Unidirectional microphone is not what you think it is.
Recently I got a question from a filmmaker here in my blog‘s comments: Adriano was wandering if Sony PCM-D100 can be mounted on a DSLR and used as unidirectional microphone for his almost speechless documentary. Intention being to have a nice stereo image in the frame and be sure the mics will reject sound coming from behind mostly produced by the camera-man himself (steadicam, dolly, zoom ring, focus ring, breathing, footsteps etc.).
As I was doing tests and trying to figure out the best answer in bunch of other ways, it struck me there must be a huge misinterpretation of microphones’ directional qualities. Especially in circles of video-centric artists as they are used to comparing things to camera lenses (those are extremely directional, right?).
So instead of just dropping a quick response I dug deeper and decided to write an article about directionality of microphones. I took unidirectional microphone (cardioid) as a reference just because it has the most defined and supposably easiest to understand polar pattern (directionality curve). Although everything you’ll learn can and should be applied to all mics.
Sound newcomers and filmmakers will find this most useful.
Please be advised that after you’re done reading you’ll have a much better understanding of what a unidirectional microphone and polar pattern in general is and be able to decide what mic will fit your next project the best.
Coincidentally lack of such knowledge on the consumer side and lack of information on manufacturer side over the years has formed wrong picture about microphones’ directional qualities between people searching for better sound capture. While this may be less complicated among more sound-involved guys, it’s highly misinterpreted by newcomers or filmmakers wanting to just pop a mic on their camera or DSLR and improve sound quality of their productions. Those people are exactly ones who may find this information most useful.
So let’s break it down and clear it out together, shall we?
How unidirectional is a unidirectional microphone?
Mistake #0. Unidirectional microphone = camera frame of shot?
First thing we need to do is clarify what exactly is directionality of a microphone. There’s a wrong idea wandering around that a microphone and its directional qualities can be compared to camera lens or to be more specific to a frame of camera shot.
Meaning if a camera captures only things that are IN the frame (most likely camera is pointed in that direction) then it’s safe and right to assume that an unidirectional microphone will only capture sound sources located in front of it. This is wrong and the best way to illustrate it is with help of a parallel.
The eyes (camera lens) & ears (microphone) parallel.
Let’s assume your camera lens and frame are your eyes and your microphone is your ears. Your eyes only see what they are pointed at plus some peripheral stuff, but your ears on the other hand can hear everything around as opposed to only in the direction they are pointed at (sides of head, that is).
Things your eyes can’t see is everything that’s generally behind say 140°. More about our field of view in this wiki article. But things our ears can’t hear are only those too far away, too high pitched (sounds pitched too low can still be felt as vibration) or too quiet. And I bet you’ll agree that none of these qualities are direction related.
So here’s a fact: while main qualities of our vision are directly direction related, qualities that describe our hearing are not.
Camera lenses and microphones function in a very similar way our eyes and ears do. Lenses allow to only “see” things they’re pointed at and everything that’s “out of frame” does not exist to the viewer.
However microphones can “hear” everything with exception of same things our ears can’t hear. That is once again:
- If a sound is too far away
- If a sound is pitched higher or lower than microphone’s actual frequency range
- If a sound is too low in volume (too quiet)
- If a sound we want to capture is obscured or overpowered by a much louder sound (this applies to our ears too – just remember last time you tried to talk to a friend during live rock band gig)
If we assume that unidirectional microphone is the one that captures “only” sounds dead up front, unidirectionality of a microphone doesn’t help to completely exclude / reject something at microphone’s back or sides like a camera lens would exclude everything that’s out of the frame. Total rejection is impossible because of one technical reason that made the world think “nice forward facing circle of focus” once unidirectional microphone is mentioned.
Why we understand directionality of microphones the way we do?
90% (or more) of microphone manufacturers build / draw a polar pattern of their products on the basis of very thin frequency range where most microphones have the best directional performance. In fact it is not even a frequency range rather than one single frequency of 1000Hz (1kHz). And how many frequencies do we have in microphones frequency spectrum? Theoretically about 20 000 and at least 16 000 that can be used in real life.
10% of manufacturers include some other frequencies to help out their customers a bit more. You can try to discern several outlines of frequencies here on datasheet of a Sennheiser MKH 416 (125Hz/250Hz/500Hz/1000Hz/2000Hz/5000Hz/8000Hz/10000Hz).
This information can help us interpret Sennheiser’s directional qualities more precisely. It’s clearly seen that this mic has quite deep “dead spots” at angles 120° to 150° measured @ 1kHz, but “dead spots” become less dead @ 125Hz at same angle while there are all sorts of shit going on in frequency range above 2kHz. Basically most microphones lose their “pretty” directionality below 200Hz and above 8kHz.
Mistake #1. Interpretation of directionality based on just one frequency.
Of course it is completely understandable that manufacturers choose to include only polar pattern responses measured @ 1kHz because it not only looks best on paper (probably there is some kind of agreement) but also justifies assigned directionality of the mic. Imagine if someone tried to graphically represent polar pattern with all those frequencies in 3D… It’s enough that it looks messy in a 2D simple diagram, right?
Thing is that by trying to make stuff prettier and more simple manufacturers allowed users (or should I say users didn’t care looking closer) to build a huge bubble of misunderstanding around directionality of microphones as most inexperienced users assume that say cardioid or unidirectional microphone will attenuate all frequencies the same amount it does @ 1kHz – thus they’ll get 100% silence behind the mic.
And “attenuation” is a strange word to use yet, because “rejection” is most often used to describe microphones performance off-axis which leads to a wrong assumption that rejection is attenuation by 100%.
Mistake #2. Sound rejection vs. attenuation.
Which brings us to another wrong assumption that microphone rejects sound. No, it does not. For example a unidirectional microphone will not reject but rather attenuate (reduce gain / volume / level) at it’s “blind spot” (180° that is).
Lets take a look at my beloved Sony PCM-D100 (you can find same picture few paragraphs lower). The pink curve represents by how much a certain frequency will be attenuated at 180° angle. While being cardioid or unidirectional microphone (combination of two mics in fact) it’ll give you about 20dB attenuation @ 1kHz and you get maximum attenuation of 30dB @ about 3kHz. With recorder’s dynamic range being equal to ~96dB you only get less than one third of attenuation @ range from 400Hz to 7kHz. Why should you call it rejection?
I am being so anal about this not because I don’t like the terms used to describe certain qualities, but because people take term “rejection” for granted and imagine they won’t get any signal at “blind spots”. Because to reject something means to not have it completely.
Bottom line – attenuation of 30dB is the most you can have on most microphones’ dead spot.
Mistake #3. Background noise reduction / rejection is not a thing.
Now once we’re clear about attenuation, let’s talk “rejection (again) of background noise”. Fact is unidirectional microphones and microphones in general are not that smart and quite ignorant. They don’t care if you consider something background noise and the other thing your main object you want to record.
They capture everything as sound waves or noise on basis of their polar patterns. Meaning a unidirectional microphone will only attenuate stuff incoming off-axis or being obscured by your main object in the frame.
By no means should you expect or even imagine a microphone being able to perform some kind of dedicated process to define useful and useless noise, based on your judgement.
The only thing defining microphones behavior is YOUR decision to point it at something on-axis (at the most sensitive point of mic’s polar pattern) with intention for it to attenuate everything that’s off-axis.
Mistake #4. Zooming and reach powers of a microphone do not exist.
Last thing to deny on today’s schedule is reach or zoom of a microphone. As you may already guessed this term is wrong as well, because microphone’s “reach” is fixed and predefined by it’s dynamic range and signal to noise ratio.
“Zooming” into something sound-wise means increasing microphone’s gain (or sensitivity) which first of all leads to “zoomed-in” background noise and second of all will eventually lead to increased thus audible self noise of the mic. By no means can we use terms “zoom” or “reach” of a microphone to describe it’s quality to dynamically emphasize something in relation to something else.
Illustration. If you have a talking person standing 10 meters away with crowd cheering behind, there is no way you can make that person sound more clear or loud in relation to the crowd without making the crowd louder. Because again a unidirectional microphone or any other microphone is indifferent to the hierarchy of sources arranged by importance by you.
Only way to increase separation is either to move the person, the mic or both closer together and further from the crowd.
Because of wrong parallels to an optical zoom of a lens there’s a belief a mic with certain polar patterns can zoom into a sound the same way a lens can; with a result being isolated person in front the crowd that is out of focus.
DOF on a wide angle lens is a great way to understand mic’s directionality for a filmmaker.
FOCUS! This term brings us to a helpful way for better understanding of microphones directional qualities trough a parallel of a microphones polar patterns vs. depth of field quality of a camera lens (not camera shot frame).
If you point your lens at a person and have a shallow DOF, it doesn’t mean everything that’s out of focus will be ignored – no – it’ll only get blurred / less clear.
Same with the mic. If you point it at something you’ll make that in focus or “on axis” of it’s unidirectionality, but everything else will be recorded as well just with less detail (detail being volume (dynamic information) and clarity (spectral information)) or “out-of focus”, but not “out-of-frame”.
Summary of facts we figured out:
- There are lots of pictures representing various polar patterns of microphones that are laying the base for misinterpretation because these are drawn only taking into considerations a thin range of frequency spectrum – most of the time 1kHz which happens to have the most off-axis rejection.
- In practice microphones are very similar to our ears (same as camera lens and frame is similar to our eyes). They can “hear” everything with exception of sounds that are (a) too far away, (b) too low in volume or too quiet, (c) to low or too high in pitch, (d) obscured or overpowered by a much louder sound.
- Polar pattern of a microphone may and will differ according to frequency. For example a standard shotgun microphone (say Sennheiser MKH-416) possesses super-cardioid qualities below 1kHz and lobar qualities above, while most likely (there is no official documentation) it gets nearly omnidirectional or extremely bidirectional at its extremes (the lowest lows (e.g.: 50Hz) and highest highs (16kHz).
- Polar pattern of a microphone tells you where its most sensitive parts are and where you can expect it to capture most natural and loudest sounding source. This is rather relative and calculated in relationship to same microphones “off-axis” or “out-of-frame” spots.
- Even in “darkest” dead spots a mic can achieve about -30dB attenuation of signal at very specific frequency or frequency range. And while 30dB difference is quite big, this is far from “rejection” as we’re interpreting it, thus it should be called and more importantly understood as attenuation (FYI: humans perceive a 6dB drop/boost in volume as twice as quiet/loud, so 30dB makes it 5 times quieter/louder).
- Microphone can “hear” no difference between “useful” sound source and background or ambient noise, thus no mic will reject or ignore something we consider “background noise”. So if someone is telling you that a unidirectional microphone or any other microphone rejects background noise she/he is wrong. It only attenuates sounds that are off-axis and if that sound (according to you) happened to be background noise, then it will be captured a bit quieter.
- Dynamic or adjustable “reach” or “zoom” of a microphone doesn’t exist. Its sensitivity is defined by dynamic range, signal to noise ratio (self noise) and polar pattern. All those things are fixed. So if you still want to “zoom” the mic into your on camera talent you’re proportionally “zooming” into every other sound around him/her as well. Oh, and you’re doing it with the help of pre-amp’s gain, which will eventually lead you to audible self noise (if the source is still too quiet) or clipping (if the background is still too loud).
- Directionality (correctly called “polar pattern”) of a microphone only tells us where the mic is most sensitive (in focus) as opposed to least sensitive (out of focus). In case of a traditionally interpreted unidirectional microphone most sensitive part is expected to be in front of it while there will be less sensitivity on its back side. That means if we put a mic on a DSLR, mount it on a tripod and walk around it talking, there will be no “dead-spot” or “out-of-frame action” for the microphone’s field of “hearing”. Only difference will be that once talked in front of the mic / camera (it depends on mic’s polar pattern, but pretend ~75° angle both left / right / above / below) the recorded source will sound louder and clearer (or fuller, or more natural) compared to everything that was recorded whilst talking from behind (source will be much quieter and will lack some tonal information).
Bottom line is – if you can hear something, the mic will definitely “hear” THAT and more. The question is only “how loud that something will be” as opposed to “will it be recorded at all”. This brings us to a fact that a unidirectional microphone or any microphone can and should not be compared to a camera lens because the mic does not have “out of frame” (as something being completely not there) quality, thus it can not just ignore something that is in its radius of reach (close, loud and clear enough).
Most accurate and understandable parallel would be depth of field quality of a lens.
And it would sound something like this: an unidirectional microphone is like an unidirectional camera lens that is able to achieve very shallow depth of field; it doesn’t completely key out stuff that is out of focus, it just blurs it and makes it less clear as opposed to things in focus.
And yes unidirectional camera lens sounds crazy and funny; exactly like unidirectional microphone does too. Because no microphone is unidirectional in a way you used to think before you read all of this.
Thanks for reading and I really hope you got some revelations on the subject. Or maybe you think I’m wrong because you had a different experience? Please feel free to let me know either way in the comments below. Questions? Go ahead! Cheers for now.