Microphones … why so subjective?
April 26, 2020
In live sound it all starts with the microphone. If the microphone does not do its job then the input signal to our system is corrupted from the very get go. Using modern signal processing some of the corruption can be corrected for, but not all. So while it is not as true as it once was: “garbage in garbage out” is still mostly true.
Recently I went to purchase a microphone. The data sheet for one of the “pro” mics I was looking at was missing a specification: the frequency response curve, and so I wrote the manufacturer. They replied that this information was proprietary. The most fundamental specification for a mic is proprietary? Really! Contrast this to the data sheet for a very inexpensive MEMs microphone used in smartphones. The data sheet gives pages of specifications.
It seems for both pro audio and audiophiles we rely to a very large extent on very subjective criteria and not detailed reproducible specifications. I will grant you the performance of a microphone is not totally described by measurable specifications, any more than the overall performance of a car can be. Even if we can measure everything (and in theory perhaps we can) a data sheet having many many pages of numbers and graphs is difficult to fully interpret. Ultimately we are interested in: “How does it sound?” and not really what is the total harmonic distortion at 3.45K Hz.
My primary problem with the subjective approach to select pro audio equipment is that, in running a small business, I don’t have the money to purchase 10 or 20 different types of microphones and even if I did, I don’t have the time to evaluate all these mics in the various environments I am going to use them. What I need is enough objective specifications that I know that any microphone I purchase will do the job. Maybe what I purchased is not the very most optimal but I need to know that its performance is far more than good enough.
OK, unlike my example above, most microphone manufacturers do give you some typical specification, usually at the bottom of a multipage data sheet where the first few pages have many pretty pictures and are mostly marketing propaganda. These specs include a frequency curve and a spatial (polar) plot at a number of different frequencies. Normally you also get the mics sensitivity. Since for most mics, the sensitivity is pretty much the same, this is really not a very interesting number. Other specifications often given are: the maximum sound pressure and self noise floor in terms of equivalent sound pressure and the maximum recommended sound pressure. With a very few exceptions, the way cardioid and super cardioid microphones are built leads to a proximity effect. This is particularly true for dynamic microphones. Simply stated, as the mic gets close to the sound source the bass response of the mic increases. Often, but certainly not always, a microphone manufacturer will add this information to the frequency response data graph.
Until I retired my day job was a (mostly analog) electrical design engineer. When you look at a data sheet for say an op amp, the data given by most semiconductors companies is pretty much the same set of specifications and much of it is presented graphically. For “critical” specifications you get both typical and worse case specifications. It would be nice if microphone manufacturers would also give you a moreless standard set of specifications so you could evaluate different microphones before you pluck down $100 to $1,000 (and sometimes even more) to purchase one. Here is my proposed “standard” and expanded set of specifications:
-
Frequency plots (20 to 20,000 Hz)
-
not just at one dynamic level but at different sound pressures. Say 60 through 120 dB at 10 dB intervals.
-
Phase response … ideally a microphone should be phase linear.
-
Proximity effect at 1”, 3”, 6”, and 12”
-
-
Spacial (Polar) plots
-
Usually you do get a graph at a number of different frequencies over the audio range. This is good.
-
Polar plots off the horizontal. Microphones get tipped. What does the polar response look like at 30 deg and 45 deg off axis?
-
A high fidelity angle (even for omni mics as omni mics are not truly omni-directional). This is the angle that all audio frequencies are the same to mithin 3dB (normalized)
-
-
Dynamic Range
-
Self Noise Floor
-
Upper Sound pressure distortion limit … the sound pressure whereby the mic has an additional 6 dB of distortion
-
Maximum Sound Pressure limit … the sound pressure whereby the mic is damaged.
-
Impulse response (time domain) at 94 dB, and at the sound pressure distortion limit. This is particularly important for “percussion mics”
-
-
Total harmonic distortion (THD)... curves showing THD over the entire frequency range and at different sound pressures, again at 94 dB and the sound pressure distortion limit.
-
Miscellaneous:
-
Output impedance
-
Magnetic shielding (for dynamic mics)
-
Lower limit for phantom voltage (for condenser and ribbon mics)
-
Handling noise … we need a standard test for this but I am sure Shure already has one
-
There is a web site named “microphone test kitchen” that allows you to evaluate 300 different microphones. Perhaps they have expanded the selection since I last looked. They test mics under very ideal and reproducible conditions. What you find is almost any pro quality microphone used in near ideal conditions does an excellent job. It is when (as in real life live sound) the mic is used in less than ideal conditions the differences start to show. Thus, many of my proposed specifications are about these real conditions such as how the mic responds when the performer has it held at a 60 deg angle?
There is admittedly a problem with this highly objective approach. It does not allow for what I will call sound color or sound spice. No microphone is perfect, they all add some distortion. For some “vintage” mics we have come to desire their particular distortions. From the performers point of view, some less than perfect mic adds a distortion that enhances the desired sound. In theory, using modern signal processing, the desired sound spice (distortion) can be added to the mix (mostly). But even here there is an objective approach: the microphone manufacturer defines the desired distortion and this becomes the reference against how the microphone is evaluated.
In analog electronics, and in live sound, there is much science that is very objective in nature, but there remains a bit of “black magic” that is highly subjective. The proof is in the pudding. My objective approach will not select the very best microphone for some particular event, but it will eliminate the “dogs” and help you steer clear of the Vebin effect … purchasing something that is very expensive simply because it has a certain aura about it. This approach is utilitarian, allowing you to select microphones that are more than “good enough”.