A Case for Quality Microphones
With the development of their Series 4000 range of studio mics, Bruel & Kjaer have earned a reputation as one of the top manufacturers of versatile, high performance microphones. In this article Ralph Dunlop, B&K's UK Sales Manager, makes a case for the use of quality microphones and outlines a useful guide for determining microphone choice and application.
Until recently, Bruel & Kjaer was renowned mainly for its expertise in electronic measurement in the fields of acoustics and vibration. However, with the development of their Series 4000 range of studio mics, designed originally to respond to the new demands of digital recording, Bruel & Kjaer have earned a reputation as one of the top manufacturers of versatile, high performance microphones. In this article Ralph Dunlop, Bruel & Kjaer's UK Sales Manager, makes a case for the use of quality microphones and outlines a useful guide for determining microphone choice and application.
It is peculiar that in these days of complicated computer-based digital recording equipment, the microphone is still regarded with a degree of suspicion and fear. And although there is little mystery about a microphone - in fact, it is probably the most straightforward part of the audio chain - attitudes towards mics tend to be either too simplistic or subjectively overcomplicated, with the two rarely meeting in the middle. It is apparent, however, that in many cases a better knowledge of transducers and the way in which they can be used could produce significantly improved recordings.
If a useful division between microphones exists, it lies between dynamic and condenser mics. The superior quality of condenser mics is naturally more suited to recording and broadcast purposes, whereas the physically robust, directional pick-up characteristics of the dynamic microphone are ideal where live work is concerned. This article, however, will be solely related to condenser mics and is guided by the basic assumption that, although the microphone has evolved through various design stages from ribbon to moving coil to condenser, it is generally agreed that a mic's primary purpose is to transduce an audio wave into an electrical signal as accurately as possible.
Without doubt, the most popular type of microphone used for recording today is the cardioid condenser. The widespread use of the cardioid is in many ways a hangover from the early days of multitrack recording, when a high degree of separation between instruments, or the instrument and its environment, was required. Omnidirectional microphones, however, are in many ways more 'complete' transducers, as the following points illustrate.
- In terms of its design and construction, the omni is far simpler than the cardioid, since all that needs close examination is its frequency response. The cardioid, on the other hand, is characterised by a complicated porting design which almost certainly results in compromises regarding frequency response.
- Because of its simple design, the omni will probably require less servicing than the cardioid and rough treatment is less likely to affect its response characteristics, resulting in a longer working life.
- The absence of the rear porting system in an omnidirectional mic allows the capsule to be virtually sealed from the outside world, making it much more suitable for use in humid environments and extreme temperatures. One of the demonstrations used by Bruel & Kjaer is to plunge an omni into a glass of mineral water, while the mic is still connected to a PA system or tape recorder, without causing any damage to the microphone. It also makes for an interesting sample!
- The omni is less susceptible to handling noise and wind noise; two important factors to consider when recording outdoor samples. In addition, less problems are experienced with vocal 'popping' when an omni is used.
- In close-miking situations, omnis do not suffer from proximity effect. This is a characteristic that all cardioid mics exhibit to varying degrees, and results in an increase in the level of lower frequencies as a cardioid is moved closer to the sound source. When people describe a sound as 'warm', the chances are that a cardioid has been used so close to the sound source that there has been a massive boost at low frequency. Often the advantages of close-miking are completely negated by choosing the wrong type of mic, resulting in the introduction of extraneous noise via a very active mixer EQ section desperately trying to compensate for the proximity effect of cardioid mics. The 'electronic' noise can be unpleasantly apparent if a drum kit is being recorded by, say, eight closely-miked cardioids onto a digital recorder. Add the distorted sound spillage from one mic to another, and the signal on tape will probably bear very little relation to the actual sound of the drum kit. This effect is important to be aware of in applications such as sampling, when it is often vital to capture the true clarity of a sound rather than its distorted substitute.
- Closely linked to this last point is the omni mic's ability to handle much higher sound pressure levels (the Bruel & Kjaer 4004 omni handles up to 168dB before clipping) than the cardioid, without suffering from proximity effect. With cardioids, clipping occurs much more often as sound pressure levels (SPLs) exceed the upper limit of the microphone's dynamic range. See Figure 1, which shows typical SPLs produced by the various elements of a drum kit.
- The cardioid, however, does have an advantage over the omni regarding off-axis rejection. Many people are under the misapprehension that a cardioid mic rejects all off-axis sound. Although there are mic's designed with a very high degree of directivity (compromised to the extreme in terms of frequency response), cardioids generally do not operate like guns - point and shoot. Rather the rejected sound is tailored off gradually, and only truly effective at 180 degrees. A comparison of how effective a cardioid is to an omni, in this respect, is shown in Figure 2. As a general rule, an omni at a certain distance will experience the same amount of leakage as a cardioid at twice that distance.
As with a magnetic field, any physical objects placed in a sound field (an area in which sound waves are travelling) will affect the way in which the sound signal behaves. The larger the object, the more drastic the effect. There is a very good reason, for example, for hanging a bulky microphone upside down in front of a vocalist - it helps to avoid sound waves bouncing back off the body of the mic, which in turn are reflected back off the vocalist and picked up as a delayed and distorted signal by the mic's diaphragm. In the same way this delay effect can exist within the body of the microphone itself: the protective cages shielding the diaphragm of a large mic act as mini reverberation chambers, in which the sound wave bounces back and forth, inevitably affecting the original signal. Naturally, the area is relatively small and the reverberation time therefore relatively short, but the sound produced is often audible and measurable by the appropriate equipment.
Subjectively, many people would describe this type of microphone as having a 'fat' or 'warm' sound: in effect, the original signal has already passed through an effects processor before it gets to the mixing console or recorder input. This is fine if the engineer is aware of this effect and is using it sparingly; but if everything, say, on a 32-track mix is recorded in this way, unknown delay times at unknown frequencies will be introduced on every track, resulting in a muddy sound which can't be rectified.
Ten years ago this may not have been a problem, in that other pieces of equipment in the audio chain were noisy enough to mask the mic problem; but nowadays, when people are recording digitally, these weaknesses stick out like a sore thumb. A signal that is distorted by a microphone is not controllable; therefore, it may be more appropriate to use mics which transduce the signal cleanly and then change the signal using reverb or delay units afterwards.
In some cases, however, the modification of a grid immediately in the front of the diaphragm can produce better results in terms of sound quality. In the case of Bruel & Kjaer 4003 and 4006 mics, for example, there is a choice of three different interchangeable grids which can be used to alter the characteristics of the pickup pattern and frequency response (see Figure 3 showing polar response with standard grid, back grid, and nose cone).
These grids are manufactured with precision accuracy and all modified parameters are measured and documented, enabling the user to retain full control of the sound signal objectively as well as subjectively. Other microphone manufacturers are following this idea of acoustic modification, at least in principle. AKG, for example, have recently released a grid which retrofits to their C1000S, modifying the pick-up pattern from cardioid to a form of super-cardioid.
When using a microphone, the sound engineer should be aware of all aspects of physical interference in the sound field. Even a mic clip can cause undesirable reflections, and should always be attached as far away from the capsule as possible. Likewise, many suspension mounts, although mechanically effective, can have an horrendous effect on the sound field purely because of their size. It is also worthwhile remembering that suspension mounts are only truly effective at a 90 degree plane to the diaphragm and rarely necessary when used with sound pressure microphones (the diaphragms are not as flexible as those of the pressure gradient design).
Where applicable, the use of foam windshields is advisable, especially with condenser mics. The shields will assist in protecting the diaphragm from moisture and grease, which can build up over the years to a point where the response of the microphone is degraded. From an acoustical angle, it is necessary to use a windshield with an inter-cellular structure - ie. the cells are all connected. An easy way to check this is to hold the windshield up to the light, and as it is turned you will be able to see through it. If a lesser quality windshield is used, it will simply mask the capsule from the sound source.
Response characteristics of a microphone are illustrated by manufacturers in various ways. Commonly published response charts show either a linear frequency response or the polar/pick-up pattern of the microphone (Figure 4).
The vertical axis is calibrated in dB (decibels) and shows the fluctuation of level related to the frequency (horizontal axis). An accurate microphone will exhibit a flat response without either dramatic peaks or dips. Often a cheap mic, at first sight, will compare favourably in this respect with a high quality, more expensive microphone, and it is sometimes difficult to determine between manufacturers' objective measurements and marketing departments' poetic license.
Response charts should be studied carefully and particular attention paid to the scale of the dB axis. The scale should be at legible increments of at least 1 dB, and preferably ½dB. At this scale no microphone will be absolutely flat, but should remain in a tolerance window of at least + or -2dB. Ideally, the user should be supplied with an individual calibration chart which refers to that particular mic, rather than a 'typical' response chart. Individual microphones of the same type can fluctuate wildly in terms of frequency response. Check also that the response is as flat as possible within the audio bandwidth (20Hz-20kHz).
Figure 4 represents the directional pick-up characteristics of an omnidirectional and a cardioid mic respectively. The charts should be viewed from a three dimensional point of view, showing the response of the microphone with diaphragm pointing 'north' at the central point. On a chart such as this, the 'perfect' omnidirectional mic would appear as an exact circle, with all frequencies at equal levels all the way round. Unfortunately, the physical size of the microphone's diaphragm dictates that the pick-up becomes directional at high frequencies - it is important to note that the smaller the physical size of the diaphragm, the less directional it becomes at high frequencies.
Referring to the cardioid polar response diagram in Figure 4, it can be seen that the back rejection of signals is clearly illustrated. This is achieved by a complicated porting system which delays sound signals arriving at 180 degrees enough to achieve cancellation with the same sound arriving on-axis. The difficulty in designing a cardioid microphone lies in retaining a uniform off-axis response (when all frequencies, in terms of level, are as close to each other as possible). An examination of the response of a lesser quality mic will show that certain frequencies do not decrease uniformly off-axis.
This point is important for a number of reasons: feedback, for example, is often caused by a microphone 'favouring' a particular frequency - if, say, the off-axis response is more sympathetic to 5kHz than most other frequencies, this will be reproduced through the speakers and will in turn be picked up by the mic again, eventually resulting in a feedback loop. This restricts the overall level, which could be greatly increased if the response characteristics were more uniform. Equally, a uniform off-axis response is important in an application such as miking a drum kit, when sound spillage between microphones is often unavoidable. If the microphones being used have an accurate off-axis response, any sound leakage that occurs will simply be different in terms of level, and not coloured in terms of frequency anomalies. To best capture the magic of drums, it is vital to begin with the most transparent sound possible, as no amount of post-processing can turn a dull thud into the 'real' drum sound required.
When choosing a mic for a particular application, there are certain criteria that need to be considered.
- The environment - is it dead, reflective, etc? If you can tailor the environment to the sort of sound you require (by screening or using different materials such as glass or stone), so much the better. This will reduce the need for using electronic EQ and enable you to control the sound subjectively by letting your ears be the judge.
- If close-miking, consider the characteristics of your microphone (eg. the proximity effect of a cardioid). If you like the acoustic sound of your source, try using an omnidirectional mic first. If it sounds too bright, turn the omni off-axis until you have the right mix. If you want to warm up the sound by enhancing the bass frequencies, use the proximity effect of the cardioid by moving it gradually closer to the sound source. Either way, you can reduce the amount of electronic EQ used and retain a cleaner signal path.
- If 'far'-miking, consider the effects of air absorption of high frequencies. Use a mic with a treble boost, preferably one that is achieved by acoustic modification rather than electronically. Remember also that the proximity effect of cardioid mics also works in reverse. Consider whether you will lose too much low frequency signal. If the environment allows, try an omni instead - this will guarantee a flat response down to 20Hz.
- To be able to use a microphone to its full advantage in the above ways, it is essential that it be of high quality [such as a B&K, Schoeps, or Neumann mic, right? - Ed.]. It is pointless trying to use the proximity effect of a cardioid if the mic starts 'clipping' because it cannot handle the sound pressure levels being produced by the sound source.
- Like an artist with his brushes or a golfer with his clubs, you need to build up enough experiences (and experiments) to be able to know the characteristics of your mics and which to use in different situations.
Up until recently, omni mics were definitely taboo in multi-microphone situations. In the last two years, however, engineers have been finding that omnis can offer distinct advantages over the trusty old cardioid. For recording purposes, they have come into their own when a 'live' sound is required, especially for jazz recordings. For example, Tom Jung, award-winning engineer/producer and President of Digital Music Products (DMP), an all-digital, jazz CD label, makes all his recordings live to a 2-track digital recorder. His philosophy is to convert the sound into digits with as little signal processing as possible, concentrating on using high quality omnidirectional microphones to retain the clarity of 'real' instruments. "Digital hears the ugliness in bad mics," says Jung. "Where I think omnis (B&K) really shine is digital, and even more so live digital; the mics are so clean and there's so much headroom - those are the things that I really like... and the phase accuracy."
Provided instruments are positioned in such a way as to account for the omnidirectional properties of the mics (that is, not all directly in front of a wall of Marshall stacks!), a stunningly realistic sound can be achieved without the need for banks of EQ and outboard equipment.
Of all acoustic instruments, drums and percussion produce the highest sound levels. If a tight drum sound is required, it is necessary to position the mic very close to the source, which often means dealing with sound pressure levels up to 160dB. Microphones which cost £100, unfortunately, will have trouble transducing accurately at these levels and transients. There is also a very subtle side to percussive instruments. Cymbals, for example, as well as being very loud, produce extremely high frequencies. With a mic that starts dropping off well under 20kHz, a great deal of the sparkle is lost.
Although microphone choice and placement techniques vary dramatically, omnis appear to be emerging as popular choices for kick drum, snare drum, and overheads. Dynamic cardioids have often been the choice for snare, but if they don't clip under pressure, they certainly distort the sound. The advent of digital recording has increased the requirement for a clean, tight sound, hence the new popularity of high quality condenser mics.
One interesting technique Bruel & Kjaer have come across recently is to use a cardioid condenser (capable of 158dB before clipping) on the top skin of the snare drum and an omni condenser (frequency response 10Hz—40kHz) on the bottom skin, the cardioid allows the engineer to fatten up the sound using the proximity effect, and the omni picks up the high frequencies produced by the snare itself. The two mics are mixed together, level-wise, to give a truly explosive sound. Furthermore, this technique is used live where acoustic dynamics are a priority. For a kick drum recording, omnis are being adopted not only for high SPL reasons, but also for their ability to transduce the drum shell reflections accurately. The general technique used is to position the mic inside the drum itself, about two inches from the back skin and pointing towards where the shell and skin meet. This gives a perfect mix between the bass thud and the slap of the beater against the skin. Many kick drums have been sampled in this way.
As left and right overheads, omnis are popular because of the very transparent, ambient sound they produce, as well as their excellent stereo imaging.
The clarity that can be achieved with top-of-the-range condenser microphones on acoustic guitar is very hard to beat. Utilising either an omni or a cardioid with a very smooth off-axis response, placed 10-20 inches in front of the upper fingerboard and angled slightly towards the sound hole, allows the guitarist greater freedom of movement generally resulting in a more relaxed performance.
The harmonic subtleties produced by acoustic instruments are a necessary part of their sound in the way we hear and react to them. Take these away and you are left with only an acoustic 'shadow', which at best only resembles the instrument.
Miking electric guitar poses different problems. In many recording situations it is necessary to capture the sound at very high volume. John Eden (producer of Status Quo's Rocking All Over The World album) uses a B&K omni positioned about two to three inches from the loudspeaker cabinet. If a more 'open' sound is required, place another microphone a few feet further back and mix to taste.
Brass instruments, with their wide dynamic range and sometimes brutal high-end transients, can be a problem to record. For trumpet, cornet or trombone, try placing an omni mic 6-16 inches from the bell. This should reproduce every subtle detail of a musician's tone.
Use the same technique for saxophone, but slightly favour the valve side of the instrument. Through experimentation, you should be able to find a musician's 'sweet' spot - which varies according to the player's personal style.
Sampling is an area where a truly versatile mic is a necessity. Because of the varied environments and sounds involved, it is important to be aware of every aspect of the mic's performance. There are really no hard and fast rules regarding miking technique, except that in general most people tend to favour the sound of a sample containing first reflections (ie. slight natural reverberation) rather than a completely dead sound. Generally, the purer the sample, the greater the possibility of manipulating it later on.
The introduction of R-DAT machines has made it much cheaper and easier to record digitally, particularly in live situations. Classical music recording has always remained a case of live recording, whether in a studio or in a concert hall. Often certain instruments or soloists are spot-miked, but in the right environment a minimalist approach (employing two mics) is often the easiest technique and produces the best results.
In the last year or so, a lot of classical music has been recorded direct to portable R-DAT, from which the CD has been pressed. Simple, effective, and cheap - for approximately £3000 worth of good equipment, master quality digital recordings can be obtained. The Denon Recording Company, for example, recorded three of Mahler's symphonies with only two B&K 4003 mics, which when released were acclaimed as the best ever recordings of these pieces of music. Many broadcast companies, including the BBC, are adopting a similar set-up (two mics mounted either side of a 'Jecklyn' disc, and a portable R-DAT machine) for outside broadcast work.
With the advent of digital recording equipment, lesser quality and often cheaper microphones and speakers are becoming the significant weak spots in the audio chain. As speakers and mics are likely to become the only two remaining analogue pieces of equipment in the recording process, their quality will be of critical importance. But unlike microchip technology, microphones are less likely to become cheaper, simply because of the precision techniques involved in making them.
So how does the cost of microphones relate to the rest of your recording equipment? Separating all the vital pieces of hardware in the audio chain and attributing them, quite rightly, with equal importance, the list probably reads as follows: microphones, mixing console, recorder, reverb, multi-processor, amps, and speakers. The price of a reasonable 8-track package is around £6000 - divide this by the number of sections in the audio chain, and the cost for each section comes out at around £1000. However, it is very unlikely that many people would spend this much on microphones in such a package.
Bearing this in mind, add up how many cheap specialist mics (ie. recommended for specific instruments) would be needed to cover the different sounds you record: probably around six at, say, £150 each - a total of £900. Take into account the fact that the mics will very likely be unmatched in terms of their frequency response and phase anomalies, and paying a little more for the quality rather than the quantity starts to look cost-effective. You will also gain a precision instrument with which to gain valuable miking experience - the type of experience that separates the men from the boys in professional circles. After all, there is very little skill in putting a tape recorder into Play mode, whether it be an Amstrad cassette deck or a Mitsubishi X850 digital multitrack. The quality mic will also stand the test of time and respond to the changes in digital technology.
Until a development like laser scanning of sound waves - and a method of interpreting them digitally - is evolved, there will be little alternative to the analogue transducer, or "the piece of wobbly metal" as George Martin once described it. But in terms of dynamic range and frequency response, the best mics available today still out-perform current digital specifications. In the short term, we will probably see analogue-to-digital convertors replacing microphone electronics, but hopefully not until the digital technology can handle the 150dB dynamic range currently achieved by analogue technology.
Bruel & Kjaer (UK) Ltd, (Contact Details).