Wildlife Recording in Stereo Part 3
by David J Tombs.
This article appeared in Wildlife Sound March 1973 ff
Parts 1 and 2 were, in the main, about microphones and because I deem them to be our number one concern, I shall, in this part, continue of the same subject.
Perhaps the greatest of our problems when placing microphones in the field is that of unwanted noises, i.e.
- Man made noise on land, water and in the air
- Electrical noise, hum, hiss, radio breakthrough etc
Taking electrical noises first, let us consider the causes of hum. Hum is most commonly due to one or more of the following conditions. Long unbalanced and unscreened cables, high impedance microphones and/or unbalanced transformerless amplifier inputs. Normally one should altogether avoid high impedance microphones (25000 ohms) for field work, especially with cable lengths of more than 3 m, if only to preserve the higher frequencies. To minimise hum, radio breakthrough and various forms of static induction, the correct type of cable has to be used, i.e. microphone cable, a balanced twin conductor cable with an outer screen. The screen should be directly connected to the chassis of the microphone amplifier/recorder. The body of the microphone, if metal, should also be connected to the cable screen.
The microphone should be low/medium impedance types not more than 200 ohms, or line matching transformers may be used at the the microphone end of the cable, and the microphone amplifier should have a balanced or transformer input.
Adhering to these conditions, cable runs exceeding 200 m may be used with confidence. The magnitude of radio frequency pickup presented to the amplifier will also be minimised although radio breakthrough is due, in tem ain, to the use of wide bandwidth amplifiers, in the design of which little attention has been given to the suppression of radio frequency signals. With the amplifier whose circuit is illustrated, radio breakthrough has not been experienced even when operated a short distance from radio transmitting aerials. However, different makes of microphone cable vary greatly in the respect to radio frequency pick-up.
The problem of hiss is more complex. Noise in the form of hiss is due to he thermal movement of electrons in the circuit resistances of the amplifiers and them ore one amplifies, the higher the hiss level. In other words, if one starts with an amplifier of good design in which the self generated noise is very low, the main source of noise generation is the very first resistance in the circuit, the microphone. As well as producing the signals desired, the microphone then also generates unwanted noise.
In the low noise amplifier circuit shown, which as a voltage gain of 85dB, let us now see how the noise is affected by input resistance. It was found that when the amplifier input was terminated with 200 ohms, a peak signal-to-noise ratio of 52dB was achieved, and when terminated with 47 ohms, 57dB was achieved, almost half the noise level and about as good as one can get.
The lower the impedance of the microphone, then, the lower the hiss. But, the lower the impedance of the microphone, the lower is the output signal from it.
Now, suppose two microphones of the same impedance are connected in parallel and their output signals are in phase – the total signal level will be slightly increased and the source resistance reduced. In my experience this method of parallel connection achieves about the best signal to noise ratio possible at the present. It is important, of course, to use sensitive microphones and the parallel system, which I use, incorporates moving-coil (dynamic) microphones having a sensitivity of 0.25mV/microbar at 200 ohms. The use of microphones of less sensitivity may lead to disappointing results.
Care must be exercised when it comes to condenser microphones which, having an amplifier as an integral part, give a constant self-generated noise level. Some manufacturers quote a signal to noise ratio for condenser microphones, usually the ratio is in respect of a signal level of 94deciels, which would correspond to the sound level produced in a concert hall by an orchestra playing loudly.
In the field we would be lucky to receive a sound level of 50 decibels, about one hundredth of the level in the concert hall, therefore a condenser microphone with a quoted noise of 060dB would, on a signal level of 50 decibels SPL have a self noise separation of 60-(94-50) = 16dB, a voltage ratio of only 8 to 1 approx. used in a reflector, where the noiseless gain could be up to 20dB (ten times) at the frequencies which relate to hiss, a microphone with this quoted figure would give an 80 to 1 (38dB)signal-noise ratio on out 50dB SPL level, which would be acceptable
The Sennheiser gun microphones MKH805/815 are condenser microphones but of a special type. The signals received are used to modulate and inbuilt radio frequency oscillator, amplification is performed at radio frequency the demodulated. The resulting audio signal output is high although from a low impedance source and the self-generated noise very low. The overall signal to noise ratio, when used with the amplifier illustrated, was found to be almost as good as the paralleled dynamic system.
Let us now consider wind noise. Wind is a mass of air in motion, which may cause objects in its path to move,thus producing sound. It also causes wind blasting on our microphone. There is little one can do about the noise caused by wind blowing thorough the trees except wait until the wind drops. There is, however, a lot we can do to prevent the wind blowing into our microphones. Perhaps the simplest form of wind shield is the plastic foam slip on type and this gives adequate protection against a light breeze. A wind break, similar tot the type used by many when relaxing on the beach, may also be used to advantage. The size will need to be about 1.5m long by 1m high to be really effective. As the wind break will often be placed between the microphone and the subject, the fabric should be acoustically transparent. Certain fabrics for loudspeaker cabinets are suitable for wind breaks and to prevent the cloth from flapping in a strong wind it should be fixed tightly stretched on a four sides frame, which can be collapsible for easy transportation. Such a screen will be very effective when erected at about twenty degrees from the vertical and sloping with the wind. The microphone should be placed in the zone of shelter and about half a meter back from the screen.
Another type of wind shield is in the form of a drum or closed end cylinder. The diagram shows the basic details of one of my own designs. The important points to observe are
- That the fabric used to cover the two drums or cylinders is of a special type of loudspeaker fabric, which I believe is known as Tygan. This is a plastic cloth woven from fibres shown to have a flattened cross – section.
- That the space between the inner an outer cylinders is at least 1cm
- That the seal or fit of the cylinders to the top and bottom support plates is good and that there are no pints at which the wind can gain direct entry to the central microphone section other than through the fine fabric. The larger the diameter of the inner drum the better, if your microphones are large, as it is the space between the working end of the microphone and the fabric which improves the resistance against wind noise.
I have used this type of wind shield in a force 8 gale with a condenser cardioid type microphone normally very prone to wind noise and got results free form wind blasting, although I had better mention that a reasonable filter was also used at the time.
A filter which reduces the intensity of low frequency rumble can be used in combination with wind shield to improve still further the resistance against wind noise. In the amplifier circuit shown, the second stage incorporates a steep cut filter which attenuates only the frequencies below 125Hz. The slope or rate of the attenuation is approximately 24dB per octave, so a 50Hz signal is reduced to approx 1/20 and a 25Hz signal to approx 1.200 of the level obtained without the filter.
The frequency at which point the cut begins may be shifted up an octave by reducing by half the values of the capacitors marked F. A multiway switch can be incorporated to perform this function, thus obtaining several positions of bass cut. A filter in no way reduces the need for a good wind shield but does improve the overall results and is very effective in reducing distant traffic rumble. Care must be exercised, however, not to cut more bass than necessary, as this may result in recordings lacking in natural atmosphere.
A point also worth mentioning is that a filter is more effective in the early stages of an amplifier. If allowed through to the recording machine large amplitude rumble signals may result in severe intermodulation distortion on the tape.
Looking at the problem of man made noise; each year sees a further aggravation of the situation. We now see electric and petrol driven grass cutters, hedge trimmers, vacuum leaf-sweepers, rotary garden hoes and many other noisy items of apparatus used to do small jobs on hedges and lawns
of a few yards which, by more traditional methods could be performed maybe in less time and almost silently. Increasingly, the clockwork model boat is replaced by the radio controlled diesel model, and the rowing boat by the speed boat just as the tractor has replaced the horse and the pneumatic drill the pick and shovel – where will it all end – as leisure time increases I can only see an increase in man-made noise.
There are many ways to overcome some of the noises such as –
- Getting up earlier than the people who make the noise
This can be very rewarding, particularly on a Sunday morning but more and more people seem to be catching on. We now see the all night traveller and 24 hour corn harvesting and to tell us how well we are moving, even the AA has taken to the air.
- Bribes to tractor drivers and chain saw men. Consider for a moment that the person carrying the tape recorder is the odd one out, and another odd fellow with a tape recorder will be showing up some time later, so spare a thought for him. It pays, therefore, to extend the utmost courtesy and understanding towards the others. Some operatives are only too willing to take an early lunch break for the cost of a few pints.
- The use of directional microphones
Highly directional microphones can indeed help us a great deal in achieving good results, which is one of the reasons why a gun microphone is high on my list of desirables. It is most unfortunate for us that the reflector is not very directional at low frequencies.
- Placing the microphone close to the subject
We have here probably the most satisfactory way of achieving good results, not only free from unwanted noise but also towards good stereo,
The time honoured method of placing microphones close to a subject seems to be undesirable and involves either watching for hours or days for a good spot to place the microphones, or directly frightening away the subjects in order to position the microphones strategically in the hope that the subject returns. The chances are that they will return, sometime, to sit behind or to the side of the microphone. This will not help us to obtain good results because for stereo we require the microphones to face the subject. We also want a nicely balanced background.
Having had this happen a few times I decided to do something about it and now have a remotely controlled unit for altering the direction of the microphone. It consists of a reversible motor and gearbox with an output speed of about 1 full turn per minute, which is slow enough to cause little concern to he subjects. Power to them motor unit is supplied via the screens of the microphone cables.
Another device, which may go a long way towards achieving good results in unfavourable conditions, is also still in the drawing board stage. The device will be a vehicle capable of transporting a car battery (to power it) and a few microphones over rough terrain, through mud and long grasses and on water. The vehicle, which could be collapsible, would be controlled remotely from the microphone cables, which it pays out as it progresses slowly towards the subjects. Mammals and birds seem to take little notice of slowly moving objects. The direction of the microphone will also be controlled remotely. The device will be halted at various distances from the subjects to obtain different sound perspectives. It might even be possible to record very close perspective with such a device. There is little point in considering radio control to such a vehicle, because the microphone cable will provide all the necessary controls. To introduce FM stereo transmitters and receivers and to do away with all cables would, I think, be taking things a fraction too far.
Carrying a tape recorder, it is possible for one to travel slowly in an open bottom camouflaged cardboard box to get closer to your subjects. The results might be accompanied by your heavy breathing! It would, of course, be possible when close enough to set up the microphones outside the back of the box through a flap, then side step away a short distance. This would avoid your breathing noises but on salt flats it might also entail some digging out by some others in a bid to rescue you from a slow and painful death.
I will not attempt to suggest or recommend a tape recorder. I have used a Nagra and a Uher but no other portable machine.
I performed a test to compare the level of recorded tape hiss among a variety of machines. I found that all the machines gave the same result and that it was the type of tape used which made the greatest difference. If you wish to get the lowest hiss level from your machine use low noise tape, then fully modulate it. Care should be taken not to overmodulate the tape as overmodulation means distortion.
The type of level meter will have a significant bearing on tape modulation levels. A fast reading meter which registers, or nearly registers, the true peak level of the signal, will be found satisfactory. These are known as peak programme meters (PPM)
Another type of level meter, more common is domestic and semi-professional equipment is the VU meter which responds to the average or RMS value of a signal. Although the VU meter may be satisfactory for recording organ music, it is not so suitable for recording bird song, many calls of which have high but short duration peaks. A VU meter may try to indicate the average level, by which time the true peak level of the signal will be running into severe peak distortion on the tape.
Many recording machines incorporate "automatic level control". This may be found useful so long as the control is allowed to operate only on the very loudest modulation peaks, which might otherwise run into distortion. To use automatic level control on the whole signal would give most unsatisfactory results, the background sounds jumping up and down in volume.
In the amplifier circuit shown, there is a means by which a treble boost of 4dB at 10kHz can be applied. When in use on a good signal level, the recording will, of course, be a little too ‘bright’ due to the HF lift. However, if the recordings are then played beck with the correct amount of treble cut, the noise level can be very low indeed.
Tape faults such as drop pouts show up very much more in stereo and the use of half track machines will results in less trouble from dropouts that quarter track machines.
Tape speed also has a large bearing on performance, though tape which at a speed of 38cm/s gives no problems, may at 4.75cm/s give many dropouts and a high hiss level.
Cassette machines for fieldwork will no doubt be popular. At the moment I know of three battery-operated stereo cassette machines but the performance of reel-to-reel machines is far superior. Editing on reel to reel machines is also much easier. It seems to me that the reel to reel machine is by far the better value. One manufacturer quotes the performance of his cassette machine using chromium dioxide tape as 30Hz to 13kHz. At these extremes the response if –15dB while the same firm's reel to reel machines response is 40Hz to 20kHz +/- 3dB at 19cm/s.
The future may be with cassette machines and it may be that one day we shall be able to use a superior type of cassette, at present only being tired experimentally for some broadcasting purposes and running at 9.5 cm/s instead of the domestic 4.75 cm/s cassettes now available.
Whichever type of machine one gets, the results can only be as good as the signals sent to it by the microphones.