It's no secret that 3-D acceleration has captured the hearts and minds of computer enthusiasts everywhere. Graphics cards are classified by the generation that they appeared in this multi-dimensional craze and are judged based on their ability to render complex, high-polygon scenes. As popular as gaming has become, there is another overlooked facet of today's power hungry adapters that is slowly and quietly being embraced by an alternative audience. Welcome to the wonderful world of software DVD decoding.
Like most emerging technologies, DVD has made its way onto the PC market slowly. Originally requiring hardware assistance for smooth playback, software decoding has come a long way in terms of performance and visual quality to enable full-frame playback on any mid-to-high-end system. Almost anyone with a DVD-ROM drive, a software DVD player and a means to get to the local video store can enjoy hours of home entertainment for a low price.
Decoding the MPEG-2 video stream isn't an easy task, and without a hardware decoder, the burden falls upon the CPU. Because of this, video card manufacturers have built in additional support for decoding to effectively decrease the load on your processor. Terms such as "motion compensation" and "IDCT" refer to such performance enhancing features.
The amount of acceleration varies by card and not all contestants entering the game are created equal. Companies not normally seen at the top of 3D shootouts are usually dominant in video acceleration, so as we take a look at the features that today's top accelerators offer, expect to see some changes to the order in which the cards fall.
It seems as though every time a new piece of hardware is released, a new set of acronyms is introduced in conjunction, and software DVD decoding is no exception. Each card's video acceleration features can be split up into either a quality enhancing or performance-enhancing group. First, let's take a look at the features that will decrease the load put on your CPU.
Motion Compensation is the term used to describe the final stage of the MPEG decoding process. The technical definition, as stated in Microsoft's Direct Draw DDK is:
If two successive frames of video data display essentially the same picture, only shifted horizontally or vertically, then there is an enormous compression gain by simply shifting the first frame and then describing the other changes that happen in addition to that shift to arrive at the second frame. This is what motion compensation does.
