View Points: Objective, Subjective & Semantic
Varying takes on microphone evaluation and jargon

By Doug Jones

Well, this is a tough one. How does one make “duh!” last for 1,800 words? I hope I’m up to the task.

Congratulations, Jack! You’ve just discovered a phenomenon that everyone who has ever used a cardioid microphone has already realized since the dawn of audio. When you get close to a cardioid mic, it provides a bass boost. Wow!

And it is different for each mic? Astonishing! And you can use an equalizer to “flatten” it out? Amazing! And you made the story of this observation last for 1,900-plus words, and made it entertaining in the process. I am truly in the presence - or is it the proximity - of greatness!

First, let’s talk about the use of language. Readers, you might remember one of my favorite rants about what I call the “problem of audio”. I’ll recap here in case your memory (like mine) is somewhat colanderesque. (It’s hard to believe Bill Gates’s spell checker doesn’t like that word - seems perfectly fine to me.)

Anyway, the rant goes something like this: three domains (or contexts) are always in play during audio analysis. Yes, mixing is a form of analysis.

Step 1: Does it sound right? If yes, go back to Step 1.
If no, go to Step 2.

Step 2: Fix it - go back to Step 1.

The first domain is the objective, where we measure and generate numbers that relate to some physical attribute of something. In audio, this is response, or power or current or sound pressure - stuff you can measure. The next domain is the subjective, where we listen to things and form opinions about our perception. The subjective domain is what is going on inside our heads as we listen. The third domain is the semantic, the collection of words or symbols that we use to communicate to others what is going on inside our heads. The semantic domain, of course, is also used to communicate the objective. Words are powerful and important things.

The evaluation subjects lined up and ready to go.

The tricky part comes when you try to make links between these domains. For example consider the sentence “This mic is really flat - it sounds great!” If we take this sentence apart in light of the domains, we see an objective element - “flat” - that is a measurement of the response. Meanwhile, “it sounds” is a subjective evaluation that occurred inside someone’s head. Finally, the word “great,” which is (usually) positive, was chosen to represent that subjective reaction.

Now here is a very important point. There is no law that states that flat equals good sounding. This is a myth, especially when it comes to microphones. Of course, one of the interesting things about language (especially English) is that if I said “this mic is flat, it sounds great,” and if, while saying the word “great,” I rolled my eyes back into my head and gave a certain kind of grimace, the listener would know that “great” meant just the opposite. It’s good and useful to know the measured response of a mic. It is also good to know what they sound like. Be careful when you try to infer one from the other. It is harder than you think.

Most of the time in this column, Jack and I are arguing about the relationship between the objective and the subjective. There is a wealth of material there, and it is a useful debate. I would like to focus for a bit on the relationship between the semantic and the other two.

In the world we call “science,” there is a very carefully crafted connection between the objective and the semantic. A volt is a volt - not a watt, not an ohm. A meter is defined in terms of some absolute. It is not arbitrary. We absolutely must have a consistent use of language for science to work.

I’m currently reading a fascinating book given to me by my lovely wife called On the Shoulders of Giants: The Great Works of Physics and Astronomy, edited by Steven Hawking. It’s an anthology of writings from some of the greatest minds in physics - Copernicus, Galilei, Kepler, Newton and Einstein. The reason we can stand on the shoulders of these giants is because we can read the symbols that are the representations of the thoughts they had.

The only way I can pick up a paper written by Copernicus 500 years ago and understand what he was thinking (to the extent that my poor brain is up to the task) is because his work was translated into a set of symbols that I understand, and these symbols have not changed significantly over time. More importantly, the only way that Hawking can debate and expand on Einstein is because Hawking understands the symbols that Einstein used.

Prof. Jones’ TEF measurement of the SM57, made at a distance of one meter from the source.

By now you’re probably wondering what either Einstein or Hawking have to do with Jack. I won’t go there, except to note that if you’re going to live in the subjective world, as Jack (“everything is about the sound”) claims to, you still must have a consistent use of language to describe the subjective impressions going on in your head. Of course in the dream world of the subjective, words do not have concrete meanings. Or maybe it is more accurate to say that non-verbal cues can modify the meaning, as we’ve seen earlier with the word “great.”

And to the point: proximity effect and presence peak are not the same thing! Now I know that every discipline has its jargon, but even jargon has to be used consistently. A few years ago, my garage burned down. Some kid set a dumpster on fire, and it spread to the garage. When I got to the scene, the fire chief said “I think what we have here is a non-spontaneous incendiary event.” Huh? Some idiot just burned down my garage!

But as a fire-safety professional, he used the terminology of his trade. I’ll bet that this terminology is precise, too. My garage did not burn down because I left oily rags in a paper bag. If you’re in a high-pressure situation and call out to your A2 at the controls “Hey, soften out the presence peak a bit,” I’ll bet he reaches for 3 kHz. Unless, of course, you are rolling your eyes when you say it...

A graph of the response of the speaker as one gets closer to it.

Proximity effect is the very real boost that occurs when a directional microphone gets close to a sound source. It is not something that mic designers arbitrarily add to a mic, but rather, is a by-product of making a mic cardioid. Omnidirectional mics don’t have it. The boost is in the low-end, generally between, say 80 Hz and 250 Hz, and maybe lower, maybe higher. It is not an absolute. The boost increases the closer you get to the mic.

Some manufacturers try to engineer proximity effect out of (some of) their products. The Electro-Voice RE20 for example exhibits almost no proximity effect. Others, like Shure, chose to leave it in. Many users of the SM 57/58 like the proximity effect and use it creatively. But hey, it’s a free country.

If you want to equalize the bass boost out of a vocal mic, knock yourself out. But understand that your EQ setting will only work for one physical position of the mic. If you EQ for lip contact, the mic is going to sound real thin at four inches. I have it on pretty good authority that mic designers pick some distance from a source and “voice” (or equalize) the mic to some desired “contour” that they think will sell. If you work it closer than that, you will get a bass boost. If you work it further away, it will sound thin.

Presence peak is a boost that is usually engineered into a mic - on purpose. It is usually between 2 kHz and 8 kHz and is fairly gentle, maybe 3 dB to 6 dB. It makes mics sound nice. (Or not, depending on what floats your boat.) Presence peak is not a phenomenon that changes with distance from the source.

I decided that although the response of the venerable SM57 is well known, I would measure it and use it to illustrate what we are talking about. See how the low-end drops off? This measurement was made at one meter from the source. It will be a bit thin. That is the way it was voiced. Also notice the lovely presence peak. This is a big part of what makes this mic so successful.

I also took the time to try to measure the proximity effect of the SM57 (with screen, of course!) to see if the “proximity-effect-curve-gradient-anomaly-presence-peak-response-thingy” looked anything like Jack’s absence dip. This is not as easy as it seems. Many variables need to be controlled if one is to make a believable measurement.

SM57 (with windscreen) normalized to one meter and corrected by the B & K data.

The first graph shows an omnidirectional measurement mic. By definition (it’s an omni), it won’t have proximity effect. Since it is a mic used for measurement, it won’t have a presence peak either (but we’re not looking at that). I decided to only show data out to 1.5 kHz for clarity. This is a graph of the response of the speaker as you get closer to it. Each curve is differenced with the measurement made at one meter, to subtract out the speaker, then the curves are corrected for the inverse square law.

It is interesting that the .5 and .25-meter curve is virtually identical to the one-meter curve. (A perfect match would be a straight line at 0.) When we get down to .125 meter and closer, we are in the true near field of the speaker, and it starts to behave differently. We actually see increase in the low-end and some high-end loss.

The next graph also focuses on the SM57 (with screen again, naturally). This graph was made by first normalizing it to itself. Again, this is calling the one-meter response flat, and subtracting it from everything else. Then I subtracted the B&K data from it. What is left is the true proximity effect of the mic, within reason and the limits of my crude technique. I stopped at two centimeters, which is a bit less than one inch from the source. Jack was making his measurements (oops - observations) with lip contact. Remember when you are working that close to a source, the inverse square law is a very powerful tool. I would expect to see even more bass boost at lip contact.

So do my measurements coincide with his observations? To some extent they do. I also think that he was compensating for the monitor wedges and the room, so those XTA settings are “fixing” more than just the mic. I certainly agree with Jack and want to emphasize that you can’t take these XTA settings and use them. You gotta EQ your own rig in your own space.

In conclusion, yes, Virginia, there is a proximity effect. It is both measurable and audible. And you can EQ for it. Maybe in a future column I will tackle the question of why there is a proximity effect at all. Until then, keep on listening.

Douglas R. Jones is the Chair of the Sound Dept. at Columbia College in Chicago. Reach him at djones@livesoundint.com

March 2003 Live Sound International