Choose a mic with the optimum polar pattern for the job, and you’re halfway to capturing a great recording.
A lot of inexperienced recordists reach instinctively for cardioid mics to record just about everything, believing that this pattern inherently captures less unwanted room sound. But, while cardioids often do work well, there can be very good reasons to make a different choice, and if you learn to understand the strengths and weaknesses of the various microphone polar patterns, you’ll be able to make better recordings.
Although I said ‘various’ patterns there are actually only two fundamental ones: figure‑of‑eight and omnidirectional. All of the others you’ve heard of are derived from these two.
All mics respond to air pressure: the capsule’s diaphragm is only moved when there’s an air pressure difference between its front and rear faces. But they behave differently, depending on the design of the capsule around the diaphragm.
If both faces of the diaphragm are open to the air, it’s what we call a ‘pressure gradient’ mic and it will have a figure‑of‑eight polar pattern. Often abbreviated to ‘fig‑8’, these mics are so called because they’re equally sensitive to sound at the front and rear, but insensitive at the sides: their pickup pattern resembles the number ‘8’ or, in three dimensions, two slightly squished spheres (shown above). Though equal in sensitivity, the front and rear pickups are opposite in polarity. Sounds approaching the diaphragm edgeways on are rejected because they cause the air pressure to be equal on each side of the diaphragm — so the diaphragm doesn’t move.
If, instead, you place the diaphragm across the mouth of a sealed chamber, you’ll have an omnidirectional, or ‘omni’ mic. These sense the difference between the air pressure at the diaphragm’s external face and the air pressure inside the chamber. Since there’s no difference between the external air pressure at the front, rear and sides, omni mics are equally sensitive to sound arriving from all directions: their pickup pattern resembles a circle, or rather a sphere, with the capsule at its centre.
Omnis and fig‑8s both retain a good balance of frequencies as the source moves off axis. While the sensitivity of the fig‑8 falls off as a source moves towards the side of its capsule, its overall tonal balance remains essentially constant; all frequencies are attenuated equally.
Also worth noting is that fig‑8 mics can be quite susceptible to handling noise and wind noise, and they have a strong proximity effect: as they’re moved closer to the source the low‑frequency sensitivity increases more than the high‑frequency sensitivity, the practical upshot being that the overall balance is skewed in favour of the lows. Omni mics have no proximity effect and are less susceptible to handling and wind noise.
Fig‑8 mics have an evenly balanced off‑axis response, even though the overall level drops as you move towards that 90 degree null.
As mentioned, the directional polar patterns (eg. cardioid, wide cardioid and hypercardioid) are blends between omni and fig‑8. So if you placed omni and a fig‑8 capsules in close proximity and kept them equal in level, their combined pickup pattern would be the same as with a dedicated cardioid mic. This happens because the fig‑8 mic’s signal is the same polarity as the omni’s at the front but opposite at the rear: their patterns reinforce at the front but interfere at the rear.
The result is the familiar heart‑shaped pattern in front of the capsule (hence the name ‘cardioid’) with a null at the rear. Alter the balance between the mics and you vary the shape, including the width, of the virtual pattern. As you increase the fig‑8’s contribution, to narrow the front pickup, a lobe appears at the rear, which is why a hypercardioid mic tends to be least sensitive 30‑45 degrees off its rear axis, rather than directly behind the capsule.
Some ancient cardioid devices were made in exactly this way, with two capsules mounted close together in a single mic. Most fixed‑cardioid mics today, though, have a single omni capsule and use a clever bit of acoustic trickery called an ‘acoustic labyrinth’ to change the pickup pattern (what follows is a deliberately simplified explanation but hopefully enough that you understand what’s happening).
The acoustic labyrinth comprises tiny holes in the rear of what would otherwise be an omni‑pattern assembly. Sounds from the rear pass through the holes and into a chamber, which is a sort of ‘maze’ that compensates for the time it takes the same sounds to reach the front face of the capsule. Thus, the capsule combines pressure (omni) and pressure‑gradient (fig‑8) operation — so it functions partly as an omni and partly as a figure‑of‑eight, with different labyrinth designs determining the width/shape of the directional pattern.
A compromise inherent in this labyrinth approach is that directional mics aren’t equally directional at all frequencies. Typically, a mic can be made fairly directional at mid and high frequencies quite easily, but the pattern steadily widens as you drop lower in frequency. This means that if the mic has a nominally flat on‑axis frequency response, the off‑axis response will tend to be skewed in favour of lower frequencies. Some leading manufacturers have managed to make the off‑axis response as even as possible. But in the vast majority of cases, if you record a sound on‑axis and then repeat the process while aiming the side of the mic at the source, and you then compare the results, the off‑axis recording will sound duller, and perhaps even ‘muddy’.
Directional capsules also exhibit a combination of the other characteristics of fig‑8 and omni mics: the narrower the pattern, the greater the similarities to fig‑8 mics in terms of handling noise and wind blasts. With very few exceptions, these mics have a proximity‑effect bass boost too, though it’s not as dramatic as with a fig‑8. Of course, there’s no proximity effect at all for sounds arriving at 90 degrees off‑axis, where the notional fig‑8 component has its ‘null’.
We’ve introduced the main patterns and characteristics, then, but where might you choose to use a mic with one polar pattern instead of another, and how do you make the most of your choice?
Despite what I said at the outset, a cardioid‑pattern mic is a perfectly valid choice. Pointing the insensitive rear of a cardioid mic at the sound you want to reject will help reduce spill from that source. In a home studio, for example, this could help to reduce the pickup of fan noise from a computer. But you do need to be aware that ‘room sound’ comprises reflections which arrive at the mic from all angles and that cardioid mics are both pretty sensitive and darker‑sounding at the sides. So, unless the room is fairly dead acoustically, you could still end up with a lot of ‘boxy’ room sound. A practical solution is to screen the sides and rear of the mic by forming a ‘V’ shape with folded blankets or duvets suspended on a frame, or by using a commercial screen such as sE Electronics’ Reflexion Filter.
When erecting such screens, always keep in mind that the polar pattern exists in three dimensions. In other words, the off‑axis response extends not just to either side but also above and below the mic, so it will capture any reflections arriving from the floor and ceiling, and you may need to improvise screens above and below the mic too. Note that screens will only have a useful effect on the mid and high frequencies.
With any mic, whatever its polar pattern, the closer you can place it to the source the greater you’ll make the ratio of wanted sound to unwanted spill/reflections. But unless using an omni mic, getting close up means contending with the proximity‑effect bass boost. Where this boost isn’t desirable you can try switching in a low‑cut filter or applying some low‑cut EQ. Note, though, that as vocals are often recorded up close the LF response of many cardioid mics that are designed for vocal use is already deliberately rolled off to compensate to a greater or lesser extent for the proximity effect. It’s most noticeable with live vocal mics, which tend to be used much closer to the mouth than studio mics do, but this ‘voicing’ isn’t exclusive to them and it presents the opposite challenge: if you use such a mic at a greater distance it can sound bass‑light, and might benefit from a low-frequency EQ boost.
Proximity effect boosts aren’t the only thing to listen for when miking up close. Importantly, the closer you get, the more the mic will favour one part of the instrument over the rest — my ‘How To Mic Anything’ article in this SOS March 2021 issue provides a good rule of thumb for your initial mic placement.
If you’re working with directional mics in any stereo array or a recording with multiple mics and sources (eg. a drum kit), you should again bear in mind the off‑axis coloration. For example, with an X‑Y pair of cardioid mics, the mics will point towards the edges of the source, and the edges are thus likely to be captured more accurately than at the centre. In a drum kit, a tight pattern can often helpfully reduce spill in close mics, but the trade‑off is that what spill is captured might not combine so well with the same sound on other mics. In other situations, the darker off‑axis sound can be turned to your advantage. For example, if miking a guitar cab with a cardioid mic and you find that the miked signal sounds too bright, you can try angling the mic to ‘darken’ the sound a little.
There’s a tendency to think of fig‑8 mics as being useful only for exotic stereo arrays such as Mid‑Sides, in which they’re combined with another mic (which might be forward‑facing, omni or fig‑8). But there can be practical advantages in using a fig‑8 mic in more conventional studio roles too. While it isn’t often that you need a mic able to pick up equally well from its front and its rear (though it can certainly be useful if recording two similar instruments facing each other), there are many occasions where you might benefit from a mic that’s perfectly insensitive to sounds coming from the side.
I mentioned earlier that fig‑8 mics have an evenly balanced off‑axis response, even though the overall level drops as you move towards that 90 degree null. Because of this, you can concentrate on aiming the ‘deaf’ side of the mic at any sound you don’t want to pick up, rather than aim the main axis exactly at the sound you wish to capture.
A common scenario for exploiting the off‑axis rejection of a fig‑8 mic is the singing acoustic guitarist, since using cardioid mics can result in too much guitar getting into the vocal mic and too much vocal in the guitar mic. By positioning two fig‑8 mics such that the ‘deaf’ axis of the vocal mic looks at the guitar, and the deaf axis of the guitar mic points at the singer’s mouth, the level of spill can be usefully reduced. The rejection won’t be perfect, as the sound sources aren’t infinitely narrow points, and reflections off the floor, ceiling and walls will find their way to the more sensitive parts of the mics. But in practice it’s a very useful technique that can get good results when other approaches fail.
You don’t have to limit the fig‑8 rejection technique to singing guitarists, though. In fact, you can use that side null to your advantage whenever there’s a sound you want to avoid capturing. For example, in a home studio it might offer a useful alternative to the cardioid in reducing the amount of computer fan noise that’s being picked up. Where the fig‑8’s rear pickup sensitivity proves problematic, some sort of acoustic screen can usually provide a remedy. Again, you could improvise such screens with duvets, try using a Reflexion Filter or similar type of device, or even point the rear of a fig‑8 mic towards a corner bass trap if you have one. Remember that, as with omni mics, such spill as is picked up should at least sound pretty natural.
You have to bear in mind, though, that the narrower pickup pattern won’t suit every source. For instance, if you’re podcasting and naturally tend to move about as you speak, you might notice your voice fluctuating in level as you move off axis. In that case, you need to move further from the mic and accept the increase in room sound, or choose a mic with a different pattern. You also need to listen out for the proximity effect of fig‑8 mics. For example, I love the sound of a ribbon mic fairly close to a guitar amp, but often have to apply a drastic low cut or high‑shelf boost to counter the significant bass boost.
The omni pattern tends to be rather underused by home recordists, presumably because people assume that an omni mic will capture more room sound, but I urge you not to overlook it: because omni mics can pick up sounds from any angle without changing the tonal balance and they have no proximity effect, it’s often possible to place them rather closer to the sound source than you can a cardioid mic, without compromising the tonal balance of the instrument.
As with other mics, you can still use screens to reduce spill if necessary: hanging a folded duvet or placing a pillow behind an omni mic can often be enough to tip the source‑to‑ambience balance in your favour. But what spill a true omni does pick up will at least be tonally accurate. As a result, I find that you can generally place an omni mic 30‑ to 40‑percent closer to the source than a cardioid one without compromising the sound; you’ll end up with a similar amount of spill, but of better quality. (I prefer to use a single‑capsule true omni mic rather than the omni pattern of a multi‑pattern, dual‑capsule mic, as the former tends to come closer to the theoretical ideal of a perfectly even off‑axis response.)
This isn’t an ‘all you need to know about polar patterns’ article: that fascinating subject could fill a book! Instead, I’ve focused on what I feel are the most important factors, and played down anything that is too deep technically, or isn’t really relevant to how mic patterns are used in practice.
We’re often so worried about getting rid of spill that we overlook the cohesive properties of well‑recorded spill. Many classic records owe their sound to the fact that they include far more spill than is common today, and key to making spill your friend rather than your enemy is a basic understanding of how mic patterns behave when picking up off‑axis sound. Yes, the cardioid mic is a mainstay of modern recording for a good reason, but you should ensure that you’re aware of what it’s picking up off axis and what it does to that off‑axis sound. And if you tend to reach for a cardioid mic purely out of habit, I hope that I’ve prompted you to think more about using omnis and fig‑8s in particular, and the situations in which they might deliver better results.
The boundary mic, sometimes known as a ‘pressure zone mic’ or PZM, is a special case. It involves mounting a microphone with an omni capsule flush with the surface of a wall (or other large area of solid material) so that they can only pick up sound from one hemisphere. Because the mic is flush with the boundary or, in some cases, placed a few millimetres above the boundary and aimed towards it, it responds only to the air pressure changes that occur at the boundary, and it can’t pick up any reflections from that boundary to cause phase cancellation. Note that the low‑end response is also affected by the proximity to the boundary (there’s a 6dB bass rise at a solid boundary, such as a wall) so while you could use any omni mic in this position, dedicated PZMs are designed specifically to deliver a flat response when positioned on a large, flat surface.
A PZM can be useful in capturing a natural sound in situations where boundary reflections might otherwise pose a problem. A practical example is when using a pair of boundary mics stuck to the ceiling above a drum kit: these may do a better job of capturing a clean overhead sound than two mics on stands, especially if the room has a low ceiling, since the ceiling reflections won’t be picked up. Similarly, two boundary mics stuck under a grand piano lid aren’t affected by reflections from that lid.
A type of mic you won’t often see in music studios is the interference tube microphone, often referred to as a ‘shotgun’ mic because of its long, thin construction. They’re also used in theatre and film sound, and smaller versions are often seen on top of video cameras. Their ‘superpower’ is that they are extremely directional, so you get the maximum sound pickup from wherever you point the camera.
At the heart of the mic is a fairly conventional cardioid or supercardioid capsule, but this is fixed to the back end of a long tube that has slots machined into the sides. On‑axis sound passes down the centre of the tube to reach the capsule. Off‑axis sound, though, is forced to reach the diaphragm via the side slots. This unwanted off‑axis sound will enter the tube via multiple slots at the same time, so will arrive at the diaphragm with varying phase relationships, because each slot is a different distance from the capsule.
The idea is that the sound received via the slots will partially self‑cancel, but the amount of cancellation depends on the frequency, the angle of approach relative to the mic’s axis, the length of the tube, and the slot spacing. Low frequencies, which have very long wavelengths, don’t benefit from the geometry of a practically sized interference tube: you’d need an impossibly huge shotgun mic to be effective down to bass frequencies! Most commercial shotgun mics have little effect below 1.5 to 2 kHz in terms of controlling directivity. Below this, the polar pattern widens until, at low frequencies, you’re pretty much back to where you were with a standard hypercardioid capsule.
It’s also worth noting that the off‑axis rejection is related to the slot spacing; you don’t get the smooth roll‑off with angle that you do with a conventional cardioid capsule, but rather a series of peaks and troughs, and if the off‑axis sound moves, so do those peaks and troughs, resulting in a distracting phasing sound. This is one reason why you rarely see these mics used in a music studio.
A cardioid‑pattern mic is obtained when the capsule employs equal amounts of pressure and pressure‑gradient operation and, as discussed in the main text, most practical cardioid mics today achieve that by using an integrated acoustic labyrinth. This allows sound waves to reach the rear of the diaphragm with a slight time delay.
The idea was first conceived in 1935 by two German engineers working for the Reichsrundfunk Gesellschaft (RRG Electroacoustic Laboratories). Dr. Hans Joachim von Braunmühl and Dr. Walter Weber developed a capacitor mic whose backplate was constructed from an assembly of separate metal discs machined with strategically drilled holes of different sizes and relative positions. This disc assembly produced multiple chambers with indirect passages from one side to the other, creating a convoluted pathway for the sound waves, taking time to navigate and hence introducing an acoustic delay. Alternative mechanical forms of this labyrinth idea were subsequently invented for moving‑coil microphones, most notably by Benjamin Bauer at Shure for the iconic Unidyne Model 55 microphone, released in 1939.
As Braunmühl and Weber’s cardioid capsule inherently employed a static charge between the diaphragm and backplate, they were concerned about dust and moisture being attracted into the labyrinth and blocking the passages over time. To prevent that, they covered the back of the labyrinth plate with a second (passive) diaphragm, forming a dust‑proof enclosure that would still allow sound waves from the rear to pass into the labyrinth. This design was patented in 1936, and was introduced commercially as the infamous Neumann ‘M7’ capsule.
Later experiments with biasing both the front and rear diaphragms resulted in the world’s first switchable‑pattern capacitor microphone. As discussed elsewhere, by changing the biasing voltage and polarity of the rear diaphragm relative to the front, and combining the electrical outputs from both diaphragms, any desired polar pattern could be created, from omni through cardioid and on to fig‑8. Although originally intended purely for dust protection, and later for enabling switchable polar patterns, the rear diaphragm of a B‑W capsule also changes the capsule’s proximity effect for close sources, compared with a single‑diaphragm cardioid mic.
Today, countless capacitor mics still use capsules based on the original M7 design, and generic dual‑diaphragm capacitor mic capsules are referred to as Braunmühl‑Weber (or B‑W) capsules, in recognition of the pair’s groundbreaking work. In essence, the rear diaphragm introduces a significant compliance at low frequencies, which means that the acoustic impedance of the rear diaphragm increases as the frequency gets lower. The result is that the rear diaphragm impedes low frequencies from entering and reaching the rear of the front diaphragm, altering the net pressure acting on the front diaphragm in a frequency‑dependent way.
What this means in practice is that for close sound sources the dual‑diaphragm capsule behaves less and less like a directional pressure‑gradient device as the source frequency decreases, and more and more like an omnidirectional pressure‑operated device. Since pressure‑operated capsules have no proximity effect, the cardioid B‑W capsule also enjoys a reduced proximity effect for close sound sources. This is a very handy side‑effect when recording vocals: it means that when a singer moves relative to the mic the bass response changes much less than it otherwise would. It also makes the mic much less susceptible to plosive popping. Consequently, dual‑diaphragm capacitor capsules are employed almost exclusively in microphones intended for recording studio vocals, even in mics with fixed‑cardioid polar patterns. Conversely, a single‑diaphragm cardioid capsule provides a more consistent polar pattern with frequency for distant (far‑field) sound sources, rejecting low‑frequency rearward sources more effectively. So single‑diaphragm cardioid mics are generally preferred for more distant‑miking applications such as in ORTF stereo arrays. Hugh Robjohns