|
What happens to the sound before the transmitter and after the receiver.
Reverberation in studios and listening rooms plays such an important part in modifying the sound which eventually reaches the ear through the medium of broadcasting that investigation of these effects should take precedence over further efforts to obtain 'straight-line' characteristics in transmitting and receiving apparatus.
One of the two chief objects of broadcast listeners is to hear the programmes as well as possible (the other object being to secure the choice of as many of them as possible; but just now we are not going to argue about that). So there is much discussion about the faults and merits of various types of receiver, loud speaker, and so on, and how to avoid distortion. And thus We tend almost to assume that if our apparatus were perfect our listening would be perfect apart from the substance of the programmes themselves, of course.
In the early days of broadcasting the imperfections of the equipment were so gross that there was good reason for dealing with them. This has been done so well that now it is possible to make a very close approach to perfection. One could build - at a price - a receiver that treats all audible frequencies of any importance equally, within close limits, and which is practically free from harmonic distortion. And yet.
Fig. 1. - Curve B, with a 10% peak, looks very much worse than A. But to the ear the imperfection is negligible; very much larger departures from a straight-line characteristic are inevitable in sound transmission, but are seldom taken into account, owing to absence of 'curves'.
Having become accustomed to broadcast listening, on a real 'quality' receiver, maybe, go and hear a programme in the concert hall or studio. Then, if you have any critical faculty at all, you will come away wondering what it is that makes such a difference. Radio experts may know all about straight-line amplification and linear detection, but everything outside the apparatus itself they are apt to term vaguely acoustics. It is gradually being realised that attempts to smooth out tiny hollows in the amplification curve are like straining at a gnat, when all the time many camels are unconsciously being assimilated. The amplification curve A in Fig. 1 looks much better than B, with a 10% peak. Yet acoustic effects, seldom considered by the technical enthusiast, commonly introduce peaks of hundreds per cent. I And what about the ear, which itself produces almost every known sort of distortion?
It seems quite clear that as the ear is common to all listening, Whether direct or mechanical, it can be left out of account and attention concentrated on producing an effect at the threshold of the ear which is as nearly as possible the same as that produced there by presence at the original performance. As it happens, however, the behaviour of the ear cannot be entirely disregarded.
The reason for this is obvious if one takes an example, say, the performance of a full orchestra or military band. This is meant to be heard in a suitable auditorium, or, perhaps, out of doors. If by the working of a miracle all the performers and their instruments could be accommodated in one's suburban sitting- room and still leave room for the family to listen, the result would not be pleasing. The room being probably about one-hundredth part of the size of a public hall, the intensity of sound might be about a hundred times as great, and the listening ear could not fail to notice it. This would, emphatically, not be a perfect re-presentation of the original.
Very well, then; the obvious thing to do is to install a loud-speaker system that reproduces everything exactly like the original, with the one exception of volume. In other words, do for the sound what a photographer does to a picture when he reduces it. Everything remains in perfect proportion, true to the original, but on a reduced scale; This surely, would give exactly the same effect in a small room as the louder original heard at a greater distance and in a bigger space?
The answer is-no. The process of hearing is much too subtle and complicated for that. And so are the acoustic effects that exist at both transmitting and receiving ends. The proud owner, in a burst of enthusiasm, may declare that it is just as if the orchestra were in the room; but he knows perfectly well that it is not - if he has ever heard a real orchestra.
Things that can be measured on a foot-rule, or even a voltmeter, are easy to study, and to arrange just as we want them; and one expert can get others to agree with him as to the results of his work. But things that depend on personal impressions usually lead to nothing more enlightening than endless arguments. So the acoustical and physiological aspects of broadcasting seem to be less definite than the electrical. The distribution of sound by films and radio is becoming so important, however, that these things are now being reduced to something more like an exact science.
What is a 'Good Transmission'
A few of the problems that are left over after everything has been done to the receiver-and transmitter will now be considered; some of them refer to the receiving end, and may help the listener to improve his reproduction. Others refer to the transmitter, and are beyond his control. As they are being actively looked after by competent people, perhaps that is just as well. But a certain degree of cooperation would be of advantage, and it is to be regretted that the BBC does not take us more into its confidence in a practical sort of way by explaining over the microphone what it is doing. Mere written publication is no substitute for actual demonstration.
Even an unobservant listener has probably noticed that a particular sort of programme can sound differently in a way that he finds difficult to describe, and which cannot be put down solely to the receiver. A good orchestra may sound dull and lifeless from a certain building. A speaker is surprised to find that a comparatively small auditorium swallows up his voice, and is more trying to speak in than a larger one. A violinist, perhaps, who broadcasts from one particular place achieves a reputation which he finds difficult to sustain when he moves elsewhere.
Fig. 2. - Some of the paths taken by the sound from a speaker (S) to a listener (L) in a rocky cavern. In addition to the sound directly received, a much larger volume comes as reflections, but as these arrive later the speech is confused.
When listening, part of the sound that is heard comes direct from the thing that is making it and part is reflected from the surroundings. The latter has a longer journey, and so it smites the ear a short time after the original. Take as an extreme example an enormous empty cavern, with smooth rocky walls everywhere, and someone a hundred yards away trying to talk to you (Fig. 2). The sound of his voice is reflected from the walls in all directions and comparatively little comes direct. So any particular syllable is still arriving when the speaker has passed on to the next, and causes confusion. It is quite easy for him to give you a large volume of sound, but it is not clear. It corresponds exactly with a badly focused photograph in which the light from any particular part overlaps other parts.
Controlling Reflections
The acoustics of a cave are hopeless for speech, but a broad, slow organ piece might sound very impressive. The detail would not be sufficiently fine to be obscured by the echoing, While the augmented volume would be grand.
For speaking purposes the reflection of sound might be reduced by lining the cave with felt. Then your friend would have to speak much more loudly. The speech would still be difficult to follow, but for another reason. It would sound 'woolly', without life or crispness. Felt absorbs the high voice frequencies that convey the intelligibility and clearness, and leaves the lower-pitched sounds that merely make a boomy noise.
One could go a stage farther and provide a more elaborate lining to absorb sounds of all frequencies. The speaker would then have to shout to be heard at all, it would be just like talking in the open air, but his voice would be heard clearly.
It may be human perversity, but it is an established fact that things sound more pleasing with a certain amount of echo or reverberation, to use the correct term. Very little is enough for speech, and slow music needs most. Probably the effect of distance, associated with spaciousness and grandeur, underlies the preference.
The ordinary small living room naturally does not impart this sense of spaciousness to any sounds coming out of the receiver; consequently, if it is to exist at all it must be put in at the transmitting end. Outside broadcasts from concert halls are, of course, coloured by the acoustics of the buildings, which are generally reasonably good. In the early days of broadcasting the studios were very small. It was found that the reverberation in a small room is bad. Although it may be possible, by having bare hard walls, to ensure a considerable amount of echo, the ear is sensitive to the time element even although in the largest hall the sound takes only a fraction of a second to travel from end to end. You can very easily test that by being led blindfolded around an unfamiliar building. Not only is it instantly apparent whether the room is bare or furnished (judging from the amount of reflection or absorption respectively), but the approximate size of the room can be estimated provided that absorption is not so complete as to allow little reflected sound as a clue.
Or when one is holding the line during a telephone call and there is a background of sound from the distant end, one can judge of the sort of room in which the instrument is placed.
There is another room characteristic besides the amount and the distance of reverberation; the room itself resonates to one or more frequencies and over-emphasises them. Next time you are in a bath-room, or other scantily furnished apartment, try humming a continuous note, beginning as low down as you can manage and gradually rising in pitch. It is more than likely that you will find a note to which the room seems to respond, and at which there would be a pronounced peak in the frequency characteristic if it were used for broadcasting. Any general noise in the room tends to cause a sound of this pitch, and unless the room is very large the effect is generally undesirable.
One of the small and heavily draped studios used in the early days of broadcasting.
So the studios were thoroughly padded with sound-absorbing materials to remove the 'boxy' effect that disclosed to the listener their smallness. The resulting absence of reverberation was unpleasant, both for artist and listener, so the next stage was to introduce synthetic reverberation by means of an echo room, entirely bare save for loud speaker and microphone. The programme was reproduced from the loud speaker, and picked up, plus echo, by the microphone. The result could be mixed in any desired proportion with the echoless programme taken direct from the studio.
Sound Absorbing Materials
Fig. 3. - Curves showing the degree to which various surfaces absorb sounds of different frequencies. A characteristic such as B would be unsuitable for lining a studio, as it would weaken the high tones in relation to the low. Something more like A is now generally favoured.
For reasons which have already been given, this fake is not equivalent to the desired acoustics of a large hall. So the present practice of the BBC is actually to use a studio of a size appropriate to the type of programme, and to aim at a desired amount of reverberation by means of carefully selected materials. This is not merely a matter of using a material with the right percentage absorption; the behaviour of any material depends very largely on frequency, as we observed in our felt-lined cavern. Certain sorts of building board are fairly uniform. Fig. 3 gives a number of examples, which show very clearly how the frequency characteristic of broadcast transmission and reception, too can be utterly altered by acoustic effects alone.
It is really remarkable what a change in the acoustics can be produced by a relatively small alteration in the surface of a room. Everybody knows that it is quite different walking into a previously empty room when a few articles of furniture have been pushed into it. Some foreign studios are provided with means for changing the exposed material on the walls; and it has even been proposed to make use of a device something like a gigantic roller towel for varying the acoustics during the performance of a musical item to get the maximum effect (Fig. 4).
Fig. 4. - Device fitted in wall recesses for obtaining rapid variation of room echo. The rollers are operated by motors, and it is proposed that they should be controlled by the musical conductor.
Microphone Position
The BBC, however, prefer to use a special studio for each job. The position of the microphone obviously has a lot to do with. the matter. If it is very close to the source of sound, the reflected sound is negligible in comparison; and no reverberation is apparent. So when announcements have to be made during a programme in a large studio there is not need for the announcer to speak from another place; all he has to do is to murmur gently into the ear of the microphone. As a matter of fact, it is possible to provide the illusion of a speaker walking away by gradually fading from a nearby microphone to a more distant one. The increase of echo gives the effect.
Have you ever happened to have a loud speaker bawling loudly in a distant part of the house or someone else's house and then turned on another near you doing the same. programme very faintly? If so, you were probably surprised to get an effect like that of walking into a lofty hall. Although the local speaker seems to contribute negligible sound, it reproduces artificially the conditions of direct and reflected sound in an 'echoey' room. Another subtle shortcoming of loud speaker reproduction is that the benefit conferred by Nature of having two ears is lost. It might be supposed that the second ear is just a. spare in case one gets damaged; but it does much more than that. It hears a sound at little earlier or later than the other (unless the sound is equidistant from both), and from this the brain is able to judge the direction of the sound. Precisely the same principle is employed in certain submarine-locating devices used on ships with two microphones spaced apart. Binaural hearing, as it is called, can be provided but; as it necessitates duplicating the entire broadcasting system, from microphone to receiver, and applying the separate outputs each to a telephone ear-piece, the system is not likely to be generally adopted, even although the results are very remarkable. In a test carried out, a whole audience of people was made to jump round suddenly when a voice appeared to shout into one ear. The speaker was actually in the studio. This form of amusement is exactly analogous to the illusion sometimes provided by supplying an audience with special coloured glasses; people wearing them duck in alarm when missiles appear to be coming straight for them from the stage.
The effect of distribution of the players in an orchestra, or of movement of actors on a stage, can be conveyed only by binaural hearing. Another and rather unexpected benefit is that owing to the ability to concentrate on a sound coming from a particular direction, there is less distraction caused by other sounds and disturbances.
There is another directional effect that causes frequency distortion. Most microphones receive equally well from all directions at low frequencies, but as the frequency of the sound rises the reception is confined more and more to the direction straight in front. Microphones have now been developed which are either directional or non-directional as desired, but in either case treat all frequencies alike. The directional types are useful for special purposes; for example, in a film studio, Where the noise of the camera is notdesired; or on a stage, where the vapourings of political maniacs in the audience, or even coughs, form no part of the legitimate programme. This technique appears to be more deeply explored in America than here.
Directional Effects
At the receiving end it is quite easy to detect a noticeable difference in tone due to the directional effect of the loud speaker. At the side the tone is woollier, speech less distinct, and hiss or gramophone scratch less apparent than straight ahead of the cone. One reason why radio-gramophones sound less brilliant in tone than the corresponding table models of receivers is that the loud speakers are pointing at the legs of furniture, people, etc., which absorb the high notes considerably, and direct radiation at upward angles is comparatively weak. A particularly effective way of introducing top-note loss is by fitting the loud speaker in the base of the cabinet pointing towards the floor. For a good dance-band 'thud' it is, of course, superb. But announcements sound as it the speaker muttered mumblingly and low, as though his mouth were full of dough.
A contributory cause is one acoustic effect that has received quite a lot of popular publicity - cabinet resonance. If you put your hand partly over your ear in a sort of deaf-man attitude, you will notice that a certain moderately high note in music is over-accentuated lowing to the resonance of the hollow of the hand. The pitch of this note can be varied by moving the hand. The hollow formed by the cabinet, being larger, accentuates a much lower tone. The remedies are, in ascending degrees of effectiveness, using good solid timber for the cabinet in place of the plywood commonly employed, leaving as much as possible open to the air, lining the interior with suitable acoustic materials, or adopting a proper box baffle filled with slag wool or similar absorbent matter.
But the most troublesome problem standing in the way of producing a perfect miniature of a large volume of assorted sound is that the behaviour of the ear is hopelessly complicated with regard to the matter of loudness. It might seem that if the volume of sound in the air is doubled, it sounds twice as loud, and that is an end to the matter. There is room for a good deal of argument as to whether the loudness increases two fold, or by some other amount. But that is relatively unimportant. The real difficulty is, first, that by whatever amount the loudness appears to change at one frequency, it is different for another frequency. And, as if that did not complicate matters enough, the whole loudness-frequency characteristic alters as the loudness alters. When one goes abroad the coinage problem is exasperating enough as it is. But if the rate of exchange varied widely according to the particular coin used, and also according to the number of them involved in any transaction, things really would be difficult.
The outstanding features are that only moderately high frequencies (1,000-5,000 Hz) are audible at very low intensities of sound. As the intensity rises the lower tones appear, but, of course, they sound relatively weak. They increase more rapidly than the high notes, however, so that the very large difference that existed at low volume gradually becomes levelled out. At very great intensities, verging on the loudest that the ear can tolerate, the rate of increase for nearly all frequencies except the very highest is reduced.
Fig. 5. - Very simple type of tone-corrected volume control circuit.
Some receivers try to compensate for the relative weakness of low notes at low volume by means of a 'tone-compensated' volume control. The simplest method is illustrated in Fig. 5. When the volume control is set low, the high frequencies are weakened by the capacitor so that the low tones are given a reasonable look in.
There are plenty of other acoustic effects that help to thwart any attempt to get perfect reproduction by electrical means alone. We have only considered some of the more important ones. It is interesting to interpret the results one gets from broadcasting, in the light of what has been discovered about acoustics.
An example of modern studio design the Small Orchestra Studio at Konigsberg. The walls are movable and the area covered by draping is adjustable to give the required acoustic characteristics.
|
