Surrounded by Sound

Dec 1, 1999 12:00 PM, By Vance Galloway and Scott Wilkinson

When the world of audio reproduction went from monaural to stereo, a new sonic vista was revealed. Individual sound sources (instruments, voices, and so on) could be localized within the left-to-right stereo sound field, presenting a much more interesting and involving sonic image to the listener.

In an attempt to improve the front-to-back imaging and general sense of envelopment, quadraphonic recording and playback was introduced in the 1970s. In this system, four channels of audio were directed to four speakers, which were generally placed in the corners of a room. However, quad never achieved widespread popularity for various technical and marketing reasons.

Despite quad's failure, the merits of multichannel audio (often called surround sound) are quite clear. This is demonstrated every day in commercial cinemas around the world. Modern movie soundtracks are routinely mixed to several channels, and speakers are placed around the theater. The sources of the apparent various sounds in the movie are appropriate to what's happening on the screen, heightening the audience's sense of being "in the action."

Now surround-sound playback systems are proliferating in "home theaters," and multichannel music-only recordings are becoming more common as well. So it's important for musicians, engineers, and producers to be familiar with the concepts and technology of surround sound.

WHAT IT IS
The most common type of surround-sound system uses six discrete channels of audio: five full-range channels and one channel that is limited to the low frequencies (typically below 120 Hz or so). As a result, the system is called 5.1.

The speakers that correspond to the five main channels are placed around the room (see Fig. 1). Two are placed just like a stereo pair, in front and to the left and right of the listener. The third speaker, called the center-channel speaker, is directly in front of the listener. The two remaining speakers are behind or to the sides of the listener; these are called the surround speakers.

The three front speakers are generally standard, direct-radiating designs. The surround speakers can also be direct radiators, but they are often of a different type called bipolar. In bipolar design, two sets of drivers are mounted into a cabinet, one set facing toward the front of the room and one set facing toward the rear. This means that such speakers should be placed to the sides of the listener. Most commonly, these drivers are angled into the room slightly. Bipolar speakers produce a greater sense of envelopment than direct radiators, and they make it more difficult to pinpoint the location of individual sound sources, which is desirable in most movie soundtracks.

The remaining channel is often called the low-frequency effects (LFE) channel, and it requires a specialized type of speaker called a subwoofer. Because low frequencies are relatively nondirectional, the subwoofer can be placed almost anywhere in the room; typically, it is in the front between the center and either the left or right speaker. Sometimes it is placed in a corner in order to reinforce the low frequencies.

Of course, each speaker needs a channel of amplification. Most modern receivers include five or six amplifier channels. (Many subwoofers have their own internal amp, so they don't need external amplification.) More sophisticated playback systems include separate power amplifiers. A lot of home-theater amps provide five or six channels; alternatively, you could use three stereo amps or five or six monaural amps (often called monoblocks). Of course, you can also use powered speakers all around.

The next step back along the signal path is the surround processor. This device takes a digital signal from the source (for example, a CD or DVD player), decodes it into discrete, analog channels, and sends them to the amplifiers. (Receivers have built-in surround processors in addition to amplifiers.)

This brings us to the source, which is typically a CD or DVD. (Other digital sources, such as laserdiscs and digital satellite feeds, can also carry 5.1 surround-sound signals.) How are 5.1 channels encoded on these media? And what surround-sound standards have been established to encourage the widespread acceptance of multichannel sound?

SECRET DECODER RING
Over the past few years, two standards have been vying for dominance in the surround-sound market. By far the most successful is Dolby Digital, which has been adopted as the standard surround-sound format for DVD-Video discs and the new digital-television (DTV) system.

Uncompressed audio at 44.1 kHz with 16-bit resolution requires a bandwidth of about 700 kilobits per second (kbps) for each channel, so six channels need a total of 4.2 Mbps. In order to conserve bandwidth, Dolby Digital uses a lossy data-compression and encoding algorithm called AC-3 to reduce the required bit rate to roughly 400 kbps. Because it uses this algorithm, Dolby Digital is sometimes mistakenly called AC-3.

The other contender is called DTS, from Digital Theater Systems. Derived from the company's commercial-cinema surround-sound system of the same name, this format has been slow to reach the market. DTS also uses lossy data compression to bring its bandwidth requirements down to about 1.4 Mbps, which is less severe compression than in Dolby Digital, but still substantial. Whether this decreased compression improves the sound quality is hotly debated within the home-theater industry.

Both formats require a player that can output the appropriate data stream from the disc and a processor that can decode it. Virtually all DVD players can output Dolby Digital, and most can now output DTS as well. Similarly, all modern receivers and surround processors can decode Dolby Digital, and most can handle DTS.

In addition to Dolby Digital and DTS, there are a couple of other formats you should be familiar with. A new DVD format called DVD-Audio will soon be available (see "Tech Page: CD? No, DVD!" in the July 1998 issue of EM). Unlike DVD-Video, the new format is designed primarily to store audio-only data on standard DVD discs (although it can include a limited amount of graphics, such as album covers, lyrics, and artist bio information).

Because a DVD disc has such a large storage capacity (4.7 GB and up), uncompressed, multichannel audio is easily accommodated. As a result, DVD-Audio allows multiple channels of audio at several bit resolutions and sampling rates from 16/44.1 to 24/192 (although 24/192 is limited to two channels).

In addition, different rates and resolutions can be used on different channels within a single mix. Even so, it is likely that lossless compression will be used to reduce bandwidth requirements by a factor of two. In fact, a scheme called Meridian Lossless Packing (MLP) has recently been adopted as the standard compression for DVD-Audio. DVD-Audio players should become available by the end of the year.

Another new type of audio disc is the Super Audio CD (SACD) from Sony and Philips (see "Tech Page: CD-The Next Generation" in the March 1998 issue of EM). This type of CD stores multichannel audio using a very high sampling rate (2.8 MHz) and 1-bit resolution, as well as a proprietary encoding algorithm called Direct Stream Digital (DSD). In addition, an SACD can include two layers: one with multichannel DSD data and one with standard stereo CD data, which means that such hybrid discs are fully compatible with current CD players. However, Sony's first commercial SACD player is stereo only.