| 16, 24 and 32-bit Color: |
16-bit mode can produce 65,536 different colors and 24-bit mode 16.7 million different colors. 32-bit mode has the same amount
of colors as 24-bit but this is only a technicality. Manipulating 32-bit graphics is usually a significantly faster process than 24-bit
graphics. You also need more memory to run in 32-bit mode.
| 2D Card: |
A normal video adapter without 3D acceleration.
| 3D Card: |
A video adapter powered by 3d-chip that does the hard work and leaves more power to the processor. Normally a 3D accelerator includes 2D
acceleration as well, but there are some cards that are only dedicated to 3D acceleration such as Voodoo2.
| Accelerated Graphics Port (AGP): |
AGP is a PC bus architecture developed to supplement PCI. AGP is more powerful and it can also use the system memory to store textures in.
In PCI-based 3d cards the texture memory is limited by the card's own memory.
| ACPI: |
Advanced Configuration and Power Interface. It enables the operating system to manage power and system configuration.
| AGP Aperture: |
The AGP Aperture setting specifies to the AGP chipset how much of a PC's system memory, including virtual memory, is addressable by
an AGP video card. The “standard” size is 64MB, but with some motherboards you are able to change it from 4MB to 256MB. A general rule is
to keep it ½ of your system memory. 256MB RAM -> AMP Aperture Size 128MB.
| AGP PRO: |
AGP Pro is developed for graphics cards in high performance workstations and will deliver four times the electrical power of the standard
AGP interface through an extended connector.
| Alpha Blending: |
| Anisotropic Texture Filtering: |
(Before reading this, look for bi-linear and tri-linear filtering) Anisotropic (which means non-uniform shape) samples from more than 8
different shapes and therefore can produce better quality, however anisotropic filtering requires even more fillrate than tri-linear filtering
does which means it is slower.
| Anti-aliasing: |
Anti-aliasing is a method of smoothing out jagged edges in order to improve the quality of a rendered image.
| API: |
API is Application Programming Interface. API is the interface between a 3D program and the 3D graphics board. It talks to the hardware
for the software -- in order to use 3D with any speed, you must be using a game that is written for the API that your video board supports.
e.g. Glide, Direct3D, OpenGL.
| Artifact: |
An artifact is more or less a graphical “glitch”. Something that is not intended to be on screen. Artifacts usually appear for example if
a graphics card is heavily over-clocked.
| Bi-linear Filtering: |
Bi-linear filtering uses an interpolation method to produce smooth transitions between different pixels in the source texture, this is done
by sampling the four closest pixels of the source texture (or most suitable mip-map) and interpolating these values before rendering each
single texel on screen.
| Bit Map: |
A 2D image used for example as textures.
| Bump Mapping (Emboss): |
Bump Mapping is a method of simulating surface graininess with textures. Emboss Bump Mapping uses multiple textures to simulate the bumps
in the geometry by shifting the bump texture in regards to a light source. It is supported by all 3D hardware, but sometimes with a significant
performance (fill rate) penalty.
| Bump Mapping (Environment map): |
Bump Mapping is a method of simulating surface graininess with textures. Environment Mapped Bump Mapping is a method that requires specialized
hardware support, and although it is gaining popularity fast, is not yet widely supported.
| Camera: |
In 3D-scene camera means the viewpoint where the scene is seen from. A moving camera is a "flythrough" of a scene.
| CAS Latency: CL2, CL3: |
CAS comes from the words Column Address Strobe, meaning the ratio between memory column (memory bank) access time and clock cycle time.
CL2 offers a slight performance increase over CL3.
| Clipping: |
Removing, from the processing pipeline points and surfaces which are outside the field of view (known as the "viewing frustum").
| Clipping Planes: |
Clipping planes could be visualized as the borders of the area that is drawn on screen. Normally front and back clipping planes are always
used, front being the minimum distance that 3D graphics drawing starts from, and back being the maximum distance that is drawn. See also
| CPU Optimization: |
It means that a specific piece of software is optimized for a given hardware platform, usually using specific instructions on that hardware
like MMX, 3DNow! or SSE. Optimizations are needed in order to maximize performance of latest hardware, and are generally found in CPU-intensive
applications such as games and DVD players.
| Cube environment mapping: |
Cube environment mapping is essentially a "snapshot" of an object from all six sides in 3D space, rendered into textures. The
end result is an accurate reflection which will change in real-time as the viewpoint changes. Cube environment mapping works best with static
environments and non-moving object since otherwise the "snapshot" would need to be re-rendered (a relatively slow operation) to
provide an accurate reflection.
| 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)
| Curved Surfaces: |
Curved Surfaces can mean either dynamically tessellating (adding polygons) round shapes in 3D geometry before rendering them. Adding a material
like gouraud shading makes the final appearance of the shape to be curved. Alternatively they are (incorrectly) used to describe higher-order
surfaces such as béziers and B-splines.
| DDR: |
Double Data Rate SDRAM memory. DDR SDRAM memory transfers two units of data per a single clock cycle. It is twice as fast as plain SDRAM
and more expensive.
| Deferred Rendering|
A technique, that ensures that no pixels are drawn if they are not necessary. This type of rendering
boosts the Fill Rate a lot. Many manufacturers have already announced their interest in this technique.
| Depth Cueing: |
A technique for changing objects color and shades depending on the distance of the target.
| Diffuse lighting: |
Also known as Gouraud lighting. See Gouraud Shading.
| DirectX: |
A big set of different APIs developed to be used in Windows. DirectX aims to make multimedia and games run smoother and they provide a very
high degree of comp ability.
| Direct3D (D3D): |
The real-time 3D graphics API in DirectX. Direct3D has become a standard supported by all 3D cards and also by most games.
| Dithering: |
Dithering is the process of intentionally mixing colors of adjacent pixels. It is usually needed for 8-bit color, but also 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.
If dithering is not used, color gradients like sky or sunset tend to show "banding" artifacts.
| DOT3: |
The abbreviation comes from a vector mathematics formula calculating a three-component dot product. It is generally used to talk about dot
product lighting / dot product bump mapping. DOT3 is the mathematical operation that combines a light vector with a surface normal, producing
a very realistic looking (bumped) surface.
| Double-buffering: |
A technique needed to maintain fluid graphics without artifacts when rendering both 2D and 3D. The video board renders into what's referred
to as a back buffer and when the image has been finalized it's flipped into the front buffer and the back buffer is cleared so that
the video board can start rendering the next frame.
| DXTn (DirectX Texture Compression): |
Also sometimes referred to as S3TC, DXTn where "n" depicts the texture compression format. There are six formats with different
number of bits for the colors R, G, B, and A (alpha). The compression algorithms are designed for real-time decompression, and are linear,
i.e. decoding a compressed block takes always the same amount of time on the hardware. The compression ratio is 1:6, and makes it possible
for developers to use larger textures in the games.
| Edge Antialiasing: |
In edge antialiasing, you render a 3D scene, then re-render the silhouettes of the objects to antialias with lines. The system redraws those
edges, blurring them to reduce artifacts. Edge AA only smoothens polygons edges, and does not blur the textures. Edge AA can be computationally
expensive and it can be difficult to determine which edges should be antialiased. For these reasons, full-scene antialiasing is often preferred.
| Fast Writes: |
Fast Writes is a technique to boost the transfer between the CPU and Graphics card in the AGP bus. When enabling Fast Writes, it performs
all writes from the CPU to the graphics card without going through the system RAM. Not all cards support this feature.
| Fill Rate: |
Fill Rate is usually used to measure how fast a graphics card is. Fill Rate is the rate how fast pixels can be drawn into the video memory.
Fill Rate is usually shown in Millions of pixels per second (Mpixels/s) or optionally Millions of texels per second (Mtexels/s). These two
units should not be mixed.
| Filtering: |
See tri/bi/anisotropic filtering.
| Flat Shading: |
A shading technique, where every side of the polygon is filled with one color without special tricks.
| Fogging: |
An effect that fades distant objects away smoothly.
| Frame Rate: |
Most commonly used as “FPS” (Frames Per Second). The rate of the frames (pictures) per second are drawn on screen. A human eye need around
24-30 to see a smooth movement. Still, 60 FPS looks better than 30 FPS.
| Front/Back/Frame Buffer: |
Frame Buffer means a screen-sized buffer where the 3D accelerator can render images to be displayed on screen. Front buffer is the one that
you see, and back buffers are used for rendering the image while front buffer is displayed. See also Double and Triple Buffering.
| FSAA (FullScene Antialiasing): |
Full-scene Antialiasing is a method of smoothing out jagged edges in order to improve the quality of a rendered image. FSAA usually does
not require any special support from the software, it can just simply be switched on.
| Floating Point Unit (FPU): |
FPU handles non-integer calculations in a CPU (it was previously called a math co-processor). Nowadays all CPUs have an FPU.
| Gamma: |
Gamma is used to control the brightness and color balance of your display output.
| Glide: |
Glide is 3dfx's proprietary API. Glide needs a 3dfx card to work.
| Gouraud Shading: |
A shading technique, where every separate color is calculated for all vertices and interpolated along the polygon. Also known as "smooth
| GPU (Graphics Processing Unit): |
NVIDIA's marketing term for a Hardware Transform & Lighting capable accelerator.
| HAL: |
Hardware Abstraction Layer. Interface between hardware and software.
| Horizontal Retrace: |
On CRT (Cathode Ray Tube) monitors, when the electron beam reaches the right hand side of the screen, it needs to be deflected back again.
The horizontal retrace is the time it takes it to return from the right side of the screen after it has traced a row of pixels. It is not
used for displaying picture data. The drawing of the picture needs to be synchronized with the incoming signal. Synchronization is achieved
by including synchronization pulses in the retrace periods. If the picture is not properly horizontally synchronized, it tears away from
the left of the screen. To ensure that they are recognized as synchronization pulses, they have different voltage levels from the picture
| HSDRAM: |
HSDRAM (High-Speed SDRAM) is a manufacturer-specific (Enhanced Memory Systems) marketing name for regular SDRAM that is claimed to run at
higher speeds than normal.
| HSR: |
Hidden Surface Removal. A term to describe the reducing of overdraw when rendering a scene by not rendering surfaces that are not visible.
| Hyper-Z: |
A marketing name by ATI for their Radeon Z-buffer optimization.
| Kernel: |
Kernel is the core of an operating system that manages the most basic operations, such as sharing the processor between tasks and handling
| Lambert Shading: |
See Flat Shading.
| Lightmap: |
A transparent texture that is used to create lighting effects on top of the normal surface detail texture.
| Mapping: |
All objects in 3D are made up of polygons. Mapping is the technique by which a bitmap is applied to the polygon to provide realistic surface
details. Usually referred to as "texture mapping" or "texturing".
| Miniport Driver: |
A device-specific kernel-mode driver that is linked to Windows 2000 / WDM (Windows Driver Model). You could for example need an AGP miniport
| Mip-mapping: |
Mip-mapping is a technique where several versions of the same texture are used to represent this texture on screen at different distances,
further away a smaller version of the texture is used and closer a bigger one. Mip-mapping can be used regardless of filtering method.
| Mobo: |
A short “nickname” for Motherboard.
| Motherboard: |
The motherboard, also sometimes referred to as the mainboard, is the communications center through which all the components that make up
your PC transmit information back and forth. Every component, from the CPU to the joystick used to play games, is plugged into the motherboard
in one way or another. Some components are plugged into slots in the motherboard itself while others are plugged into ports that are accessed
from the back of your PC case.
| Motion Blur: |
An effect often visible in film or TV when objects move faster than the camera's frame exposure time would allow. Motion Blur is always
simulated in computer 3D graphics, as it does not occur naturally.
| Multisampling: |
Multisampling is the way DirectX 8 does full-scene antialiasing. Multiple samples (called subsamples) of rendered pixels are taken, and
blended together before being output to the screen.
| OEM: |
Original Equipment Manufacturer.
| OpenGL: |
A high-end API originally used mainly for CAD. Nowadays almost all 3D accelerators are shipped with OpenGL-drivers. It is very popular among
programmers because of it's easy use. The biggest competitor to Direct3D.
| Overclocking: |
Overclocking is done by increasing the MHz of some parts (CPU, Graphics card etc.) to a higher setting than default. This voids your warranty,
and might damage your hardware. Only users with more experience are advised to overclock (and with your own responsibility).
| PCI (Peripheral Component Interconnect): |
A standardized connector inside your PC used for connecting add-on cards such as modems, network cards and 3D accelerators.
| Perspective Correction: |
"Perspective correction" also applies perspective projection inside polygons. To achieve full realism, both textures and gradients (colors) should be
perspective-corrected. The alternative is to perform projection on the object vertices only, which causes textures to bend and wobble.
| Phong Shading: |
Phong Shading is a lighting technique that produces good looking highlights and is not very dependent underlying polygons, like Gouraud
Shading. It is not currently used in real time accelerated 3D graphics because to produce correct results, it requires computationally intensive
per-pixel calculations. There is a "Fast Phong Shading" method that emulates the true method, but it too is not widely used.
| Pipeline: |
In 3D engines, a pipeline means the processing path for the 3D graphics. This includes steps like transform, lighting, culling and rasterization.
Some of these steps are done by the 3D accelerator, some by the CPU.
| Pixel: |
Abbreviation of “picture element”. The smallest element of an image represented as a point, with a specified color or intensity level.
| Pixels/sec vs. Texels/sec: |
Shortly explained, Pixels/second means how many pixels a card can draw to the display buffer per second. These pixels may blend together
many texels. Texels/second means how many texture elements the card can fetch from memory in a second. These terms are often confused, especially
in marketing materials
| Polygon: |
Polygon is a more general definition of a 3D surface, that can be composed of three or more vertices.
RAMDAC is an abbreviation for Random Access Memory Digital-To-Analog converter. It is a part of the video card which converts data in video
memory to analog signals that get sent to the monitor. Usually for economical reasons, the RAMDAC is integrated on the graphics chip, but
it can also be on a separate chip. The speed of the RAMDAC is referred as Megahertz, and simplified, the more MHz the RAMDAC has, the higher
resolutions, color depths and refresh rates are available.
| Refresh Rate: |
Refresh Rate indicates how many times per second the image on your monitor is drawn. It is not directly related to how many times per second
the content of the screen changes (that's called frame rate). Normal refresh rates are for example 60Hz or 75Hz.
| Rendering: |
This is called the overall process of creating a 3D image with the chosen features e.g. fog, alpha blending, textures, lighting etc.
| S3TC (S3 Texture Compression): |
S3's trademarked term for the texture compression in DirectX (from 6.0 onwards). See DirectX Texture Compression.
| Scan Line Interleaving (SLI): |
SLI makes it possible to connect 2 Voodoo2 cards together with (theoretically) double-speed.
| SGRAM: |
Synchronous Graphics RAM. Slightly modified SDRAM.
| SDR: |
Normal SDRAM (also may be used as abbreviation of Single Data Rate). It is slower and cheaper than DDR memory.
| Sideband Addressing (SBA): |
With sideband addressing, the AGP bus utilizes 8 extra "sideband" address lines which allow the graphics controller to issue new
addresses and requests simultaneously while data continues to move from previous requests on the main 32 data/address wires. This adds performance.
| Single Pass Multi-Texturing: |
Multi-Texturing is a process of applying two or more textures to a 3D object. This is usually done to increase texture detail or to create
specialized effects such as environment mapping or bump mapping. Single Pass means that the 3D accelerator supports drawing more than one
texture to the same 3D object at once.
| Sprite: |
Sprites are typically 2D-images textured on a rectangular 3D object. Examples of typical uses for sprites are billboards and particles.
| Stencil Buffering: |
At its most fundamental level, stencil buffering enables applications to mask off sections of the rendered image so that it is not displayed.
Applications often use stencil buffers for special effects. It is supported by most of the new 3D accelerators.
| Supersampling: |
Supersampling is a way of doing full-scene antialiasing. It reduces jaggies and pixelation on the screen by rendering the screen many times,
or rendering to a (four or more times) larger surface and downscaling the image on the fly.
| T&L: |
Transform and lighting (T&L) are two major steps in the 3D graphics pipeline. They are computationally very intensive. Transform phase
converts the source 3D data to a form that can be rendered and Lighting phase calculates lighting for the 3D environment. These two steps
can be performed simultaneously or consequently to a triangle. Traditionally the CPU performs them, but nowadays some 3D accelerators also
have dedicated hardware T&L solutions.
| TC&L: |
Transform, Clipping & Lighting – basically same as T&L since clipping is an essential part of the 3D graphics pipeline.
| T-Buffer: |
3dfx's marketing term for a technique called accumulation buffers, available in Voodoo4/5. T-buffer makes it possible to do some special
effects such as FSAA and motion blur.
| Tearing: |
Tearing is an effect that is caused when the VSync is disabled. See "VSync" for more details.
| Texel: |
A texture element, i.e. one pixel from the surface of a texture (does not correspond to pixels on monitor).
| Texture / Texture Mapping: |
Texture is a bitmap (for example a normal photo) mapped onto a polygon or several polygons, to provide a more realistic surface for 3D geometry.
| Texture Memory: |
The part of memory where 3d accelerators store the textures that are needed rendering a scene.
| Transparency: |
Makes transparent surfaces (e.g. water) possible. Very important effect in 3D graphics to make (e.g. explosions) look much better.
| Tri-linear Filtering: |
Works very much like Bi-linear filtering, but tri-linear filtering also uses the four closest pixels in the second most suitable mip-map
to produce smoother transitions between mip-map levels. All in all tri-linear filtering samples eight pixels and interpolates these before
rendering, twice as much as bi-linear does. Tri-linear filtering always uses mip-mapping.
| Triangle: |
Triangle is formed by three vertices.
| Triple-buffering: |
Triple buffering is very similar to double buffering but it adds a third buffer, by doing so video boards that support this feature can
maintain a more fluid output and start taking new commands from the application sooner. Triple buffering of course adds to the memory requirements.
| Tris: |
Abbreviation for "triangles". See Triangle.
| Vertex: |
Vertices are the “basic unit” of 3D graphics. All 3D geometry is composed of vertices. Vertices contain X, Y and Z positions plus possible
vertex normal and texture mapping information.
| Volume Textures (3D Textures): |
They are three-dimensional textures – just imagine standard textures piled on top of each other. They can possibly be used for special effects
like fog or volumetric lighting but take up a lot of memory.
| Voxel: |
Volumetric texel. A non-3D-accelerated technique used by some games to depict for example terrain or characters. Nearest real time 3D technique
would be "3D Texturing".
| VSync: |
Vertical Retrace (VSync) is a signal that occurs once per screen refresh, when the electron beam on the monitor starts drawing a new frame
from the top-left corner of the screen. Usually all programs wait for Vertical Retrace to occur and then flip front and back buffers. This
eliminates any "tearing" effects on the screen and results in a steady image being shown. If VSync is disabled, the programs do
not wait for the electron beam to finish updating the previous frame but can change in the middle of the refresh, as soon as the back buffer
image has been rendered completely.
| W-Buffer: |
W-buffering is a slightly improved version of depth buffering. The depth values are more evenly distributed in a W-buffer than in a Z-buffer.
This allows applications to support larger maximum depth ranges, while still getting relatively accurate depth buffering close to the viewer.
Not all hardware supports W-buffering.
| Z-buffering: |
Z-buffering calculates which pixels are visible in the finished image and if the current pixel is behind the older one, it is not drawn.
Essentially, this provides nearly perfect visibility-checking.