7  Stereo

Part 2

Microphone techniques for stereo

Here we can look a little more deeply into microphone usage for stereo. Various techniques exist, but here we shall look at three, and only briefly at two of those. The reader who wishes to go further into the matter is referred to the list of books headed ‘Further reading’.

The first technique, and in many ways the simpler to understand, is that of multi-miking – making use of several microphones whose outputs are ‘positioned’ in the stereo image with panpots. From the point of view of the non-professional user, the difficulty with this approach may be not having a sufficient number of inputs and associated circuitry on the mixing equipment. Also, of course, a large number of microphones may be required – in the opinion of some professional sound balancers the drum kit alone may need six to ten or even more!

The main advantage of the multi-microphone approach is that there can be good control over the sound balance – that is, weak sections of, say, a band or choir can be raised in level to match louder sections.

The disadvantages are:

1.   Technical complexity can be considerable and, paradoxically, can sometimes make it more difficult to achieve good results. There is also more to go wrong!

2.   It is possible for microphones to form unwanted stereo pairs. This means that in addition to the panpotted images there may be spurious additional images caused by more than one microphone picking up the same sound. In practice, this is unlikely to mean that there will be multiple images; rather the panpotted images will be shifted from their intended positions. As a further hazard, moving a fader to alter any one microphone's output may result in some of the shifted images moving. The only way to avoid this effect is to have all the instruments, band sections or whatever ‘closely miked’, i.e. have the microphones as close as possible to their respective sources. This can sometimes be difficult, especially as it is not always easy to retain the characteristic sound of an instrument with closely positioned microphones.

In short it's fair to say that multi-microphone techniques are generally best not attempted unless there are many microphones available and the mixing equipment is suitably comprehensive.

The second microphone approach is to use coincident pairs, which we have already referred to. Here a great deal depends on the polar diagrams of the two microphones. The assumption must be made that the diaphragms of the two microphones are as close together as possible, although it must be admitted that satisfactory results can be obtained when they are several centimetres apart (see ‘Binaural stereo’ on p. 101).

The main effect of the polar diagrams is on the angle of acceptance of the microphone pair. By ‘angle of acceptance’ we mean the angle in front of the microphones, which corresponds to a sound image stretching the full distance between the loudspeakers. Photographers will realize that the angle of view of a lens is a very similar concept.

Knowledge of the angles of acceptance for different polar diagram systems is essential if satisfactory results are to be obtained. However, it isn't difficult to carry this information in the head when it is remembered that hypercardioid microphones have polar patterns which are intermediate between figures-of-eight and cardioids. Then, for microphones which are at 90° to each other, the angles of acceptance are approximately:

Figure 7.5 ‘Angle of acceptance’ for a coincident pair

Notice that we are assuming the microphones are at 90° to each other. This is fairly standard, and a few degrees difference on either side of 90° is unlikely to have any significant effect, although the angle can be increased considerably with cardioids, resulting in a wider stereo image.

Perhaps the big question now is, what polar diagram system to use when? Often, of course, there is little choice. Commonly the only microphones available are a pair of cardioids (as, for instance, in the built-in stereo microphones of most camcorders). However, the positioning of a pair of coincident microphones, of whatever polar pattern, is largely a matter of geometry.

Let's assume we have two cardioids – and we repeat that they should be as closely similar as possible, same manufacturer, same type number and so on. Figure 7.6 shows the probable image spreads for two positions for such a pair of microphones. In (a), the microphones would be close to the front of the sound source (band, choir, orchestra, etc.). In (b), the microphones are further back so that the axes of the cardioids are pointing approximately at the edges of the source. Remember that the angle of acceptance for a pair of cardioids is about 180°. In (a), the spread of sound images will fill the region between the loudspeakers almost completely; in (b), only a part of the region will be filled. Note that the state of affairs in Figure 7.6(b) is not necessarily wrong. A solo piano will usually sound a bit silly if it appears to be the full width between the loudspeakers! Equally, a classical string quartet should normally fill only part of the picture.

Figure 7.6 Likely image spreads for a pair of cardioids

With a rock band it is difficult to say what the reproduced image should be as there is often no attempt to reproduce the reality of the original sound.

In either case, though, reverberant sound will tend to fill the stereo image, thus adding to the realism, even if the basic image of the instrument(s) is narrow. It all depends on what will seem reasonable to the listener on the replay.

A pair of figure-of-eight microphones has rather different characteristics. To begin with, the angle of acceptance is smaller than for cardioids (90°) and for full loudspeaker-to-loudspeaker coverage the microphones should be further back – about where the two cardioids are in Figure 7.6(b).

Secondly, there is a rear pick-up region of 90°, which may or may not be a good thing. For example, it can be useful if plenty of reverberation is wanted; on the other hand, there may be unwanted noises from the rear – a coughing audience, for instance! A final and sometimes very important point is that the side quadrants are out-of-phase regions and this means that sounds arriving from these directions may be difficult for the listener to locate. The sides therefore have to be regarded as ‘forbidden zones’, although reverberation which arrives from the sides doesn't suffer any peculiar effects because it is, by its nature, so diffuse that phase has no real meaning. This is illustrated in Figure 7.7.

The point has been made that, other things being equal, figure-of-eight microphone pairs need to be further back than cardioids. There are some pros and cons as a result:

1.   The further back the microphones, the better (probably) will be the relative balance between the front and back of an orchestra or band. Put another way, a close pair of microphones such as cardioids may result in the nearest instruments or voices being too loud in relation to the further ones. The relative sound perspectives (see Chapter 5) will be exaggerated.

2.   If the acoustics of the hall or other recording area are poor, for example with excessive reverberation or external noise, then closer microphones like cardioids may help.

3.   The closer the microphones are to the band, orchestra or choir, the more important it may be to have additional mono microphones to reinforce the more distant sections. Then panpots have to be used to put the outputs of these microphones into their correct positions in the ‘picture’.

Figure 7.7 Figure-of-eight microphones as a coincident pair

It may be worth mentioning that, given a good band, orchestra or choir in favourable acoustics, a pair of figure-of-eight microphones can, in the opinion of many experienced sound balancers, give the most natural reproduction. It is also the simplest arrangement technically speaking.

A hypercardioid pair has not been mentioned but the reader will quickly be able to adapt what has been said about cardioids and figures-of-eight to hypercardioids, remembering that they can be regarded as being half way between the other two.

A third microphone technique is that of binaural stereo. It is intended primarily for headphone listening, but carefully done gives results which are acceptable on loudspeakers. The idea is that the ears of a person are replaced by microphones, or at least microphones are placed at, or even in, the ears. The theory (rather incorrect!) is that if those microphones are connected to headphones the listener is ‘transported’ to the position of the wearer of the microphones.

Special dummy heads have been designed and marketed commercially, but much simpler and cheaper is to have a disc of perspex, some 6–8 mm thick and 200–250 mm diameter. Two small microphones are mounted on opposite sides of the disc, on its axis and roughly 100 mm from it. These form the stereo pair. (Note that the two microphones have a spacing which is approximately that of human ears.) Interesting results can be obtained, although most headphone listeners find that the sound images tend to be behind them. There is scope for experimentation by the reader with two small microphones and modest workshop facilities.

Headphones for stereo monitoring

Having mentioned the use of headphones, it may be worth adding a comment about their use in stereo.

To begin with, serious monitoring of stereo is almost impossible with headphones, because the positions of the sound images in general bear almost no relation to those perceived when listening to the same stereo material on loudspeakers. Some sounds can appear to be behind or even inside the head. I once carried out some experiments to see whether there was any correlation between the positions of ‘headphone images’ and ‘loudspeaker images’, but even experienced listeners differed so much among themselves in their location of headphone images that the final conclusion was there was no significant correlation at all.

Lastly, in setting out to do a stereo recording, it is worth studying the situation in advance to form some idea of what microphones to use and where to place them. Then, and this is most important, a rehearsal should be used to test as thoroughly as possible whether the optimum microphone positions have been achieved. This means listening critically, preferably on loudspeakers. If this cannot be done in the recording venue, then it means making a recording and playing it back in the best possible listening conditions.

All this is often easier said than done, especially as there are complications like the fact that the presence of an audience will almost certainly modify the acoustics, and the musicians will play louder and better when there is someone to play to! However, with experience it is possible to make some allowance for these things.

Questions

With a coincident stereo pair of microphones angled at 90° to each other, what are the effective ‘angles of acceptance’ for the following?

1.   Cardioids

2.   Hypercardioids

3.   Figure-of-eight

4.   Omnis

5.   With what coincident pair arrangement is there likely to be a significant out-of-phase region at the sides?