds and 2D sprites in games, like Nintendo’s Donkey Kong country. Unlike scenes that are rendered in real-time, prerendered images cannot have their viewing angle changed during runtime.
Procedural Texture Mapping
Texture maps created by an algorithm often based on a fractal noise or turbulence and an algorithm for the material to be drawn.

Advantages include being able to have unlimited levels of detail as each texture can be made precisely to the size that is currently being viewed and can always be continuous by using fractal patterns.


Usually the conversion of three dimensional coordinates to two dimensions, thus making drawing the polygon on a two dimensional screen far simpler.
ProMAT Animation Library
A FLI/FLC Software Development Kit (SDK), for Allegro
Potential Visibility Set. A sorting process used in 3D engines to determine that polys can be seen from a given position.
Common term used in modeling when refering to a model made up of four point polygonal faces.
Recursively using 2 cutting planes to subdivide space.

Used for image/model storage and visibility determination (i.e. terrain rendering.)

See also: Spatial-partioning, Binary Space-Partitioning Trees (BSP), and Octrees

An extention to normal complex numbers, invented by Sir William Rowan Hamilton. Usually used in 3D graphics to represent rotations and the orientations of coordinate axes. It is possible to use them to create smooth interpolations between two rotations.
A method of indirect lighting which calculates the light diffusion/deflection off of not only the light to face angle to viewer (specular) but calculates the diffused light coming from the light source, hitting the faces in the scene and diffusing light from those faces onto other faces then to the viewer.
The process of creating an image from 3D components.
ray casting
Raycasting is a technique used to produced 3-D worlds based on a 2-D matrix. In this technique, rays are shot out from a specific point and their intersections with the world (as represented by the matrix) are recorded. The information recorded from these intersections is then used to create a 3-D perspective.
Graphics technique used in Wolf3D where a number of rays (or lines) where drawn from the current position out to the obstacles they would encounter. The distance of the line and the surface they hit would determine what would be drawn to the screen.

If there was a 320x240 sized screen, there would be 320 rays cast out to get a sliver for each pixel of the screen.

Refresh Rate
The rate at which the screen is drawn, usually measure per seconds as frames per second (FPS) in software and Hz in hardware.
Relative Coordinate system
Also called the View-Centered Coordinate system. The Viewpoint (camera) is always at coordinates <0,0,0> and everything else in the Universe is based relatively to this home position.
Rendering Context
aka: RC The active area of memory that OpenGL writes information derived from the Model View, Projection View and Texture matrices, that is displayed on screen. Similar to the Device Context (DC) used by Windows when drawing graphics with GDI.
Developed by Criterion Software Limited in 1993, RenderWare (RW) is a 3D "middleware" applications programming interface (API) graphics rendering engine that has become more popular since version 3 and the Playstation 2 video game console where it has been used in various games (most notably Grand Theft Auto 3), many of which have been ported to the PC.
The accuracy of something. Often used as the resolution of a screen, the number of pixels in X and Y dimensions.
Red, Green, Blue. The 3 primary colors that make up all other colors by being displayed at different intensities next to each other. In a monitor this is referred to as a triad, which equals one pixel.
A 3D CAD NURBS modeller.
Right Handed Coordinates
A version of the Cartesian coordinate system where positive X points right of the origin, positive Y points up from the origin, and positive Z points behind the origin.
Run Length Encoding. A type of compression that reduces file sizes by shortening sequential "runs" of color of the same amount.

For example, 20 pixels of red horizontally could be abbrieviated to a number equalling a row 20, followed by the color of red. Then when the image is decoded it reads that there are 20 pixels of the following color and draws them.

To change the angle an object is being drawn at, or the camera is looking at.
Rotation Matrix
A type of matrix that, when applied, rotates a point.
Real video is shot and then drawn over to give very accurate looking cartoon animation. This was done on movies such as The Lord of the Rings and was also done on the game Prince of Persia.
Texture compression developed by S3.
A value that one can represent with one component.
The process of altering the size of an object.
Scaling Matrix
Generally appiled to an entire object or group thereof, a scaling matrix lets you either enlarge or decrease the size of the desired object.
Scan Conversion
Method by which a graphical object (such as a line, circle, ellipse, etc) is converted from its algebraic equation into its approximate representation on a raster display.
Scene Graph
A scene graph is a tree where the nodes are objects in a scene arranged in some sort of hirearchy. These nodes may be actual physical objects, or simply 'abstract' objects. For example a transformation node would apply some form of transformation to any 3D objects that are below the transformation node in the scene graph. A scene graph can be used for many things, depending on the way you order the nodes in the graph. For example you could have an octree containing object to be rendered in a scene, this would be a limited form of scene graph. You could have a scene graph that contains an octree as well as an alternative way or organising the same data, e.g. by render state. So you could use your scene graph for culling unseen objects as well as ordering objects to be rendered by render state.
An image taken from a game to show what was on the screen.
An assembly-like program which replaces part of the rendering pipeline with custom code. Shaders that affect vertices (vertex shaders) replace the normal transformation and lighting stage of the pipeline, while shaders that affect pixels (pixel shaders), work at the rasterization stage, affecting how the final screen color is determined. Shaders are supported in DirectX 8 and later, and in OpenGL through extensions (and as part of the proposed OpenGL 2.0 standard).
Skeletal Animation
Animation that is based on a model having a skeleton instead of being drawn as a series of different models (same model saved in different positions, key-frame animation). Skeletons are set up with joints or bones to determine how the unit will animate. Hierarchical Skeletons and Skinning Skeletal Deformation Example
A texture that is used to wrap around an entire model. Normally skins are drawn on a single bitmap, and then the coordinates are mapped onto the vertices of the model.
Solid Shading
Rendering a polygon a certain color without regards to the possible effects of lighting.
Recursively using n cutting plane(s) to subdivide space, where n typically ranges from:

1 = BSP 2 = Quadtree 3 = Octree

See also: Binary Space-Partitioning Tree (BSP), Quadtree, and Octree.

Reference: Foley & van Dam, Computer Graphics 2nd ed., pg 548

Specular Highlighting
A graphics technique which creates the illusion of light reflected on a surface. A specular highlight is the brightest point on an object.
"Spacial line" is a linear pattern that has a tri-coordinate system (x,y,z) rather than the planar system which forms a line(x,y).
A small bitmap image, often used in animated games but also sometimes used as a synonym for icon.
Refers to sub-surface scattering
Stencil Buffer
A buffer that holds information about what pixels should be drawn or not. Often used for creating shadows and reflections.
Sub-Surface Scattering
A rendering technique used on objects to determine how much light is allowed through the object. This is a very demanding technique since it calculates the amount of the material is it going through and the denisty of the material. This is most commonly used on skin, plant material, and cloth materials.
Swept Sphere
A swept sphere is a 3D object that can be created by pulling, or sweeping, a sphere along a path, leaving a trace that kind of resembles the shape of toothpaste when it comes out of the tube. While sweeping, the radius of the sphere may be changed, and the path does not need to be a straight line. Quite often, though, swept spheres are used in collision detection as an alternative to bounding boxes. In these cases, most of the time the path is a straight line and the radius stays fixed. The object that is created by sweeping a sphere like that is a cylinder with hemispheres, which have the same radius as the cylinder, attached to both ends. Detecting if a point is within this object is computationally quite simple, often easier than doing this with a bounding box.
Tranformantion and Lighting Recent 3D accelerators now have special features for hardware based transformations, which were traditionally control by the software, as well as hardware.
To divide an object into geometric primtives, such as triangles, for the purpose of simplification. Used either to make rendering or to reduce the complexity of the object.
Short for texture element, a texel is an individual pixel that is part of a texture.
A 2D image which is is used as a kind of wallpaper for the basic polygons used in 3D graphics. Textures are usually images of real-world objects and are often repeated over an entire object in a tiled fashion. Eg: a retouched photograph of a brick wall may used to provide the basic texture for a house's walls.
Texture Filtering
Using a filtering method, such as point sampling, bilinear, trilinear, or anisotrophic filtering, to resolve problems caused by applying a 2D texture to a 3D object.
Texture Generator
A small program that generate textures, mainly for use in 64k demos. The generators often have generate, filter and distortion functions.
Texture Mapping
The process of mapping a 2D image to a polygon. Often the polygon is rotated and a different size so that the texture must be rotated and scaled.
The process of applying a texture to something.
TGA is a graphical format for saving files.
A bitmap that can be placed to create a picture, usually used for background maps.
Transformation Matrix
Any kind of matrix that is used to alter the position and/or orientation of an object.
The act of linearly altering the location of a point.
Translation Matrix
A matrix that moves a point linearly.
Having the property of admitting and diffusing light so that the objects beyond cannot be clearly distinguished : partly transparent. (ie. Alpha-blending two polys together).
When an image is drawn so that images drawn before it can still be seen. This is done through blending the source colors with the destination colors at different percentages.

Also called Blending.

Having the property of transmitting light without appreciable scattering so that the bodies lying beyond are entirely visible. (ie. Color keying or transparent sprites - not drawing the mask color).
Converting a polygon into a number of triangles.
Trilinear Filtering
Combines bilinear filtering from 2 Mip levels (See Mip Mapping) to create a smoother version of the textures current size. Doubles the memory useage and can cause bandwidth problems.
Tripple Buffering
A video buffer consists of a memory allocation for the information that is drawn to the screen. The first buffer is what is actually drawn to the screen, a second and third buffer are used to create a workspace to draw to that doesn't require synchronization to the vertical retrace of the monitor.

Tripple buffering allows the advantages of a double buffer, where the program has a buffer to draw on that is not dependent on the retrace, but adds the additional advantage of being able to draw to the third buffer while the second buffer is waiting to flip to the front of the screen.

True color
When color information is stored in the form of its red, green and blue components it is said to be stored in true color. Opposite is Indexed Color.
Tru-Walk Technology
An animation technique used to create the illusion that the characters are engaging in various forms of locomotion.
User Interface
An interface in which the user interacts with the game. There is output given to the user through the monitor and speakers, as well as input taken from the user through the keyboard, mouse, joystick and other devices. User Interface is commonly used to describe the layout of the screen and screen elements that the user must interact with. For instance, buttons or windows that the user must use to access different features or give commands in the game.
In 3D graphics, a set of numbers representing magnitude and direction.
A line or movement defined by its end points, or by the current position and one other point.
Vector Graphics
Made famous by games like Asteroids and Battlezone, these are graphics that are defined by objects like lines, triangles or boxes, instead of by a bitmap.
The point of intersection of lines or the point opposite the base of a polygon or other object.
Video Electronics Standards Association. An organization that sets standards for video and multimedia in PCs. They created the Super VGA (SVGA) standard and the VESA Local Bus.
Visual Surface Determination
Ensuring polygons are drawn in appropriate order. Popular methods are depth soring, Z-buffering, portals, and BSP trees. Often abbreviated VSD.
Volume Rendering
A method of rendering that deals with the space that is used, as opposed to the faces that make up the outside of an object. Originally used for medical purposes only, it has also been adapted to do environmental effects, such as light through fog.
Volume pixel. These are basically another way of drawing 3D objects, instead of defining them as the polygons that make up the outside of an object, you define it as a number of points. Voxels can be the size of a pixel on the screen, or they can be large spheres, sprites or any graphical representation of a point.
Wings 3D
An increasingly popular free modeling utility. It has a very nice interface, and doesn't take much time to learn. It's good for both high and low polygon models, but is not particularly well suited for entire scenes or rooms. http://www.wings3d.com/
Antialiasing algorithm that extends the Bresenham-line by drawing additional (background-)blended pixels above and below each original pixel. What is considered "above" and "below" is decided by the direction of the plotting. Some implementations also blend the original pixels.
XGA is a Monitor Graphics Type which contains 15-inch units with a native resolution of 1024x768 pixels.
A pixel format comprised of luminance (the Y component) and chrominance (the color components U & V). YUV was developed for the compression of motion video data. See also RGB.
An array of numbers that store the proper point on an polygon and its distance from the camera. The array is usually comprised of all the pixels in a screen and only the closest points are stored and then drawn. This method eliminates the problem of overlapping objects.
The flickering distortion effect caused by two or more textures which overlap each other / are too close to each other / occupy simliar planes. It is called z-fighting because the multiple textures appear to be 'fighting' for dominance of the z-axis.

Add Definition

The Game Dictionary™ is a trademark of GameDev.net LLC. No duplication, reproduction, or transmission of the Game Dictionary or its content is allowed without the consent of GameDev.net LLC.