Audio Recording, Stereophonic and Surround Sound

It is perhaps difficult to understand the problems experienced by early listeners when they first heard ‘‘music coming out of a hole.’’ It was an effect that annoyed some people much more than others, although few had even encountered it before the twentieth century. Together with the difficulties of low amplification and high background noise, this effect proved significant and was a further problem in sound recording and reproduction.

In the late twentieth century, the words ‘‘stereophonic’’ and ‘‘stereo’’ refer to a sound along a horizontal line, usually between two loudspeakers along one wall of a room. Good stereo can simulate a live performance, but the expression is restricted to loudspeaker listening. Listening on headphones is quite different, because our hearing has evolved to take advantage of two different phenomena. For natural sounds (and for loudspeakers), our heads form a ‘‘sound shadow,’’ so high-pitched sounds from the left are picked up by the left ear very much louder than the right ear.

For low-pitched sounds, the shadow-effect is practically nonexistent. What little directionality we possess depends upon differences in the time the sound arrives at the two ears. If recordings were made with microphones the same distance apart as our ears (about 17 centimeters), and something the same general shape and size as a human head were put between the microphones, we might have much more faithful stereo reproduction than any loudspeakers could give.

This is called ‘‘dummy head’’ or ‘‘binaural’’ stereo. We learn directional hearing in the first year of life, correlating the senses of sight and stereo sound with feelings of balance and the sensations our muscles give us as they move. With headphone listening, the acoustic fidelity is certainly better; but when the listener moves his head, the earpieces move too and the sound remains the same, yet his other senses indicate the exact opposite.

The first publicized experiment into spatial sound reproduction was performed by Clement Ader in Paris in 1881. He placed a row of microphones along the footlights of the Paris Opera House and sent signals by telephone lines to earpieces at an electrical exhibition 3 kilometers away. A spatial effect was reproduced. American workers were first to achieve stereo sound recording.

Closed-circuit experiments established that just one pair of microphones and loudspeakers would reduce the disadvantage of ‘‘sound coming out of a hole;’’ but attempts to reproduce physical movement showed that if the microphones were more than a few feet apart, a new ‘‘hole in the middle’’ might appear. A third central microphone-and-loudspeaker arrangement reduced this ‘‘hole in the middle,’’ demonstrating that three were much better in practice.

In 1932 a significant experiment was performed using landline connections in Philadelphia, which seemed to show that for orchestral music, three channels would be sufficient. While significant, it was not a definitive experiment because there were apparently no comparisons with more channels and the vertical element was not featured. Nevertheless this established the benchmark against which future systems would be compared.

This experiment led to the first stereo recordings in 1932 of the Philadelphia Orchestra conducted by Leopold Stokowski, recorded on a two-track disk at Bell Laboratories in New Jersey. Because there was only one experimental stereo disk cutter, continuous performances could not be recorded. In 1979 Bell Labs reissued the surviving material on two charity stereo long-playing records (LPs) with the missing sections in monophonic sound, clearly showing the advantage of stereo.

On the other side of the Atlantic, work started from different basic principles. It was noticed that when mono sound was reproduced on two identical loudspeakers equidistant from the listener, the sound appeared to come from a point midway between the two speakers. By dividing the same mono sound into suitable proportions and sending it to two speakers, it could appear to come from anywhere between the two. This was the claim in Alan Blumlein’s British patent in 1931. Blumlein was an employee of the record company EMI, but he considered the principal application to be soundtracks for films, unlike in the U.S. where the initial work was for sound alone.

Blumlein’s work showed how suitable signals might be captured by two bidirectional microphones at the same point in space. It also showed how the correct balance for monophonic listeners would result from mixing the left and right in equal proportions, so downward compatibility would occur. It also discussed how stereo sounds might be recorded on disk or film, including some remarkable predictions of what occurred decades later. The important point was that Blumlein showed that only two discrete channels could be sufficient.

Blumlein’s work also mentioned a device to ‘‘steer’’ a mono sound between two loudspeakers, now called a ‘‘panpot.’’ In 1934 the EMI Amateur Dramatic Society made a short stereo film called Move The Orchestra, in which a single restaurant scene achieved exactly that.

The public heard stereophonic sound for the first time in 1939 in the Walt Disney film Fantasia. In the ‘‘Fantasound’’ system, three discrete channels were printed onto a separate optical film that held the three soundtracks side-by-side, together with a fourth track that controlled the volumes of the other three, giving wider dynamic range. During the last musical piece (Ave Maria), loudspeakers at the back of the auditorium could be switched on for the choral effect. This was not only the first time the public had heard stereo, but it was the first time they encountered ‘‘surround sound.’’ The outbreak of World War II brought all these advances to a halt.

After LPs became established in the 1950s, techniques were researched to get two channels of sound into one groove. For commercial master sound recording in the U.S., the three traditional discrete channels were initially recorded on i-inch magnetic tape. In Europe, EMI issued the first British ‘‘stereosonic’’ 1-inch magnetic tapes in 1955 using Blumlein’s principles.

They were intended for loudspeaker listening; Blumlein’s ‘‘coincident microphone’’ technique did not capture the time information needed for headphones. The standard form of stereo disk that carried two sounds in one v-shaped groove was launched in 1958. A single cutter vibrating in two directions mutually at right angles imparted two sounds onto the two groove walls, which gave compatibility between mono and stereo disk reproduction.

Although human eyes see an almost circular field of vision, most practical action takes place on a narrower plane. Widescreen filmmakers followed American thinking: a different microphone for each loudspeaker behind the screen. But dialog was found to give problems since it was upset both by performance considerations and the acoustic properties of studios. Subsequent stereo films used sound picked up on mono microphones and ‘‘panned’’ into place. Ultimately, mono became standard for dialog, with only music and effects in stereo.

During the first half of the 1970s, the high fidelity, or hi-fi, market stimulated the idea of ‘‘quadraphonic’’ sound reproduction. The idea was a commendable one, to reproduce the entire acoustic environment of a concert hall rather than just the stage. The concept failed for several reasons, but it can be argued that the sophisticated engineering research needed for quadraphonic LPs helped prolong the life of mechanical disk records until the end of the century.

Until digital methods came to the rescue, the shortage of analog channels continued to hamper surround sound. Dolby Laboratories had invented methods of improving the perceived signal-to-noise ratio of analog recording media, and they also found ways to encode more channels of audio with extra information for cinema audiences, while retaining downward compatibility with stereo and mono. The first ‘‘Dolby Surround’’ films date from 1976.

When video recording became suitable for domestic use, the same principles became available to domestic listeners. At the end of the twentieth century, there were several ways of encoding surround sound. The dominant one used 5.5 channels (the “half” being low-pitched sounds such as aircraft, guns, and bombs normally requiring a large loudspeaker, which could be placed behind the viewer).

It seems certain that further developments will occur in the twenty-first century. The vertical element of surround sound, for example, is not yet featured, although it is an increasingly important part of our environment.