by Pablo Bellinghausen –
As we saw in our previous instalment, the large majority of professional microphones are directional, which means they will pick up sound differently depending on the angle the original sound is coming from. This is a lot harder to measure and visualise than just the frequency response from the front, so there are several different ways that this information can be shown.
The easiest way to present this directionality is through a frequency response chart at different angles, as per the following third-party measurement graph for the Røde NT5:
Although it’s very informative about the degree of rejection against any sound coming directly from the back (as well as the colouring that said sound will be subjected to), it is of course unhelpful for any other angle.
For an accurate representation of how a microphone picks up audio information related to the position of the incoming sound, the best graph type to use is called a polar chart, in which the attenuation relative to the front response is plotted in a circular “polar” grid:
This plot will create different shapes, which dictate the name of the microphone’s polar pattern. There are several main types, amongst which the main ones are Omni (picks up everything equally), Cardioid (picks up most of the sound from the front and rejects most of the sound from the back, such as the above picture), Super / Hypercardioid (picks less sound from the sides at the expense of some leakage from the back), and Figure of 8 (picks up from front and back equally). Shotgun and parabolic microphones used for location recording are even more directional but for several reasons are mostly unsuitable for both live and studio work.
If you’re still having trouble understanding how these graphs work, here is a great little flash web app from Indiana University that might help clarify it (click on the different patterns to see the relative sensitivities):
Polar Patterns and Frequency
In an ideal world, cardioid microphones would achieve a 100% sound rejection rate from the back. This means that, in theory, a measured frequency response at 180° would be pure silence. As it happens, the NT5 is a cardioid microphone, but as we can see in the first graph, the level attenuation is on average 20 to 30 dB – far from the full rejection one would expect!
The reason why, of course, is that the electric and acoustic techniques used to create a microphone’s polar patterns won’t work perfectly at all frequencies, and that perfect rejection is pretty much impossible in the real world. Most of the time manufacturers will focus on the directivity at a frequency of 1 kHz, since it’s perceptually in the middle of our hearing range (which works logarithmically) and is also a very easy frequency to reproduce and measure. In the real-world however, the polar pattern of any directional microphone will start to look increasingly erratic the more one deviates from said frequency.
Understanding this fact is crucial in both live sound and studio recording, since in a large majority of the cases the microphone will be in an enclosed space, which will bounce reverberation into the mic from all angles, and will pick up “bleed” from any other source of sound in the room, whether another part of the instrument, or a different one altogether.
This is what the polar patterns of the NT5 look like at several frequencies, and they are typical of most small-diaphragm cardioid condensers. They aren’t any worse than most high-quality models (and it’s worth noting that the NT5 is a pretty good-sounding microphone in its own right) but they’re certainly not perfect – because no microphone is.
There aren’t enough different frequency measurements to convey the actual “jaggedness” of the frequency response at different angles, but this graph does give us a pretty good insight as to how the microphone will react in different positions; for example, from the slight bump at the back at 16k we can tell that there will be more bleed from high frequencies, but can see that the pattern is otherwise quite consistent throughout the front of the mic, except at the highest frequencies, where most microphones will become more directional.
This consistency is why cardioids are the most common type of microphone; they are a great compromise between isolation and sound quality. Most studio vocal and instrument mics will be cardioid, and pretty much every single popular music album ever recorded will have used one on most instruments.
Cheaper microphones will often sound acceptable right at the front, but most likely will start sounding a bit odd as soon as the sound source moves towards a certain angle, so more care needs to be taken than with better models, which will usually sound good throughout the whole front half.
An omnidirectional pattern should technically be the same throughout the audible range, but we can see here that the DPA 2006A becomes a lot more directional at higher frequencies. To a certain extent this is actually by design, since a smooth, polite drop in the highest frequencies will make the sounds at the back feel slightly more distant than those at the front, while still sounding more open and natural than a cardioid model. Due to their construction larger omnis like the 2006A will always be more directional than very small lapel or headset capsules.
If the room one is recording has a great reverberant sound (when recording classical music in a church or hall, for example, or when recording solo instruments in professional studios), it’s often a good idea to try an omni mic. The sound will feel a bit more distant due to the added room sound, but it will invariably be more “real” and accurate. They also don’t suffer from proximity effect, which is something we will see in one of our next instalments.
A rule of thumb is that the more directional a mic is, the more irregular the polar pattern will be, and this certainly shows in the hypercardioid pattern on the Neumann TLM 107. Neumann’s published charts are usually quite smoothed out (the precise graph plot lines for any large dual-diaphragm condenser such as this one should look far more jagged than those of the previous two mics) but it’s enough to see huge differences in response at certain angles (almost 20 dB in some places) which means the microphone will in many positions sound rather odd!
The TLM 107 is an award-winning model by one of the most respected brands in the market; why would Neumann even offer a hypercardioid option on a multi-pattern mic if the response looks that “bad” at certain angles?
The reason why is that the audio world is one of constant compromises, even at the highest quality levels, and that a hypercardioid pattern offers one of the highest overall rejection levels of any studio microphone design. Even though many dynamic models can be even better in terms of sheer isolation, the Neumann does offer that flattering large-diaphragm “sheen” to anything that’s placed in front of it; when a condenser sound is desired but a cardioid lets in too much extraneous noise, it’s a good idea to try a hypercardioid pattern.
Hypercardioid dynamics are very common in loud rock and metal stages when the amount of feedback rejection is critical, but it is important to understand that this will always be in detriment to the quality of the sound of the bleed from other sources; when bleed and feedback are not a consideration, it’s always best to try an omni or a cardioid first.
Advanced: Polar Patterns in 3 Dimensions
Microphones don’t pick up sound in just one plane – they receive sound from all directions. The published polar patterns are measured in one plane only, but if the microphone is not perfectly cylindrical, then the directivity will be different from another plane (usually top-to-bottom as opposed to left-to-right). As if the charts weren’t complex enough even without accounting for that!
However advanced sound engineers can use this fact to their advantage when positioning microphones. It is pretty hard to get that sort of information from manufacturers, but luckily one of the few that do is the BBC.
It may strike some people as unexpected, but the BBC actually designed one of the most iconic microphones in history, a figure of 8 ribbon design now made under the name of Coles 4038. Both the original prototype called the PGS and the commercial-grade 4038 were the BBC’s workhorse microphones for decades and are still in use today.
The BBC PGS prototype and the commercial Coles 4038
One of the published white papers for the PGS has frequency charts not only from different angles around the mic, but also from towards the top and bottom of it:
Here we can see the smooth treble reduction from the side (orange line) is still pretty polite-looking, and besides a reduction in sibilance and slight darkening of the sound, at this angle the results should be more than acceptable. The blue line denoting sound coming from the top is more interesting however, since it displays a pretty nasty cancellation around 10kHz that is caused by sound being reflected inside the mic. The green line, although showing a bump in the 3 to 6kHz area, is far smoother.
This is an aspect that is inherent to most side-address microphones but is rarely discussed; tilting the microphone away from the source of sound will tend to give a more “present” yet duller sound (green line), whereas tilting it towards the source (blue line) will make the sound bounce back inside the microphone, which can sometimes affect the sound negatively. This is particularly important when recording close-up vocals, where tilting the mic is often used to reduce plosives or sibilance.
As with everything related to audio work, it’s important to note “good” and “bad” are inherently subjective, and so what is a technically incorrect way to record might be exactly the sound the song requires! Knowing the reasons why audio equipment behaves the way it does is however not only useful when experimenting, but often critical in situations where time is limited or extra takes are not an option.
Even though there are no official standards in published graphs and intrusive marketing practices often cloud the information, making informed purchasing decisions rather more complicated than they should be, polar pattern plots can tell us a lot about what we can expect from a mic, both in sound and in real-world usage.
There are still several aspects of microphones to go over and we will eventually get to those, but next time we will instead look at some more basic and hands-on advice for vocal recording. Until then, happy recording!
Graphs and other data measurements have been faithfully redrawn based on various sources from manufacturers or online/print publications.
For any further information just drop us a quick message.