This review page is supported in part by the sponsors whose ad banners are displayed below

Directional sensitivity. Aside from how a microphone converts air pressure into voltage, directional sensitivity is arguably the most decisive factor. Essentially this is the inverted principle of a loudspeaker’s radiation pattern. At least in theory a poorly working microphone is the one which captures sound regardless of which direction it arrives from. That’s called an omni. But even omnis get directional at high frequencies where their microphone body itself shadows them. Because with a nearly omnidirectional sensitivity room sound becomes dominant, this type of microphone is exploited predominantly with classic music where hall sound is a vital ingredient. These mostly small-diameter condenser pressure receivers are amongst the most neutral and natural microphones and exhibit a very flat expanded bandwidth. It’s why they often show us as measurement tools. Directional mics are sorted by how strongly they attenuate off-axis reception and how their specificity looks graphically. A microphone that’s maximally insensitive to sound at the rear whilst continuously increasing in sensitivity around the front will exhibit a kidney-shaped diagram. Hence they're called cardioid. More directionally selective types are called super or hyper cardioid.

Carnegie Hall [Wikipedia!] | omni, cardioid or figure eight

Specialty microphones, most often ribbons, are equally sensitive front and back but dead to the sides. This trait is called figure eight. The knowledge of microphone directionality is essential to the sound engineer who can hide neighbouring sound sources in the microphone's ‘dead zone’. In the mixing process it’s far easier to work with maximally isolated performers and sound sources. This explains the rise of switchable mics. This usually combines two cardioid mics back to back whilst switchable circuitry for addition and subtraction can generate all patterns between omni and figure eight. Incidentally and different from speaker design, nearly all microphones are widebanders since separation into two bands would create more issues with amplitude and phase differentials than it solves even if some types will fade out in the extreme highs and lowest bass. But with most sources that’s not problematic. They don’t operate across the full audible bandwidth to begin with. The human voice for example has barely any components below 100 cycles. The final aesthetic of a sound mix may also not rely on any ultimate definition in the high treble.


It’s interesting that over the past half century microphone design has occurred predominantly as an optimization or refinement of existing concepts which save for a few exceptions remain the most commonly used types today. What’s more, vintage models with very strong signature colourations are hawked for extreme stickers. Though both more than 60 years old, the Neumann U47 and  AKG C12 remain the most legendary condenser models. Even today Abbey Roads Studios uses their inventory of them in daily production. Most sound studios offer their technicians a good variety of microphones. These include flexible work horses as well as specialty mics with strong personalities for very particular applications, perhaps even dummy head and boundary microphones. This microphone subject is thus the most emotional and important to the trade. Not for nothing is the symbol of the microphone a universal sign for recording. A special sector of microphony is the recording of classical music. Whilst contemporary recordings of a single stereo pair of microphones have become rare, it’s not uncommon that two ideally placed mics are responsible for the majority of the signal capture and dominant localization of individual sound sources. Here the technician exploits a range of stereo principles.


One of the most basic and simplest microphone arrays is to set up two of the same type in one place slightly rotated against each other. This XY principle works with amplitude differentials. If as so often this relies on small-membrane cardioids, the frontally arriving sound obviously exhibits the highest levels.  If the microphone later associated with the left loudspeaker ‘looks at’ an orchestra’s first violin section for example, it will capture it very clearly whilst the right microphone only does so very attenuated. The advantage of this approach is that it works with just one microphone position. Such recordings can exhibit extreme localization focus whilst missing some spatial depth. Obviously the recording engineer will determine not only the microphones’ position but other parameters of his pair including directionality, type and, vitally important, the angle by which both are rotated to determine stereophonic width. But our ear/brain determines location not merely on amplitude but also path-length differentials. That's because sound takes time to travel. Rather than attenuate the signal in one playback loudspeaker, one can delay it for a similar effect. If a simple stereo mic recording is supposed to contain such path-length data, its microphones rely on a certain distance between themselves. That’s prosaically known as AB. This could be a few centimetres or a few meters. A/B is popular with large ensembles in premium acoustics. Whilst image focus suffers, other systems won’t capture the same ‘volume’. Most AB setups rely on omnis which guarantee very linear response down into the lowest bass.


Aside from XY and AB, a number of mixed forms exist like ORTF with its fixed parameters relative to directionality, distance and angle.  And there are other ways to generate a stereo image. One is the quite complicated MS where one mic is dedicated to central mono signal, the other to stereo capture;  the triple-mic Decca Tree; the Blumlein method of two figure-eight mics at 90°; systems with central divider walls; dummy head and more. Obviously surround sound has its own equivalents like the Fukada tree, the IRT cross and INA5. But enough already. The main microphone system is often supplemented with supportive spot microphones on specific instruments or instrumental sections, sometimes only to heighten certain musical passages. Should a harp, woodwind, soloist or the violas be a bit subdued on the main mics (hard to imagine with violas), they can be captured discretely. Aside from the obvious gain in amplitude, these data can now be manipulated discretely. This might include a lift or softening of the highs. In the later mix it’s simply important to assure that the main microphones remain dominant. This is achieved not only because the auxiliary mics are set lower but because their signal will get electronically delayed to not arrive ahead of the main microphones which sit farther away.


Microphone signal relies on extreme amplification factor to achieve useful levels. That's similar to vinyl cartridges whose output signal is comparably tiny. This becomes a job for the microphone preamp which might also be built into a mixing console or PC audio interface. These options too span the gamut from dirt cheap to painfully dear, from purist ultra-clean transparent to highly coloured variants where valves and transformers often leave a deliberate aural fingerprint. But just as with microphones, the particular construction or use of certain parts alone isn’t predictive. There are very lucid ultra-fine tube mic preamps. Often this component includes extra functions which allow the removal of bass, phase inversion or direct-to-tape manipulations which otherwise are reserved for the mixdown phase. Again as with microphones, it’s important that a sound engineer knows his tool kit and arrives at the right choice of hardware combinations. As you will appreciate, it’s a difficult process that relies on comprehensive know-how just to record some music with a few microphones. And that’s merely the beginning.