SENSITIVITY

OK, first up this time, sensitivity. This is a very misunderstood specification, but it's really very simple. It's usually stated in units called dBV/Pa, One Pa is defined as a sound with the volume equivalent of a 1kHz tone that measures a volume of 94dBSPL at the microphone capsule.

How loud is that? What, you STILL haven't bought that SPL meter I told you to buy? All right, it's pretty loud, but fairly normal for a normal singer at a normal distance from the mic. Not a whispery pop vocal, not an opera singer, just your average vocalist. With that input, the mic's voltage output will be the stated level in dBV.

Let's assume you've read through Glen's interactive online applet on gain staging, so you already know you want your mic preamp to boost the microphone's signal to somewhere around line level. With the sensitivity specification in hand, a little simple math is all you need to determine the amount of preamp gain you'll need to do so.

Let's say you're using that SM81 on a 94dBSPL source, and you're shooting for 0dBV so as not to tax your converters. The SM81 sensitivity spec is -45dBV/Pa, kinda quiet for a condenser. So you would need a preamp that can cleanly deliver at least 45dB of gain.

Some manufacturers expect you to be able to perform higher math functions in your head, and annoyingly state sensitivity in mV (millivolts) instead of dBV/pa. Use this formula to make the conversion from mV to dBV and to compare two mics that use the two different types of sensitivity measurements: dBV = 20 * log (mV).

There is one big misconception that many folks have about sensitivity; that there is a relation between sensitivity and off-axis response. We often hear, for example, that large-diaphragm condenser mics (most of which tend to have higher sensitivity specs) "pull in" more of the room because it has a greater off-axis response. In fact, the mic sensitivity spec has nothing to do with off-axis response. Nothing. The polar response charts we talked about in Part 1 tell us all want to know about off-axis response.

The mic's capsule innocently transduces what it hears according to its frequency response and polar pattern. The sensitivity spec simply defines the voltage levels into which the sounds that it hears are converted. The higher the sensitivity rating, the “louder” the mic’s output.

The second big misconception about sensitivity is that it affects or is related to the mic's noise specification. It doesn't and it isn't. In order to understand that, we first need to talk about noise specifications a bit more.



SELF-NOISE

All mics have noise, or self-noise; i.e. noise created by the the mic itself. Usually you'll only see a noise spec for a condenser mic. That's because a dynamic mic's self-noise is rather low, so nobody worries about it too much. (As long as the source is loud enough that it doesn't push the limits of the preamp, anyway.) So let's talk condenser mic self-noise spec, since this is usually the next spec we'll see most often.

Self-noise is most often stated in “dBSPL equivalent, A-weighted”; or dBA for short.

The SPL equivalent part simply means that if you're recording, say, a 94dBSPL source (as specified in the sensitivity spec), then the mic will contribute an amount of noise to the signal equivalent to what room noise would contribute were it at that SPL volume level. Still using that SM81 and an example, it's Self-noise is rated at 16dBA, meaning that the amount of noise in the mic itself is the equivalent of what we'd get from 16dB of room noise.

What is A-weighting? It means you pretty much ignore low-frequency noise, and high-frequency noise above 16kHz. Your ears are very insensitive to noise at those extremes, but very sensitive to noise in the A-weighted range. Or, to put it another way, A-weighting represents a frequency response curve that very roughly mimics the response of the human ear.

(Some people hyperventilate about A-weighting, and claim that only unweighted specs are honest. To paraphrase William Shatner, those people need to try kissing a girl sometime. So long as everybody uses the same A-weighting curve - and it is a published standard - all microphone specs should be comparable. And if somebody is going to be dishonest, well, it's just as easy to lie about an unweighted spec.)



SIGNAL TO NOISE RATIO (S/N)

The signal to noise ratio (S/N) of the mic is simply sensitivity minus the self-noise. For the SM81, that yields a S/N of 78dB (94dBSPL minus 16dBSPL equiv.)

Note that signal to noise spec is often printed in lieu of the self-noise spec so you don't have to consider the mic's sensitivity when evaluating noise. A mic that is 10dB more sensitive, but with the same noise rating, will have the same amount of noise on your final recording once you match gain at your preamp.

Is there a correlation between noise and sensitivity? Yes, there is. Larger diaphragm capsules, by virtue of their size, have a hotter signal, which keeps the signal farther above the noise of the electronics (and the capsule itself). But don't take that as a firm rule, because it's easy for a small diaphragm mic to use a gain stage in its internal circuitry that boosts its output--but the noise gets amplified along with the signal.

The bottom line underneath all this is that if you are recording a very quiet source, look for a mic with a very low noise rating, NOT a mic with high sensitivity (although that doesn't hurt, so long as you mind the noise!)



MAXIMUM SPL

On the other end of the loudness spectrum, we have maximum SPL ratings. I’m reminded of that old Calvin & Hobbes cartoon where they figure out how much weight a bridge can carry by piling on the weight until the bridge can no longer hold it and it collapses. I’d love to be able to extrapolate that to describing the maximum SPL rating as being determined by simply bombarding the mic with louder and louder noises until it catches on fire and explodes. But that’s really a bit of an exaggeration, though some people believe the mic will be permanently damaged at that rating.

Not quite that extreme or fun, though, the maximum SPL rating is simply a measure of distortion. In a condenser mic, that's an overload of the internal electronics, generating distortion at either 0.5% or 1% (manufacturers vary in their selection of the threshold). Nothing will get damaged.

Is the mic "clipped" at that level, the way a digital converter would be? That depends. Usually not; usually distortion increases gradually, so even 6dB above that rating on a fast transient might not be noticeable. But don't push it; those ratings are there for a reason - the preservation of a nice clean recording.



DYNAMIC RANGE

Similar to signal to noise ratio, the manufacturer might do some more subtraction for you, and state the difference between the mic's maximum SPL rating and self-noise as the mic’s dynamic range. Put simply, this is the range between the loudest sound the mic can hear without distorting, and the quietest sound it can capture that won't get drenched in its own noise.

Directly affecting the mic’s maximum SPL level, and therefore also it’s dynamic range, is the mic's internal pad, if it has that feature. When switched on, the pad will reduce the capsule’s output level before it hits the internal circuit. The maximum SPL rating will increase accordingly, but so will self-noise. For example, if you have a mic that boasts a –10dB pad, that means that when engaged, the mic’s maximum SPL level will increase by 10dB. But it will also at the same time push the signal 10 dB down closer to the mic’s self-noise level, effectively raising that by 10dB as well. You shouldn't care about the noise too much, because you should only ever use the pad when your problem is having to mic an extremely loud source. Just make sure to switch it off before you use the same mic on acoustic guitar!



OUTPUT IMPEDANCE

Getting down to the dregs here . . . let's talk about output impedance. Impedance is massively misunderstood, and will be the topic of a future column. Neophytes often hear--perhaps from very old-timers or maybe ham radio operators--that you want to match impedance. In this case we’re talking about mic output impedance and preamp input impedance, and here, no, you probably don't!

Suffice to say that for any condenser mic, you want to bridge impedance, not match impedance. For those who are Ohm’s Law-challenged, this simply means you want the preamplifier to be of much higher input impedance than the mic’s output impedance in order to maximize the transfer of voltage from one to the other. So if a mic has a 150 ohm output impedance, you typically WANT to use a preamp with 1K ohm or higher input impedance. With dynamic mics, you can get different sounds by "loading down" the mic with a matching impedance, but this is very optional.

In most cases you shouldn’t need to worry about it. Professional mics are designed with appropriately low output impedance, and professional preamps have appropriately high input impedance. This is one area that truly should be "plug and play" for most of us. So plug it in . . . and press record!