02:36 AM

MPEG-4 vs. Microsoft VC-1: why high-definition video software standards are irrelevant

By Damien Stolarz for SiliconValleyWatcher

Teapot-Tempest.jpgAbout a year ago, Microsoft made great strides in legitimizing its technology for broadcasting by getting its Windows Media 9 video codec (now grandly titled Video Codec One, or VC-1) accepted by the industry's standards body SMPTE.

And both the HD-DVD and Blu-Ray groups, representing the two competing High-Definition DVD hardware formats, have agreed to support Microsoft's codec as well as the MPEG-4 Advanced Video Coding standard in their new hi-def players.

This presents an interesting question. The Windows VC-1 and MPEG AVC camps are fighting over which is the "best hi-def codec". But with the future media players supporting both codecs, does it matter which one content producers choose? I think the software doesn't matter anymore.

Everyone's familiar with the names of some codecs because their acronyms are widely used, but here's a little more background: MP3, Apple's AAC, WMA, RA, and OGG are all audio codecs. They all compress analog audio down to less than a tenth of its original size when storing it in digital form, then convert data back to analog so that we can hear it.

To use them on an audio player, such as a computer, you need to pay a fee. The popularity of MP3 is its ubiquity; the popularity of AAC is due to its use in the iTunes music store. The popularity of WMA is that it is the default format for any music ripped on a Windows machine. The popularity of OGG is that it is royalty-free. But they all do their job, and a computer (and many portable music players) can easily play them all.

The modern video codecs are also very similar to each other, and use similar mathematical techniques. They all break the picture into a lot of squares, and then use fancy mathematics (like those used in JPEG pictures) to represent these small squares with the smallest amount of bits. (You can find a technical comparison of the differences here.)

But they are all roughly comparable in quality, and the the VC-1 vs. MPEG AVC battle is a technical stalemate; the differentiating features have much more to do with implementations, price, and of course politics.

Some will still argue that it's essential to have a single software standard, to make sure that a 2005 hi-def DVD video player can play media produced in 2015, just as a 1995-vintage DVD player can still play DVDs made today. But that's also less relevant than ever.

The real thing that is causing the software standards to be irrelevant is powerful hardware. Traditionally, "computerization" of an industry required application-specific computer chips. For instance, automobiles, PDAs, sewing machines, and DVD players all used completely different microchips.

But now, powerful general-purpose computers are getting so tiny that it doesn't make sense in many cases to make hardware that's too specific.

Sure, if you need to mass-produce a cheap, single-function portable video device, you design or purchase the right microchip that implements, say, some version of MPEG-4 or Windows Media. But a home theatre PC or hi-def DVD player doesn't have to be power-efficient or small (in fact, DVD player cases are mostly empty - they're just a small circuit board and an optical drive).

Initially, HD DVD players (or Blu-Ray DVD players) may be expensive. But they don't have to be. In fact, the Mac Mini has a street price of $499 and can play back some HD-resolution content. You could assemble a PC for less and play back hi-def Windows media. And a full-fledged home theatre PC can play back anything you throw at it.


As I commented in an earlier blog, it's entirely possible to design hi-def DVD players that can support new video codecs through software updates alone. People already patch the firmware on their DVD players to unlock new features; a sufficiently general-purpose processor in a DVD player (or a video decompressor chip that can be reprogrammed) could adapt to future Windows or MPEG AVC standards. It's not inconceivable. In fact, people are starting to expect the hardware to do post-processing to the video to tune it to their own HDTVs.

Even if my flights of fancy about upgradeable DVD players turn out false, people have gotten used to $29 DVD players. It's going to be hard for vendors to sell hi-def DVD players with stratospheric prices. And when players become cheap enough, people WILL go out and buy a new one if they need to.

So the fighting between groups supporting different hi-def codec standards may seem important right now. But in the long term, it's only a minor skirmish in the battle for the living room.