What is stereo? A little history

Ask yourselves what "stereo" means to you. Most of you will probably say that it is two-channel audio reproduction. Stereo is short for "stereophonic sound" and, by definition, should be understood as the use of more than one independent audio channel to achieve a sensation as close as possible to natural listening. The opposite of stereo (or stereophonic sound) is monophonic sound, good old mono.

In everyday usage we have associated stereo solely and exclusively with audio reproduction through two loudspeakers. In fact, stereophonic comes from the Greek stereos (solid) and phone (sound), and it was a word introduced by Western Electric by analogy with the word stereoscopic. This first demonstration took place at an SMPTE convention in 1937.

For sound engineers, stereo means the technique that allows us to place recorded sources in a relative position within what we know as the sound space. That being the case, it is not wrong to call 5.1 sound (or 6.1, 7.1, etc.) stereophonic sound.

The above definitions, especially that of stereo, are not accepted by everyone. Some insist that "stereo" must be associated solely and exclusively with two-channel reproduction, while others understand the popular generalization but stick to the description mentioned above. We will not argue here about which of the two is the true one; let everyone choose whichever they prefer. Although it is true that in this article we will be talking precisely about stereo sound as we have defined it: the ability to represent and position sounds within a sound space, thanks to the use of two or more audio channels.

Nor is this the first time we have pointed out that human beings do not, in fact, hear in a purely two-channel way. Constant head movements allow us to locate any sound source in physical space almost in real time. It is indeed possible to control the position of a source using two points, as long as we are working in a two-dimensional space. But our world is three-dimensional, so instead of having three ears, what we have are dozens of sufficiently different samples that, taken together, allow us to locate the source in both the horizontal and the vertical planes.

The first stereophonic "device" was a pair of binaural headphones devised by Harvey Fletcher who, together with Irving Crandall of Western Electric's electrical engineering department, presented them as the result of a study on hearing and speech. That was in 1929, and those headphones led to the first stereophonic recording some time later, when Fletcher changed jobs, this time joining Bell Labs as director of acoustical research. Years later, together with Arthur C. Keller and Leopold Stokowski, Fletcher devoted himself to improving audio recording systems, and at the Academy of Music in Philadelphia he began recording and transmitting monaural and binaural sound. In December 1931, Alan Dower Blumlein obtained a patent in England for stereo recording.

The first record made in stereo was the work of Stokowski, on March 12, 1932. The record had a dynamic range of 60 dB and a frequency range of 10,000 Hz. Curiously, Bell Labs saw no commercial application for the two-channel recording system, so it took 4 years to act on the patent. Meanwhile, Arthur Keller proposed improving the pickup method with the 45/45 system, a system that was forgotten for commercial reasons and was revived and improved in the fifties by Westrex.

The beginnings of stereo, or stereophonic sound, were not at all promising. Those working on it managed to continue their research purely out of scientific interest, although they received little support from the companies they worked for. In fact, part of the interest lay not in music recording as such but in transmission, as was demonstrated on April 27, 1933, when stereophonic sound was broadcast from the National Academy of Sciences to Constitution Hall in Washington. There, several guests were able to hear a symphony orchestra (conducted by Alezander Smallers) performing several pieces of music live. Three microphones were installed (right, center and left) whose signals, via cable, first passed through the hands of Dr. Stokowski who, in Washington, "mixed" the three channels according to his personal taste.

While experimentation continued quietly in recording studios, there was even time to use stereophonic sound as yet another measurement tool. In 1949, General Motors asked Magnecord to make stereo recordings to improve the spatial analysis of noise in its cars. To do so, Magnecord modified its PT-6 tape recorder, a device it had presented at the National Associaton of Broadcasters (NAB) show in May 1948. This modified tape recorder was presented in 1949 at the Audio Fair in New York as the first two-channel tape recorder for recording and playing back 1.5" tapes, with two channels that each had their own amplification system.

In those decades classical music was the star, so most recordings were aimed at recording, manipulating and transmitting this type of orchestra. In the mid-20th century, Emory Cook made the first stereophonic recording of passing trains for his LP Rail Dynamics, one of the first sound effects records in history. His idea was to thrill listeners in the same way that the Lumière brothers had done in the previous century with the train in what is perhaps one of their best-known films. The movies also wanted to take advantage of developments in stereo sound. The Robe was the first CinemaScope film with 4 stereo tracks, leading the 33 productions that would end up in stereo that same year, 1953. But the industry was not ready from a technological point of view, so we would have to wait until 1975 to enjoy good stereo sound again, thanks to Dolby's optical system. It was not that the industry abandoned it, but that there was no standard that even regulated the method of stereophonic recording and playback.

As an anecdote illustrating this "mess", we can recall the suicide of Edwin Armstrong, who threw himself from the tenth floor in Alpine, New Jersey, because of the bureaucratic problems he had with his 1939 invention: FM radio (his wife continued the legal proceedings for 13 long years, and eventually won). FM radio, as we know today, proved to be a frequency modulation system that lowered noise and achieved a better frequency response than the AM radio of the time, which also helped to stimulate stereo sound. That same year, Murray Crosby demonstrated the multiplex system proposed for stereo FM radio on Long Island, in front of 16 RCA executives. His demonstration was the result of a request from Leopold Stokowski to David Sarnoff: it was the first time that senior executives from the music industry had heard it. And that encouraged RCA to launch stereo reel-to-reel tapes, the first being Richard Strauss's Also Sprach Zarathustra, priced at $18.95, recorded by the Chicago Symphony Orchestra conducted by Fritz Reiner. A two-track tape recorder running at 30 ips was used, with two omnidirectional Neumann M-50 microphones. RCA immediately began its own stereo recordings and, almost at the same time, EMI in England did the same, this time at the Abbey Road studios. Decca followed, in this case using three microphones, in what was called the Decca Tree, a design by Roy Wallace. Roy proposed using three directional microphones with a cardioid pattern (Neumann KM-56 condenser microphones), suspended at a certain height above the floor and slightly above the conductor. The recording was made on an Ampex 350-2 at 15 ips.

Three years later, Westrex revived the 45/45 recording system through public demonstrations at the annual AES convention (New York, October 1957). The RIAA (Recording Industry Association of America) adopted the Westrex system as its recording standard on March 25, 1958. That date marks the beginning of high-fidelity components. That same year, the ever-attentive BBC began its stereo broadcasts, initially on Sunday mornings.

Little by little, broadcasters turned to stereo transmission, to the point that John Koss (who gave his name to the prestigious Koss headphone brand) began to promote his idea of personal listening through headphones, although very few audio devices had an output for them.

In 1970, one of the stereo audio systems that has perhaps been talked about more than any other was presented: the Quadraphonic system. It offered the experience of 4 audio channels, a new stereo concept that had one small problem: listening had to be almost personal, since the listening position was limited. Nor did it gain much technological support, barely JVC's CD-4, leaving all the success to the traditional two-channel stereo system.



THE SOUNDSTAGE



When the first symphonic recordings were made in the forties and fifties, even on an experimental basis, the intention was to reproduce the sound space that exists in real listening. The goal is as simple as allowing the listener to discern the position of each of the sound sources.

To understand how stereo sound can be simplified into the two traditional channels, we need only turn to several premises that all sound engineers know when they make not only a studio recording but even a live event. Although the human ear needs two or more pickups to locate a source in a real environment, when listening to music from the front the need to discern whether the sound comes from the front or the rear is, to some extent, lost. Our two ears, from a single pickup, can tell how far away a source is, for example 3 m from our head. But to know whether those 3 m are in front or behind, they will need another pickup. The advantage of knowing that all the sound we will hear comes from the front spares us from discerning the third plane: everything comes from the front.

We always listen facing forward, so it is obvious that our loudspeakers are placed in that position. Panning (placing a source on the plane parallel to us) is achieved by adjusting the level of that source in each of the channels. The more presence it has on the right, the further toward that side we will perceive its location. A sound that is equally loud in both channels will give us the sensation that it is right in the center. This is what we call stereophonic sound.

Simplifying things to obtain better reproduction means that it is standardized with just two independent channels. Phase problems come into play and, therefore, adjustment or setup. With two channels we can already obtain a perfect spatial representation. If "personal" music listening at home poses no problem in this respect, why does cinema use 5 independent channels? First of all, we have to understand that on a real musical stage, and aside from the reverberation of the hall, the orchestra is always located in front of us.

In a movie scene, the "sound" is not just the dialogue, but also the ambience which, on the other hand, we usually avoid in a music recording. This ambient sound comes from all directions, and locating it does not require a third channel to provide all-around positioning.

The addition of the center channel has its origins in the commercial movie theaters of the time, which were much larger than today's commercial cinemas. It counters the Haas effect, a deficiency related to the listener's distance from the two main channels.