There are two ways to divide the different systems available or proposed for use as stereophonic systems. One of these concerns the difference between the program or information contained in the separate channels of the system. The other concerns the kind of material in which the program is conveyed, either recorded or transmitted. As there is no basic relationship between the two ways of dividing the systems, we will consider each separately.
First we will discuss the different ways of dividing the program or information between different channels. For this part, the discussion is concerned with how realistically the program can be presented in the livingroom. From the discussion in the preceding sections we realize that the livingroom is fairly small, compared to many spaces in which sound may be recorded.
Specifically, the lower audible frequencies have wavelengths so large that the livingroom may only contain one wavelength, a fraction of a wavelength, or at the most just a few wavelengths. In any event, frequencies in this range do not have room to travel around as they would in free space. Standing waves are immediately built ·up because of the restricted size of the room relative to wavelength.
At the very low frequencies, where only a fraction of a wave length can be contained in the room, the loudspeaker does not produce a complete wave at all, but virtually pumps air alternately in and out of the room at the desired frequency. So there is no stereophonic system that can do anything to give a sense of direction or discrimination to the extremely low frequencies. The sound coming from any number of loudspeakers will be approximately of uniform intensity and in phase. (If the loudspeakers are out of phase, cancellation will occur in the room, and diaphragms will move much farther than they should.) Consequently, with correct connection the effect must be cumulative, as it would be if a number of loudspeakers were connected to the same single-channel program.
It should be stressed that this remark applies only to the very low frequencies. At middle and higher frequencies intensity and phase differences (according to the theory usually presented) can be of importance. To be more precise, each channel will carry different information about the transient components of the sound in the middle and upper frequency ranges.
Basic Systems
In considering the potentialities of different systems and also examining the performance and the degree of realism they repro duce, we should keep in mind both the critical and the tolerant aspects of the human hearing faculty. It would be utterly fruitless, for example, to go to great expense to accurately reproduce all the different components of reverberation. This will add very little, if anything, to the realism of the presentation. On the other hand, it is worthwhile expending effort to get realism in the reproduction of the transient components that convey information equivalent to the original program material.
The reverberation part is not unimportant; it has to be there to achieve realism. But it is not important to achieve accurate precision in the reproduction of the reverberation.
For the home user economy is an important factor. The system has to be made at a cost that a reasonable number of people can afford to buy, or it is not a commercial proposition.
This applies not only to the reproducing equipment, but also to the recording medium used, where the program is recorded rather than received over radio channels. This we shall discuss in more detail when we come to consider the different media. For the moment we can think of the question of how well the channels are utilized--whether the number of channels we have is used to good advantage--because whatever kind of medium is used to carry the information, the cost of recordings is usually dependent on the number of channels required, and the time.
Closely associated with economy is compatibility. Can the stereophonic program play over a single channel and still give high fidelity? This question is related more closely to whether we use radio, disc or tape, and is discussed later.
1. Two-Channel Stereophonic
This system uses two microphones and two loudspeakers. The microphones may be spaced apart by the approximate distance of the human ear, in which case the proper technique for playback is to have the loudspeakers as far apart as possible, so that in a normal listening position, each ear hears predominantly the loudspeaker reproducing its channel. This way of using two-channel stereophonic is basically an adaptation of binaural reproduction.
The recording is made in exactly the same way as the binaural re cording, but the reproduction uses loudspeakers.
An alternative, and more generally used placement of the micro phones for recording, puts them a foot or two (or maybe even more) apart, so the two channels receive the program with a bigger time and intensity difference than will occur at the spacing corresponding to the human head. This method is based upon the simplification of the stereophonic theory we discussed in the previous section.
Whichever method of recording is used, the program material re corded on the two channels is very similar. Listening to one or the other, the differences require critical listening to detect. But there are differences both of intensity and phase.
It should be noted here that the differences are not just in the same program material: it is not that the composite audio has its intensity and phase modified in the two channels: different components of the audio, such as violins, wood-wind, brass, etc., have varying combinations of intensity and phase.
Thus, if the violins for example are playing nearest one micro phone while the drums are nearest the other, number I channel will carry the violins at higher intensity than the drums, and the phase of the drum reproduction will be just a little behind that on number 2 channel. Number 2 channel, on the other hand, will carry the drums at greater intensity than the violins and phase of the violins will be a short time behind that of number 1 channel.
Even so, whichever method of pickup is used, the time difference between the two channels is extremely small and the intensity is not usually very great either. Consequently you could listen to either one of the channels singly and scarcely note the difference.
It is only by playing them both over separate loudspeakers at the same time that the stereophonic improvement is noticed.
But to achieve this improvement we need two complete recordings--two pickups, preamplifiers, and amplifiers for the whole playback operation, as well as the two separate loudspeaker systems. Each complete channel carries a high-quality full-range presentation of the program, although they are both very similar, requiring critical examination to detect the all-important difference.
If the program is broadcast over the radio, two separate broad cast channels are required to carry the separate stereophonic channels. These may be two FM-channels, or sometimes an FM- and an AM-channel.
With this system, if the two loudspeakers are played fairly widely apart, according to the shape of the room where the reproduction is presented, there is apt to be a complaint about "lack of center". When the reproduction is intended to represent a source of sound equally distant from two loudspeakers (or in the middle) , it is apt to seem somewhat lost. True the impression, if any, is that the sound comes from the middle, but it seems to do so in rather an indefinite way, as if the sound has to go all around the room before you hear it properly.
This effect with two-channel stereophonic has been called a lack of body, because the center, or body, of an orchestra reproduced by this means, seems to be missing. The sound comes over but it has a kind of "hollowness" about it. If the loudspeakers are placed closer together, the lack of center or body is to some extent over come. One arrangement puts them back to back in the same en• closure, facing outward (Fig. 28). This arrangement is quite suitable for many average livingrooms.
This kind of stereophonic reproduction can be very impressive on demonstration tapes, such as an aircraft flying from right to left or left to right, and similar dramatic presentations. But this is not the kind of program you usually want to live with. If you happen to live near an airport (who doesn't these days?!) you get plenty of aircraft flying over, without presenting it over your re producer system!
Fig. 28. Single-enclosure stereophonic loudspeaker cabinet, using loudspeaker
units at each end, with doors as reflectors.
As far as musical presentation is concerned, the two-channel stereophonic gives the best realism on program material with low to mid-range frequencies--organ music, and orchestral music with the string bass, cello, bassoon, tuba and similar low-to-middle tones predominating. Also the bigger drums can sound quite realistic.
This is because the time and intensity differences are apt to be slightly exaggerated as compared with natural listening, a condition arising from the fact that the microphones are further apart than the normal spacing in the human head. This is least noticeable on frequencies where the wavelengths are still at least as large as the spacing between the microphones and loudspeakers.
At higher frequencies, where the spacing between the micro phones may amount to several wavelengths, the directional effects may become quite confusing, because the relative intensity and phase differences can multiply up several times the natural amount.
If you listen to a recording of an orchestral program with the strings played pizzicato, it will give you the impression that there are just about twice as many strings being plucked as there really were, because each microphone will pick up the same string pro gram with just sufficient difference in time to give the impression that each string is plucked twice instead of once.
2. Three-Channel Stereophonic
This is virtually the original stereophonic system. It utilizes a center channel as well as the two side channels, which is the only respect in which it differs from the first system. The use of a center channel overcomes the "lack of center" and improves the body discussed under the previous heading.
Fig. 29. Best position for loudspeakers with three channel stereophonic
is along one of the shorter walls of a listening room.
This system is always used with the microphones spaced several feet apart. Consequently the effect on different kinds of program material is quite similar to that with the two-channel stereophonic.
The best results arc obtained on program material in which the dominant frequencies arc in the low-to-middle range.
The remarks about each channel carrying almost exactly the same information, with just slight intensity and phase differences, so that any channel could equally well be listened to as a single-channel reproduction, applies to this system as much as to the two channels.
FIG 30. Microphone pickup patterns for making the stereo-sonic recordings.
Consequently it means the system is just about three times as costly as a single-channel arrangement, to produce its particular approximation to stereophonic sound.
Unless you have a larger-than-average livingroom, speaker placement can be a problem. The ideal arrangement seems to be with the three loudspeakers spaced along one of the shorter walls of the listening room (Fig. 29).
3. "Stereosonic"
Notice that this is a different spelling. It is a system patented by the EMI [Electrical and Musical Industries.] in England and utilizes quite a different principle from the two-channel stereophonic. It is the only system that basically uses directional microphones and deliberately docs not employ a time difference as well as an intensity difference in the material on the two channels.
The pickup arrangement uses two bidirectional microphones, with their directivity pattern at an angle (Fig. 30). The whole 90° between the maximum points of the two microphones can be used as a pickup area for the program material.
An artist playing on the extreme left-hand side will be picked up only by the left microphone, pointing in that direction, because he will be edge-on to the sensitivity of the right-hand microphone.
An artist playing at the extreme right-hand side will be picked up only by that microphone, while one in the center will be picked up equally by both microphones.
This means there is a much bigger difference between the pro gram content of both channels, but there is no phase difference.
The only difference is that the individual instruments are recorded at different relative intensity on the two channels.
This method of pickup avoids one of the chief deficiencies of the two- and three-channel stereophonic--the excessive phase difference between the program content of the individual channels which at the higher frequencies can produce quite an erratic, resulting frequency response of the system, as well as making the directional effect of these upper frequencies quite indefinite. By recording everything on both channels in exactly the same time or phase relationship, there is no possible conflict from this source, any more than there is from just running two loudspeakers off the same single channel.
The paper giving the theoretical foundation for this system ex plains, with mathematical support, that intensity differences from the loudspeakers produce effective phase differences at the ears of a listener anywhere in the room, so as to give a similar equivalent sense of direction practically independent of where the listener may be situated. This is not achieved by either the two-channel or three-channel stereophonic system to the same degree.
A better way of showing that this principle works (and far more convincing than the mathematics when we remember the discussion we had about the relative importance of continuous tones and transients) , is to demonstrate the effect of varying the amount of power fed to two loudspeakers from a single channel. This experiment is illustrated in Fig. 31.
As the control is turned so the loudspeakers receive either equal power, or more to one than the other, the apparent source of the composite audio appears to shift from the center between the two loudspeakers to the left or to the right. This effect is quite definite, although the power fed to both loudspeakers is in the same time phase at all times, because it comes from the same channel--a single-channel recording.
What the stereosonic system does is to feed varying intensity combinations of different components of the program (different parts of the orchestra) to each loudspeaker. This means the different parts of the program will come from different apparent positions in the same way as the whole program is shifted by using the control of Fig. 31.
FIG 31. The effect of only loudness difference between two loudspeaker:
In the main listening position the best illusion of direction is obtained,
but some variation is also noticeable at secondary positions.
Fig. 32. How the output from the two microphones in the stereosonic system
can be combined to give a major channel 1, and a minor channel 2.
There is another possibility with the stereosonic system. The paper describing the system showed that the program material picked up by the microphones can be combined so as to produce a form of coded channels. The output from the two microphones is first amplified, and then these channels combined in such a way that one output channel takes the two microphone channels in phase with one another, while the other output channel takes the two microphone channels out of phase (Fig. 32). These two combinations or output channels then form the program that is re corded, on what we will continue to designate as channel 1 and channel 2.
Any instrument that is on the 45° line, in between the two microphones, will only be recorded on channel 1. Nothing will be recorded from this position on channel 2 because both microphones pick it up at equal intensity and there is no difference.
An instrument on the most sensitive position of the first micro phone (which is the "dead" position of the second) will be re corded on channel 1 and also on channel 2, both in the same phase, we will say. However, an instrument on the most sensitive position of the second microphone will also be recorded on both channels, but this time with channel 2 out of phase with channel 1. Notice that with this arrangement, channel 1 carries all the program material in almost uniform intensity, while channel 2 carries only the program at the two sides, with practically nothing from sounds originating in the middle.
This method has the advantage that channel 1 may be used as a single-channel program and played over a single loudspeaker, maintaining the balance of the original program material. With the original method, neither microphone gives a truly balanced pickup, because an instrument on the maximum pickup position of one microphone is not picked up at all by the other or vice versa.
Another advantage of this arrangement is that channel 2 carries only a minimum amount of information and thus does not need such a high-fidelity recording channel as channel 1. This may prove to be a considerable asset of the system. While all the other systems so far described, require--as many channels as are needed--of uniform fidelity, this one will give quite successful performance with only one channel of maximum fidelity. The system will give almost indistinguishable performance if the extreme low frequencies are completely missing from the second channel and also if some of the high frequencies get dropped off in playback.
Of course, the channels are recombined in the same way, by taking sum and difference arrangement after the first playback stages, producing recombined loudspeaker channels (Fig. 33). This method involves a slightly more complicated electronic equipment, but means it should be possible to effect economy in recorded program cost, or programs for transmission over two separate radio channels. The highest fidelity channel, for example the FM, can be used for channel 1, while a relatively low-quality audio channel, an AM transmission, can be used for channel 2, without noticeably deteriorating the quality of the resulting stereo phonic presentation.
4. Coded Single-Channel Stereo
At the moment of writing no such system is available for home presentation of stereophonic recording. It is presented here for two reasons: (1) it is known that certain companies are experimenting with a variety of possible systems; and (2), this approach gives at least a theoretical possibility for considerable economy in effective stereophonic presentation.
In its simplest concept, it differs from all other systems of stereo in that only one channel of audio is needed. In the stereosonic system only one high-fidelity channel was needed and the second channel could have quite restricted range, but in this system only one channel of audio is used. To effect different distribution of sound between the two or three loudspeaker channels used, some kind of coded information that is not audible has to be recorded, along with the program material that is audible.
FIG. 33. Recombining on playback with a simple electronic "adding" and "subtracting" arrangement
restores the original A and I Information picked up by the stereosonlc
microphones.
The code information can either be on a separate channel, in which case the channel separation prevents it from being heard, or it can be recorded in some inaudible fashion on the same channel with the audio. The latter, of course, represents a very considerable economy over other systems.
In this case it can use either sub-audio frequencies, that is, frequencies that are too low to be audible, both in frequency and intensity, or it can use supersonic frequencies--frequencies too high to be heard. Yet a third possibility is that it use a narrow channel of frequency right inside the audible band, but the system takes special precautions to see (a) that the control frequency is not audible, and (b) that program frequencies do not operate the coded control.
The use of supersonic control-code frequencies means that the recording mediums must be capable of maintaining good reproduction up to these frequencies, without any intermodulation from frequencies that are audible. This proves to be a serious restriction.
The last possibility mentioned means that unless the recording is played over the system designed for it, the code frequency will be audible in a single- channel reproduction. This would be a serious drawback, because one of the big advantages of a single channel, coded stereophonic is that it could be made compatible with existing single-channel systems, so that if a coded stereo record or input were to be played over an ordinary high-fidelity system it should give good reproduction (but not, of course, any stereo effect) . Using this third possibility for coding would, at whatever frequency was chosen, produce a peculiar sound that had nothing to do with the music.
This would appear to leave us with the sub-audio control frequencies as the most logical solution. They also happen to be most logical in relation to the kind of coding information they need to carry. We will not take up much space on this as the system is not vet obtainable, but a short description of how the method works to achieve a satisfactory stereophonic presentation is included so the prospective merits can be judged.
Such a system can be designed to feed either two channel or three channel loudspeakers (or even more, without requiring more recorded channels). In the two channel variation it differs from the stereosonic in that both loudspeakers carry exactly the same program material at all times but in different overall loudness from time to time. In the illustration we used for the stereosonic, the violins could be playing at the left and the drums at the right at the same time, and each would be predominantly reproduced by its own loud speaker. With coded stereo this is not strictly possible.
However, a little thought will show that a program of this type is rather unusual--especially the desire to be able to differentiate between the two all the time, through the program. In practice we do not try to hear the different parts of the orchestra separately unless the part we are interested in happens to be playing what is essentially a solo--the rest of the orchestra playing merely an accompaniment for the solo part. In these circumstances the com poser directs our attention to different instruments in the orchestra in turn by his composition of the music. This may be quite a quick succession of transfer of interest from one to another, but just the same it is a transfer. We do not actually try to listen to both at the same time.
In quite a lot of modern music there is a primary melody that is possibly played by the strings, together with a secondary musical theme that appears in the intervals of the primary melody, the secondary theme being played by some other musical instrument, possibly the wood-winds. Coded stereo can adequately take care of this situation. It concentrates the attention on the appropriate loudspeaker each time the strings play the primary, and refers our attention momentarily to the other side every time the secondary theme comes in.
From demonstrations of this system in experimental form it has been found capable for the presentation of a great variety of pro gram material. The impression conveyed is at least as realistic as the ordinary three channel stereophonic and sometimes more so.
For example, when strings play pizzicato this system does not pro duce two or three times as many strings!
Fig. 34. The basic working arrangement of single-channel coded stereo.
The coding works by changing the amplification of the same composite audio fed to each of the loudspeaker channels (Fig. 34). The system has another advantage: that it will improve the effective dynamic range of the medium used. At present, although high-quality tape-recorders such as those used for professional purposes, can achieve a dynamic range about equal to a disc recording, this is not possible in the low-priced recorder suitable for home high-fidelity use. Consequently, as used for high-fidelity playback, disc recording still gives the best signal-to-noise ratio, or dynamic range. Tape-recorders lag behind in this respect.
The coded stereophonic system helps to overcome this by providing what is effectively a volume expansion, when the program is playing quietly the amplification of all three channels (or two channels, as the case may be) can be turned down so as to reduce background noise as well as the program loudness. When the full power of a crescendo is required, the two or three channels arc turned up to maximum amplification and thus a greater effective dynamic range is obtainable than with any other stereophonic system.
Stereophonic Program Mediums
So much for the systems available, or potentially available, for use in the home. We now come to consider the different mediums that can be used to present the program material over these systems.
Basically the source mediums, as the home user sees them, are radio, disc and tape. There are other hypothetical possibilities but they probably will not be used for home systems. In discussing each we want to see how they will line up for home use, on the score of:
(a) Simplicity to use: they should not be difficult or complicated to put on.
(b) Economy: recordings should be available at a price suitable for as many as possible to obtain copies.
(c) Dynamic range: they should be able to present the fullest possible dynamic range in recorded material; and
(d) Compatibility with other systems.
This last feature also affects the cost question, because it will mean that recordings do not have to be purchased exclusively either as single-channel or stereophonic; a single recording can be issued that will serve both purposes if the method of presentation is compatible.
1. Radio
This is not, strictly speaking, a medium in itself in most in stances. Occasionally perhaps, live programs will be broadcast over stereophonic channels. But most often stereophonic presentation of program material will be broadcast from some kind of recording.
Consequently the radio is not a medium in itself but a link between the broadcasting station and the listener, and as such is constitutes the input to his system.
It also has an economic factor from a different aspect than the other mediums, because of the limited channel allocations avail able from the FCC. For two- or three-channel stereophonic presentation the ideal arrangement would be two or three FM-channels as near as possible to the same frequency allocation, so any fading or fluctuation in quality on the different channels keeps more or less uniform, and can be adequately controlled by the AGC and the AFC action of the receiver, more or less in sympathy.
Use of the FM- and one AM-channel for two-channel stereophonic has been fairly successful over short distance transmission where the quality of reception is not subject to appreciable variation. For transmission over even medium distance this method is to be deprecated because of the varying way in which the totally different kinds of transmission will fluctuate in quality, producing some quite spurious stereophonic effects not intended in the program.
For stereosonic presentation the use of an FM-channel for channel 1 of the combined signal, with an AM-channel for channel 2, would be quite satisfactory because of the lesser degree of fidelity required on the second channel.
Similarly in transmitting coded single-channel stereo: this could, by some sacrifice at the low frequency end, be transmitted entirely over a single FM-channel; alternatively, by using an AM-channel of extremely restricted bandwidth and therefore low inherent noise pickup, the combined FM and AM arrangements should produce extremely good transmission results with coded stereo.
In both the stereosonic and coded single-channel stereo trans mission over radio, the fact that the important, high-quality audio is all transmitted over one channel, minimizes the possibility of spurious stereo effects being introduced by noise or fluctuation in transmission quality.
If there should ever be any extensive interest in stereo broad casting, to the extent that live program transmission would become a common thing, there is a lot to be said for the stereosonic system of working, with FM- and AM-channels. The stereosonic system has a great advantage of simplicity of control for the audio engineer at the studio. This system requires no "fudging" to get the de sired stereo effects.
The two- and three-channel stereophonic require careful monitoring to ensure the desired effects being put over correctly, while the coded stereo requires careful gain-riding by the mixer when recordings are being made. Careful rehearsal and reworking can get an ultimate product that gives the desired results all the way through. But in processing live program there is no opportunity to rework if an error is made, even though there may be opportunity to rehearse the program before the actual transmission time.
The stereosonic system also has an advantage of compatibility, in that the program will sound like single-channel if only the FM is picked up. Coded stereo transmitted the same way would be equally compatible for single-channel presentation. However, stereosonic and coded stereo would not be compatible for one another.
2. Disc Recording
When tape first appeared as a recording medium, many people prophesied that it would eventually out-date discs. So far it hasn't.
High quality tape-recording machines, with the best quality tape, can produce a dynamic range far better than the disc records of a decade or so ago. Even the early tape-recorders, and tests with carefully developed tape, showed this was a possibility. However, modern development with discs and with improved phono-pickups, has brought us to a stage where a well-recorded master, used with careful pressing technique, can produce a disc with at least as much dynamic range as the best tape.
With both tape and disc, the question of the ultimate in dynamic range seems to be a combination of two factors: (a) How much care and attention is given to the development of the system and better materials; and (b) How much you are prepared to sacrifice, quantity- and price-wise, in achieving an increased dynamic range.
For disc recording, it is possible to increase dynamic range by utilizing a greater modulation width--having the groove "wiggle" further. For tape recordings, the maximum magnetic density on the track is set by the saturation density of the tape. But dynamic range can be improved by using either a higher speed or a greater track width. Use of a I-inch track width, with very carefully aligned heads to ensure good reproduction of the extreme high frequencies, would improve the dynamic range by 10 or 20 db on the quarter inch tape now used as standard. So the argument about relative dynamic range between disc and tape proves to be dependent upon other factors than the simple choice of one medium or the other.
The big advantage claimed for tape is its long-playing time.
By using the very thin tape now available, a 7-inch reel will play two hours each way on a half-track recording at 3¾ inches per second. At 7½ inches per second each side plays for one whole hour. This is certainly a considerable improvement in playing time over anything in the corresponding "bulk" on disc.
However, where the disc recording scores--and this seems to be what promises to keep discs in business indefinitely--is in the readiness with which a certain passage can be selected. The modern LP record with from 4 to 6, or even more, recorded bands, arranged with run-on grooves from one band to the next so that if desired the whole record will play continuously, make it very easy to pick out one particular band of a selection at a moment's notice.
Simplicity of use is another advantage of disc. Even the most conveniently arranged tape-recorder requires a little more initiation to use than does a disc-player. But this advantage is about re versed when comparing the Cook binaural disc with stereo tape.
With a tape recording such selection is by no means so convenient. If the tape comes with a footage register for each of the bands, and the tape-recorder has a footage indicator, it is possible to use fast forward or rewind to find a particular place and play just that section. But even this takes a little more time and attention than just placing the stylus at the beginning of the band you want to hear.
Fig. 35. The clip-on cartridge holder to utilize the two-track type of
binaural disc recording. Courtesy Cook Laboratories.
FIG. 36. Different form, of stylus movement used. In disc recording: (A)
vertical (or hill and dale) used In early phonographs; (B) modern lateral;
(C) the experimental single groove stereophonic uses both.
Of course, tape shows much improved prospect for several kinds of stereophonic presentation. To apply stereophonic recording, of the conventional type at any rate, on disc, one needs to have two or three styli. The Cook binaural record (Fig. 35) has two bands of recording concentrically arranged, with a standard spacing between the pickup styli. The two pickups arc carried on a common arm and the record delivers simultaneous output from the two cartridges to the two-channel amplifier.
Some care is necessary in placing the styli in the grooves however, because with the modern microgroove recordings which this uses, the grooves arc very close together and some latitude must be allowed so the stylus can adjust itself to compensate for the slight variations in tracking at different positions of the arm. This means it is easily possible for the stylus to be one or two grooves away from the one corresponding to the groove the other stylus is playing.
Naturally the playing time is approximately half, because the number of grooves per inch is on the same order as for regular LP records, and the space has to be divided between the two bands for the two pickups. Also it is not so practicable to use separate bands as in regular LP's because of the difficulty in getting both styli into the right groove at points other than the beginning of the records.
This method of using discs for stereophonic presentation does not seem to offer too optimistic a future. Three-channel would be even more of a problem than two-channel.
Fig. 37. How use of the stereosonic system, with major and minor channels,
fits in with single-groove disc, 10 the stylus movement reproduces air-particle
movement of the stereophonic wave at the microphone.
The question next arises whether some other kind of disc presentation, using only one groove, might not offer an economic and versatile possibility for stereophonic presentation. Two possibilities might be considered: (1) that all the modulation is contained in a groove with lateral vibration; or (2) that the groove provides movement in two directions, up and down, as well as sideways.
The latter method (Fig. 36) would enable twice as much information to be carried in the one groove, and for a long while this possibility has intrigued designers as a means for applying two channel stereophonics to disc. There are two problems: (1) a very intricate and accurate pickup is needed, and (2) it is difficult to avoid "cross-talk" between the up and down and sideways movement, due to the complicated way the stylus has to move in following the groove.
The stereosonic does offer one improved possibility in this regard. It arises from the fact that the "sum and difference" mixing method puts the main "intelligence" requiring high-fidelity in one channel, while the other channel requires only "difference" information. In this system the lateral vibration of the stylus, as used in all modern phonograph recordings, could be used for channel 1, while channel 2, containing only the "difference" in formation, could be the vertical one.
This method makes a certain amount of sense on another score: the fact that the stylus movement will now be in exact correspondence with the air particle movement in the vicinity of the pickup microphones. The vertical component of movement one way will combine with the horizontal movement to make the stylus move at 45°. In opposite direction of vertical movement the stylus will move at 45° but in the opposite direction--at 90° to the previous one. This corresponds with the line of movement in front of the original microphone (Fig. 57) . The fact that both channels contain program in exactly the same time phase, but in different relative magnitudes, means that any cross-talk that occurs would not produce serious interference effect as with other systems, but will only slightly modify the directional impression conveyed by the system.
It still remains of course, to satisfy condition (1)--to produce a high-quality pickup that will work with this kind of disc re cording, and also to figure out a way of pressing such recordings which are somewhat more complicated in the kind of engraving produced, than the ordinary lateral-groove recordings used for modern LP and other kinds of records.
Another possibility is that the vertical movement of the stylus would convey the coding frequencies for the coded stereo. As the up and down undulations would be only at the very lowest frequencies in the audio range, this would considerably reduce the cross-talk problem as compared with any system that requires both directions of vibration to contain the same total frequency range.
The stylus movement could be made quite flexible in the vertical direction so that the transmission of the vertical movement does not get into the lateral transducer.
This possibility may make for easier design in the rather complicated pickup needed. On the other hand extra care is needed to make sure the vertical vibration does not get into the horizontal movement, because unlike the stereosonic, in which such a transfer would not produce any form of distortion or unwanted program, this method would produce an unwanted low-frequency buzz.
The alternative way of using a single groove for stereophonic recording requires all the intelligence to be combined in some way, using lateral, or side to side, vibrations only. This means different frequencies must be used for different purposes. If coded single channel is used, space has to be found somewhere in the frequency spectrum for the coding that will control the distribution of the audible frequencies.
If multichannel is used, only one channel can use the actual frequencies we hear. The other (s) must use supersonic frequencies by conversion. With this system, accepting a response limited to 12,000 cycles on each channel, a two channel system would require response to 24,000 cycles. To get three channels into the same bandwidth would require each to be restricted to less than 8000 cycles. As the very high frequencies arc very susceptible to damage in a disc, these possibilities arc not very practical, because the channel occupying the higher frequencies would be apt to disappear altogether. Another disadvantage is that they are not compatible, because such a disc could not be played on an ordinary system.
Fig. 38. Multichannel single-track (or groove) can be achieved by frequency
conversion, but this method has several disadvantages (see text).
Fig. 39. The difference in position of recording on the two tracks for
stacked (A) or staggered (B) heads.
Suggestions have been put forward for other methods of making three-channel stereophonic on disc, but they are all too complicated to be regarded as practical, so we shall not go into them right here.
3. Tape
The great advantage of tape as a medium for stereophonic re cording is that it is relatively simple to put as many tracks as we want on tape. If we could not get three tracks comfortably into a ¼-inch tape, the width could easily be extended. In practice some good three-track tape has been produced within a ¼-inch width.
The principal problem is in the design of heads of sufficient durability and sensitivity to give good dynamic range. Three tracks on the tape require to have two spaces between them, as against only the single space between two tracks. Consequently quite a large proportion of the tape is unusable because of allowing sufficient spacing to prevent cross-talk. The use of narrower track means the output from the playback head is that much lower for full magnetization of the tape. This means the dynamic range is reduced.
Apart from this, use of stacked heads (i.e., with all the three tracks in line) means that the air gaps have to be in line and consequently the head windings need to be very close together.
This means only small windings can be used, and consequently it is more difficult to produce an efficient head. Use of staggered heads improves the possible efficiency of the pickup heads. This is an alternative system, where the tracks are spaced apart lengthwise so the heads running along the different tracks can be separated (Fig. 89). But the use of staggered heads makes it still more important to avoid cross-talk between channels. Cross-talk between channels of a stacked head would merely mean that some of the program intended for one loudspeaker would break through into the channel for another loudspeaker. Using staggered heads however, means that break-through of this nature results in program getting from one loudspeaker channel to another, but at a different time, producing some very spurious reverberation-like effects. Consequently adequate track spacing, and the use of only a very narrow track must be very carefully followed to avoid this kind of thing happening.
Tape is also a very suitable medium for either stereosonic or the coded single-channel stereo. The code can be carried either on the same channel by the use of filters (Fig. 40) , or it can be run on the separate track. A possibility here, that may be worth considering to improve overall dynamic range, is to use a somewhat wider track for the audio and a narrower one for the coding. This would improve the possible performance of tape-recorders, and make them easier to produce at a lower cost, because adjustment to get a good dynamic range without excessive background noise and hum would be much easier to achieve.
Fig. 40. Two possibilities for coded stereophonic on tape: (A) using one
track only, enabling the tape to be used "two-sided" as for single
channel; (B) using two tracks of different width to achieve much superior
quality.
A final advantage for tape as a medium, at least in a temporary sense, is its excellent possibilities for doing experimental work. For any of the disc systems the whole thing is entirely in the region that must be explored experimentally by professional people--companies with laboratory facilities. But the use of tape as a medium brings the experimental work into the scope of the average home user. Extra tape heads can easily be bought as separate items and mounted up on the deck of an existing tape-recorder. Extra electronic equipment can be built to produce necessary coding frequencies or extra amplifying channels, so the home user can easily build up his own system and try out different variations and combinations that may appeal to him.
4. Other Possibilities
When a satisfactory basic system has been developed by necessary experimentation on tape, it is then not impossible that the same basic system would be transferred to disc for the greater convenience which this medium gives--that is, if it is possible either to con dense the total program material into a single channel so a lateral groove could be used, or into a form suitable for lateral and vertical at the same time.
Ever since the advent of recorded program material, other methods of making the recording have been explored. While to date the only popular forms are disc and tape, there are others in the experimental stage that may still ultimately come to the fore again.
Wire of course, was another form of magnetic recording that was popular before oxide coated film in the form of tape took its place. Tape is so much more manageable that the use of wire has become almost extinct. Wire is very difficult to handle, especially if it should break, when it inevitably becomes tangled up in the most impossible manner.
Another possibility that has been explored is different kinds of optical film. Various sound tracks as used in the motion picture industry have been tried out from time to time for home phonograph purposes. Unfortunately, none to date have proved to be competitive, either in quality or cost.
An alternative system that uses a very similar principle is the Miller system. This does not use optical recording but employs optical playback. The recording is achieved by means of a wide angle cutter on a film base consisting of transparent plastic coated with a black layer. The wide-wedge cutter plows into the plastic and removes a variable width of the black layer, producing a track very similar to the variable-area track made optically. This system has the advantage that the photographic processing is unnecessary, and the cost of the cutter-head is probably lower than the complete optical system necessary to produce optical recording.
Playback, of course, is almost identical to any optical playback arrangement, so for the home user buying a complete recorded program to be played at his leisure, and not particularly interested in making his own recordings, it is immaterial whether a recording is made optically or by the Miller system. The playback requirements are the same.
(Adapted from: Stereophonic Sound (1957) by Norman H. Crowhurst)
= = = =