How Surround Sound Works
By definition, surround sound is sound that surrounds youin particular, music or movie soundtracks that come at you from all sides, thanks to multiple audio channels fed through speakers positioned in all corners of the room. In a home theater system, the surround sound is initially encoded onto a programming sourcea DVD, a cable or satellite broadcast, and so on. When the programming is played, the surround sound tracks are extracted from the source by a piece of electronics called a surround decoder. The individual surround channels are then amplified and fed to the appropriate speakers. The result is sound that envelops you as you watch your movie or listen to your CD.
How do all those channels of audio get from their source to your ears? It's all a matter of technology. Two primary technologies are used for surround sound today: the older matrix surround, typified by Dolby Pro Logic, and the newer discrete surround, as used in the Dolby Digital system. We'll look at both, starting with the oldest first.
The original home surround technology tried to pull off the impossibleturning two things into four. That's because, at the time (we're talking the early 1990s herea virtual lifetime ago), the videotapes and laserdiscs we were watching barely had space to carry two stereophonic channels, let alone the four channels necessary for a complete surround experience. Engineers could fit right and left channels onto the tape or the disc, but no more than that. So where to put the extra channels?
To stuff four channels of information into the space normally used by two, the engineers at Dolby Laboratories used a matrixing technology. This enabled them to combine four streams of information into two tracks, and then retrieve the original four channels on playbacksometimes referred to as a 4-2-4 processing system. (They took four channels, crammed them into two channels, and then separated them into two channels again: 4-2-4.)
Does that sound a little tricky? It was. Imagine mixing streams of red, green, blue, and purple sand together in a bucket, carrying the bucket across a room, and then trying to extract the individual colors at the other end. Messy at best, and perhaps even impossible to separate into the original colors.
The Dolby engineers figured out how to do it, however. By analyzing the audio information contained within the left and right channels, they could identify the information that was the same in both channels. They realized that they could mix information from a third channel into both the left and right channels, compare the two channels, and then extract that information (the third channel) that was identical in the two channels. They could even get a fourth channel into the mix, by recording it out of phase with the third channel and then also feeding it to the left and right channels.
Using this approach, the Dolby engineers devised the following data streams:
Stream A (left channel)
Stream B (right channel)
Stream C (the data that's identical in streams A and B)
Stream D (the difference between the data in streams A and B)
Obviously, streams A and B carried the normal left and right channel information. The engineers used stream C to carry center channel information, and stream D to carry information for the surround channel. (This technology only allowed for a single surround channel, even though this single channel was often fed to two rear speakers.)
Here's how the whole thing works. Left channel information is fed into data stream A, and right channel information into stream B. Center channel information is placed into both the A and B streamsthus resulting in the "identical" stream C. The surround channel information is also recorded on streams A and B, but placed out of phase to the second stream. This creates stream D, the "difference" stream.
Thus, the videotape, videodisc, or television programanything that can carry a stereo signalis encoded with these four data streams. If you listen to the program through a normal stereo audio system, it sounds like normal stereothe center channel information comes out (in equal portions) through both the left and right channels, as does the surround channel information. No one knows the difference; the positioning of the new streams is effectively "hidden."
But when you run the program through a surround decoder, everything changes. The decoder passes the data in streams A and B directly to the left and right amplifiers. It identifies the identical data between these two streams and extracts it into the decoded stream C, which is sent to the center channel amplifier. The decoder then identifies the differential data in streams A and B (the out-of-phase information), shifts them relative to each other so they're back in phase, and sends the stream to the surround channel amplifier.
The only bad thing about this processcalled Dolby Pro Logicis that it's not terribly precise. That's because when everything gets all mixed together, it doesn't always separate out completely, resulting in some degree of leakage from one channel to another. But the results are surprisingly effectiveparticularly when the surround channel contains primarily ambient effects. However, if you want to clearly position a sound source in the surround channel, the effect is less than impressive; no matter how effective the decoder, there will always be some duplication of the sound in the front channels, which spoils the effect. How, then, to better separate the surround channel sound?