Graphic Rendering The Real World
The real world is filled with complex objects: curves, textures, shadows, and especially the commonplace flaws found in everyday structures that make them real. The challenge for game developers is to duplicate these objects in a realistic manner. A character's face shouldn't look perfect, because there are always small blemishes, wrinkles, and other imperfections. Also, the physics of movement and interaction between objects in the real world have to be faithfully duplicated.
So the challenge for video card manufacturers is to create hardware that is able to support all of the programmatic techniques developed to present a realistically rendered world within a computer. When comparing video cards, you'll often run into terms such as bump mapping, anisotropic filtering, bilinear filtering, MIP mapping, etc. All of these (and others) are techniques implemented in the hardware that let game developers create realistic objects in their games. Let's take a very quick tour of what some of these terms mean.
GPU – Graphics Processing Unit; as video cards become increasingly complex, it pays to offload a lot of the heavy-duty visual processing from the computer's main CPU to what amounts to a mini-CPU on the video card. The video card's GPU does a lot of the computational heavy lifting required to render 3D graphics.
DDR – The term DDR is often used to describe a video card's memory. DDR stands for "double-data rate" and is a type of SDRAM used in video cards today. Double-data rate means this type of memory is able to move data on both the falling and rising edges of the clock cycle. DDR1 SDRAM is the most common memory in use with video card technology today and is able to move data in 2-bit chunks.
Again, the purpose here is to offload graphics rendering tasks out of the computer's main memory and CPU, hence the idea to equip video cards with loads of memory. Nowadays, a cutting-edge video card should pack 256MB of DDR1 or DDR2 SDRAM. Unlike DDR1, DDR2 memory moves data in 4-bit chunks, so data transfer performance is doubled. The newest video cards on the market today are beginning to adopt DDR2 memory.
RAMDAC – Your computer only understands digital, and your eyes can only see in analog. The video card RAMDAC's job is to bridge the gap between the digital and analog worlds and let the computer display analog images on the monitor. Faster RAMDACs (RAMDAC speeds are measured in megahertz) enable the computer to display higher resolutions on a monitor without annoying flicker.
3D Terminology – Today's video cards render highly realistic 3D images. To do this, video cards must faithfully reproduce what gamers see in the real world, including lifelike textures and realistic movements. There are a number of techniques implemented in both hardware and software that let video cards generate realistic 3D images. Let's take a look at a few of the most important:
Z-buffering – A technique implemented in the video card's memory that is used to store information about pixel location as fluid scenes are rendered. For example, if a pixel moves behind another, that information is stored in the Z-buffer.
Texel – Stands for "texture element." A texel is a graphics unit similar to a pixel, except it is used to define 3D graphics. Think of a texel as the base unit that defines 3D graphics.
Fogging – Technique used to simulate haze, fog, or smoke.
Trilinear filtering – Video cards use texture maps to apply textures, such as skin, surfaces, etc., to simplified models of objects called primitives. Essentially, these primitives are "colored" with textures stored in 2D images called texture maps. Textures must not only look realistic, they must also be adjusted, if need be, according to the position and movement of the object. Trilinear filtering is used to make this mapping appear more realistic.
Full-screen anti-aliasing – Anti-aliasing is used to smooth out transitions between pixels with different colors and reduce image "jaggies." This technique minimizes the jaggies at all resolutions.
Programmable vertex and pixel shading – These techniques allow the rendering of realistic lighting and shadow effects. For example, ATI's SMARTSHADER technology allows the implementation of DirectX 9 floating-point programmable vertex and pixel shading. NVIDIA's version is called CineFX.
DirectX 10 – This is Microsoft's latest version of the DirectX APIs that let video developers programmatically exploit the capabilities of existing video hardware. The API lets software developers quickly create games without having to know the intimate details of each
on Dec 27, 2009 | Computers & Internet