From the first experiments which Edison conducted with wax cylinder sound reproduction, the objective of sound reproduction was to accurately RE-produce an audible event, whether music, speech, or some other form of sound. Not surprisingly, music reproduction was an early mainstay of both the cylinders and the flat Berliner discs, and thus the idea of reproducing a symphony orchestra in a living room was born. In the nineteen thirties, the Bell labs conducted a series of experiments with three channel stereophonic sound, and produced some discs of these sounds. At about the same time, Alan Dower Blumlein invented a stereophonic recording and reproduction system, in British patent #394,325, filed on December 14, 1931. In his patent discussions of spatial illusion, spatial perception, microphone techniques, surround sound, the stereo phonograph record, and stereo radio transmission, Alan Blumlein anticipated nearly all of the body of theory which has come to be the body of fact in twentieth century sound reproduction.
The idea behind high fidelity reproduction is the RE-PRODUCTION of an original acoustical event at the highest possible level of fidelity, and as many have observed, here, there are many problems. First, in remaining with the classic examples, we can consider the original acoustical venue itself, the concert hall. On the stage is the modern symphony orchestra, a group of perhaps one hundred players, seventy five feet wide, forty feet deep, and seated, and some standing, four and one half to six feet or more high. We are in the audience, and if we want a nearer field experience, in the first two or three rows. Now, in keeping with the technical questions asked by the BELL LABORATORIES, in their concert hall recording experiences, what if masons came and bricked up a wall that extended from one side to the other of the proscenium arch, and from the top of the arch to the top of the stage.
Now, the orchestra would be in a relatively small six sided room, and we would be in a relatively much larger six sided room. The members of the audience, and the members of the orchestra would have a wall in common, and the BELL LABORATORIES question would be, how many bricks do we remove, and from where in the wall should they be removed to produce stereophonic, or solid, sound?
The answer turned out to be deceptively, perhaps too deceptively, simple: remove some bricks from the left hand part of the wall, at some point from a vertical wall center line, and some bricks from the right hand part of the wall, at the same relative point from a vertical wall center line as the left hand removed bricks.
There are conceptual and actual problems with this solution: now there are two discrete sources of sound where there was only one, large, source of sound before. Also, as others have pointed out, a large percentage of the right hand section sound of the orchestra will now also come through the left opening in the bricks, and the same is true for the left hand section sound as well: whereas the full orchestra was a large scale wide monophonic source, that is the musicians do not sit in two discrete groups on either side of the stage, now we have two discrete sources. What made the wide continuous single large symphony orchestra source stereo was the two ears of the listener, not the two artificial sound sources. Further, if the air molecules in the concert hall are the transmission channel from the orchestra to the listener, how many channels are there in the concert hall? One, and only one.
The BELL LABORATORIES added a third, center channel in their experiments, which was dropped from the stereophonic record, and stereophonic tape software incarnations of sound reproduction, perhaps because discrete three channel sound is difficult, if not impossible to achieve on record, or because three full channels of sound are difficult to distribute in the listening room.
There are also some thought experiment clues in the brick experiment above for home stereo reproduction. Microphones are essentially placed relatively close to the orchestra, as initial surrogate ears for the eventual listener. If the plane of the microphone is extended, then we can imagine microphones in the plane of the brick wall between the orchestra and the listener. When the signal from the microphone is archivally stored, and then played back, where should the playback loudspeakers be. Out in the concert hall with us, or in the plane of the dividing wall itself? The problem introduced here is the first and second venue problem. When a recording is made in a concert hall, if full high fidelity is to be served, all of the acoustics of that hall are stored for playback. Our playback room, no matter what the acoustics of that room are, is not the original venue, it is a second venue. Any acoustical contribution which our room makes to the sound is a form of distortion, an addition to the original acoustics of the original performance hall.
Another thought experiment will further clarify this idea, an example drawn from audio life. Imagine a pair of free standing loudspeakers, out some distance from the rear wall. The sound does come directly from the loudspeaker to you; it also goes to the rear wall, and is reflected back, past the loudspeaker to you. The reflected sounds are a contribution from the second venue room. If a full range, low distortion loudspeaker is mounted directly on the reflective plane of the wall, the front radiation from the loudspeaker, and the reflected radiation from the rear wall are concident in time and in space; one entire class of room reflections has been eliminated.
The basic idea then, of stereo reproduction, or multi-channel reproduction, is to minimize the contribution made by the second venue, as any acoustical contribution from the second venue is necessarily a distortion of the production acoustics of the first venue.
By Stanley Marquiss