As soon as you start to talk to recording studio engineers they all express one thing in common. Their other expressions on the subject of stereophonic recording may differ quite widely. But what they all say is that sometimes a stereophonic recording will come out beautifully while at others the thing seems to be a complete flop. This, of course, is quite apart from the kind of faults that can develop in any recording session.
If you pursue matters, expressions will differ. By "coming out beautifully", some will say the production is "realistic", while others freely admit they are not looking for realism, but for a satisfying impression, or in some cases an impressive satisfaction!
Different Schools of Thought
Some studios have used pressure, or unidirectional microphones exclusively, partly on the theory that the human hearing is a pressure-sensitive device rather than a velocity-sensitive one, and that the directional faculty is achieved only by comparison of what two separate ears pick up. Consequently, based on this theory, these recording studios have worked entirely with pressure-type microphones spaced apart at various distances.
This kind of technique leads to a preferred position above the performers and slightly to the fore. This is necessary because keeping on a level with the performers, it is impossible to get close enough to them to get (a) sufficient discrimination against an exaggerated reverberation effect, and (b) sufficient differential be tween the two or three channels. The only way to get "close enough in" without being too close to some, is to use an overhead position. Our comment would be that using a "natural" pressure microphone evidently necessitates an unnatural position! Some studios on the other hand, have expressed themselves as finding directional microphones--either of the cardioid or the bidirectional pattern--much preferable. In most of the larger studios with normal reverberation characteristics--that is, characteristics usually associated with rooms or auditoria of this size--a cardioid gives the best results, because the back pickup of a bidirectional ribbon still makes for a somewhat higher reverberation than seems natural.
The best way to use a ribbon successfully in these circumstances is to completely rearrange the performers so that back and front are used as live pickup directions, and everyone is just a little nearer to the microphone than they would be if all were arranged on the same side (Fig. 41). This introduces another factor en countered by the recording studio engineer.
Musicians are Important, Tool
This is the extreme desirability of not interfering with the musicians' positions or playing technique more than is absolutely necessary. Musicians are people with a long background of experience playing in orchestras, using a more or less normal layout that has conditioned each one's hearing to a certain musical perspective in his work. Each one can only produce his best performance when he occupies as nearly as possible, a position that gives him this "conditioned perspective". The conductor is accustomed to placing the musicians in a position that suits him. Other musicians are used to a comparatively limited variety of positions, dictated by various conductors with whom they may have played. If the conductor is to get the performance from his musicians that he expects, he must use his well proven techniques. To give him a position with musicians placed fore and aft in different directions to suit a peculiar microphone pickup pattern, would completely throw him from his inculcated ability to achieve tonal balance between the different sections of the orchestra.
The sounds he normally hears from specific directions will now be coming from completely new directions, at least as he hears them, although the microphones may get the correct overall impression for the two channels recorded. An arrangement that gets correct balance in pickup, but disturbs the musicians so they cannot give their best performance, loses more than it gains.
The same argument opposes the use of studios with deader-than usual characteristics. Some of the recording companies have developed studios of normal size but much shorter reverberation time.
This makes it much easier to use whatever type of microphone and microphone placement may be desirable for achieving the balance or stereophonic effect required, without having to consider what the reverberation effect will be. Then whatever reverberation is needed can be added by means of a special reverberation room or echo chamber.
This makes the recording engineer's job very simple, except for one thing: the musical director has serious difficulty in handling the musicians and their conductor, because these artists find the acoustic environment extremely unnatural to work in, and they can not give their best performance under conditions so strange to their aural conditioning. Some engineers have dubbed musicians crazy, hide-bound, fussy and similar adjectives, because they adopt this "difficult" attitude. Correct environment is as necessary for a musician to give his best performance as good microphone placement is to make a good recording of it. Further, the rather-dead studio with make-believe reverberation is an artificial method of achieving a close approximation to the intended illusion, without simulating the apparent conditions at all.
This can hardly be called "realism"!
FIG 41. Comparison of distances from microphone necessary to cover the
so total area with (A) cardioid microphone, (B) ribbon, using only one
aide, and (C) using both sides of the ribbon.
Another approach is to use more or less normal studio characteristics but bring the microphones much closer in to the artists, and if necessary use a greater number of microphones, almost one per artist, perhaps. This method uses a general pickup microphone--or three or four--and then adds individual microphones to pick up solo instruments--or groups that are important to the overall effect. The latter are operated at a level to dominate the general pickup.
This method of achieving tonal balance, and a balance between the direct sound from the performers and the reverberation, has been used extensively for single-channel recording. But some of the stereophonic proponents regard this as being too "artificial" to apply to the new medium. We should eliminate all the "trickery" and get back to "genuine" recording, now that we have a stereo phonic medium to give "exactly the real thing". Those who adopt the latter attitude appear to be the very ones who most often find that their results don't come up to expectations. It would seem that when they do come up to expectation, it is something of a fluke, rather than because they have conditions more accurately correct.
Some companies have done a considerable amount of work with stereophonic recording and have gone to almost endless pains trying different types of microphones and microphone placement.
It must be conceded that the presentation should be as realistic as possible although we cannot expect identity with the original, and that we should be able to establish an ideal microphone arrangement for each session with some theoretical "justification". But suitable theory has yet to be completely defined.
... And The Composer
For this reason a dear apprec1auon of the principles stated in the earlier part of this guide should help to improve the approach to stereophonic recording. First let it be realized that there is no such thing as a single correct microphone placement pattern, for stereophonic recording in a given studio. There is not even a correct microphone placement for a given orchestra playing in that studio. Microphone technique has been found to be dependent, among other things, upon the precise piece of music the orchestra is playing--particularly upon who the composer is.
This statement may at first sound heretical, but a brief consideration will show there is foundation for it. Each composer has taken the variety of musical instruments available to him and orchestrated his composition in such a way as to produce the sound effects he intends. Different composers have characteristic ways of using the musical "tools" of the orchestra.
Some use them in successive order, playing one group and then another, so that groups have a separate entity. Others make the different groups play simultaneously, but still with individual parts or characters. Yet other composers hear the sound as a complete merging whole, and the music does not benefit by having the different instruments "separated"; the composition was never designed that way. Differences in the composer's use of orchestral instruments considerably affect what stereophonic sound can "do" for his music, and consequently what microphone techniques will best serve the purpose, or how stereophonic sound can enhance the particular presentation.
Perhaps, in the future, composers will write music especially for stereophonic reproduction. But music written in the past has been intended for live performance in specific surroundings. Musicians and engineers should work together to create the best possible illusion to follow the composer's intentions.
Monitoring and Playback
Listening to an orchestra in an auditorium our hearing faculty can hear the musical instruments somewhat in stereophonic perspective, but mainly the quality is due to the separation between the direct sound coming from the musical instruments and the reverberation coming from all around the hall. Whatever microphone pick-up arrangement we may use, it is not possessed of anything like a human brain to differentiate between the individual sounds that each microphone picks up, and convey the overall result of the sound to a recognition faculty that separates direct sound from reverberation before transcribing it for magnetic or electrical re cording. We cannot have such a simple transfer of "intelligence" as occurs in the human brain, listening to a large performance.
This being the case, we must endeavor to create the most successful illusion for the particular musical performance in hand.
In the recording profession, some idealists are endeavoring to adhere to a particular technique without modification, placing microphones in specific positions, adapted as little as possible from the theoretical ideals with which they started out. With some performances this can be quite successful, but the success is more by accident than design.
Those without any inhibitions due to preconceived notions as to how it should be done, who thus leave themselves free to experiment, have discovered that different microphone arrangements will give best results, even in the same auditorium and with the same orchestra similarly placed, when different musical compositions are being handled. These are not the only differences ...
The next difference we consider is one which can be a real source of confusion, the fact that the best microphone placement depends not only on the sound being picked up, but also on the manner in which the channels recorded will be played back or reproduced. During recording the engineer and musical director use a 2-speaker or 3-speaker system in a monitoring room to judge how the recording will sound when it is played back. The monitoring room is built to be as near as possible representative of an average livingroom in which the recording is intended to be played.
But the loudspeaker placements can differ. It may be one of the double-ended loudspeakers, or two separate loudspeakers at different spacings along a wall. In the same way that loudspeaker positioning will modify the effect of a specific recording on play back, the positioning chosen will also modify the choice of micro phone position to get the best results with this particular loud speaker placement.
Consequently a recording produced to get ideal results for ex ample, from a loudspeaker consisting of a single enclosure with the units mounted in each end, will not give such good results when played over a system where the loudspeakers are mounted in the corners of the room. Correspondingly, a recording made in which the monitoring arrangement used loudspeakers either in the corners of the room, or spaced a little distance apart, will probably not sound so well over a loudspeaker in which both units are mounted in the same enclosure.
As each studio has standardized on the loudspeaker placement for monitoring, a result of this will be that the user will probably prefer recordings by a studio that uses a monitoring loudspeaker placement most like the loudspeaker system in his own listening room. This will undoubtedly lead to arguments and conflicting preferences as to which studio turns out the best stereo recordings.
Fig. 42. The danger in mixing more than one microphone that picks up the
same instrument is the kind of erratic frequency response shown here.
Recording Technique
Some studios allow themselves a little more leeway in making up the final masters for ultimate release, by recording the original program on more than the final two or three channels. Some use four, five, six or more separate channels to record the original program, and then work on these in the laboratory, combining them in various proportions until the most satisfactory illusion is achieved.
While this procedure does save time for the set-up and occupation of the musicians, all of which is quite expensive, and allows the delicate work to be done more cheaply by a few engineers working at their leisure, it has serious limitations. There are definite limitations to the way in which several channels can be combined to make one or more common master channels that are satisfactory.
Some studio engineers express a preference for using as few microphones as possible, and the reason is not difficult to see. When more than one microphone picks up sound from one or more of the instruments, the combined effect on the ultimate recording will have a phase difference between sound that is brought together in the same channel. If all the separate channels are reproduced over separate transducers, a spatial effect of some kind can be achieved.
At least cancellation or partial cancellation at some frequencies, and augmentation at other frequencies due to phase difference, will not occur.
But combining separate channels into one, with phase or time differences from individual components in the musical program, will introduce an artificial "liveness" to the reproduction that is not natural. It will produce peaks and valleys in the effective frequency response of individual musical instruments, in spite of the fact that the individual microphones may have extremely smooth frequency response (Fig. 42) . This is why some engineers prefer to use not more than two to four microphones (in the latter case two on each channel) , to make the original recording.
However, the opposite technique, using microphones for each soloist or each group of instruments, avoids the problem for an opposite reason. Each microphone only picks up a comparatively small "area" of sound at the final intensity. Any stray-over from other areas will be at a much lower level, so it does not interfere with the sound picked up by the local microphones concerned.
The trouble with using a large number of comparatively local or close-in microphones however, is that the reverberation is reduced to an unnaturally low level. This can be added, it is true, by the use of an echo chamber, but then the result is even more artificial.
For this reason the most commonly accepted method uses micro phones fairly close in, but not too close, so as to use a studio of more or less normal reverberation characteristics. The apparent reverberation on the ultimate recording is governed by the spacing between the microphone and the sound source--the musical instruments. A position closer in will reduce the effective reverberation by increasing the relative sound from the sound source. A position further away will increase the apparent reverberation.
For this reason several studios are finding it preferable to utilize microphones with some directional characteristics, either of the condenser type or one of the cardioid variety. This enables fewer microphones to be used further away from the performers, while still getting an acceptable relationship between the original sound and the apparent reverberation in the final recording. The overall result achieved this way tends to be more realistic, although it may offend the academic susceptibilities of some people who feel that omnidirectional pressure-type microphones only, should be used.
No professional studio recordings are made using the binaural microphone technique with two separate microphones placed in a dummy human head (Fig. 4). This is intended for home binaural recording and similar applications, rather than for serious professional use.
Fig. 43. The use of a relatively sharp filter to cut off frequency ...
above a certain point (6000 cycles shown here) produce, greater interference
with frequencies below this point thon a more gradual filter: (A) compares
the measured response; (B) shows corresponding delay time... Curve 1 indicates
a sharp filter; 2 indicates … a more gradual one.
Stereophonic recording is still a new field. Talking to recording engineers from the various studios, one gets the impression that as yet, a great many poor tries have been "buried" for each one that has proved satisfactory enough to release.
In discussing stereophonic theory with different recording engineers, one meets some conflicting opinions as to the nature of stereophonic sound, and also the pertinent factors for satisfactory binaural perception. Many independent investigations have been conducted for example, into the effect of frequency range, or the effect of different sections of frequency range, on binaural perception of the location of individual instruments, both "laterally" and in "depth", from the reproduction of two loudspeakers.
Conclusions from these varying sets of experiments seem to disagree, not because people with different kinds of ears were listening, but because the explanation of the results has been somewhat faulty.
Unfortunately, many are still hypothesizing on the basis of steady tone composition--discussing intensity and phase differences. This can lead to some faulty conclusions.
For example, to determine whether frequencies above, say 6000 cycles, are necessary, a low-pass filter which allows all frequencies up to 6000 cycles to be amplified, but removes frequencies above this point virtually completely, is used to compare the effect of recordings "with or without" the top range, from 6000 cycles upwards.
On such a test it seems that these high frequencies are vital to proper stereophonic perspective. However, a factor that seems to have been overlooked in this particular experiment, is the affect of such low-pass filters upon the transient response of the system.
It has been proved, both theoretically and experimentally, that any filter with a sharp cut-off that suddenly ceases to amplify above a certain frequency, also produces time differences between the frequencies well below the cut-off point.
This means that frequencies in the region of from 1000 cycles up will have the time relationship of transients modified by the use of such filters, although the response magnitude-wise, may be quite level (all the frequencies are reproduced in proportionate amplitude). But correct binaural perception is dependent upon correct timing of all the frequencies heard, as well as correct magnitude.
In fact, other things being equal, it is more dependent upon correct timing. Notice here that we are concerned, not with the correct phase relationship between the two channels, but the correct timing of the individual frequency components of the waves from each channel. As the sharp cut-off filter upsets the timing of frequencies from 1000 cycles up, it is understandable that addition of these filters would considerably mar the stereophonic illusion.
Had a filter been used which did not cut off so sharply, so as to destroy the timing of frequencies below the cut-off point, quite different conclusions would have been reached (Fig. 43) . Experiments conducted in this way show that the high frequencies are less necessary to stereophonic illusion than the mid-range, up to somewhere between 6000 and 10,000 cycles, as has been discussed elsewhere in this guide.
This once again underscores the importance of thinking about stereophonic effects in terms of sound impact rather than frequency response.
Another question connected with record-making is the kind of medium to be used for final release of the recorded program, whether to use tape or disc. So far, of course, the only available disc method uses two separate tracks spaced apart radially. How ever, the idea of using both vertical and lateral movement of the same stylus has intrigued a number of people for some time and experiments are going on with a view to developing this system.
The fact that magnetic tape does not use any moving parts, and thus avoids the mechanical complications bound to be involved in any two-way pickup, makes most people in the industry confident that magnetic tape is really the ultimate solution. The big appeal for disc, however, is its extreme versatility.
On the other side of the score, versatility is being improved for magnetic tape by the development of tape cartridges. These will save the necessity of threading the tape through the machine, which at the moment is a job for only one member of the average house hold--the owner of the tape-recorder. Almost anyone however, can put on a disc.
Years of work have already gone into the production of satisfactory disc techniques, whereby discs can be pressed quite rapidly.
On the other hand, pre-recorded tape production is a relatively recent development. Even the quality of magnetic oxide on the tape has only recently developed to its present standard, and it is highly probable that it will go on improving for some time to come.
This may result in tape ultimately being a medium of much better quality than discs could ever be, especially in view of the fact that disc depends upon a mechanical stylus vibration for the reproduction.
The problem in tape duplication is that of speed. There is no counterpart to the pressing method, whereby discs are produced, stamping them out in rapid succession. However, multiple duplicators have been designed capable of running the tape through a number of heads simultaneously at many times the actual playing speed used for reproduction. This enables the transfer to be effected much more quickly, and a great many tapes to be printed at the same time (Fig. 44). Further development in this direction may yet yield a tape duplicating process that competes favorably as to ultimate cost to the consumer, with the pressing method for discs.
FIG. 44. A modern multiple tape-copying machine. With high precision built
into it, and a much higher bias frequency, this machine can copy at several
times the normal playing speed, without losing original quality.
(Adapted from: Stereophonic Sound (1957) by Norman H. Crowhurst)
= = = =