With the advent of the PZM (Pressure Zone Microphone), sound engineers began to revise their ideas of what a microphone should look like and where it should be placed. PZMs once more revealed the importance of comb filter effects created by sound reflections from boundaries. This article deals with the PZM approach to that problem and offers hints on how to avoid comb filter effects when using 'traditional' microphones.

The basic theory of comb filter effects is explained in the December 83 issue of HSR: but basically the same sound signal is reflected off boundaries, is picked up by mics and arrives at the mixer several times, with a very, short delay in between. The addition of these delayed signals gives a combined signal that ultimately sounds bad and lacks clarity as part of one signal cancels out and/or enhances parts of the others. There are two typical situations that may cause comb filter problems:

Too Many Mics

Here the different sound travel paths to mics 1 and 2 translate into a short delay between the arrival of the two identical signals at the mics, causing comb filter effects and sound deterioration (Figure 1). Solution: fewer microphones. Have two back-up singers - remember Paul McCartney and George Harrison - and use one mic together. Your drum kit, too, may benefit from fewer mics. For vocal mic placement on stage, a good rule is: distance from mic to singer x 3 = distance from mic to next mic

Reflections

In this kind of situation, the sound travels both directly to the microphone and to any nearby surface (boundary). From there it is reflected back into the microphone where it will arrive a little bit later than, and merge with, the direct sound. This will cause comb filter problems.

Before turning to the PZM solution to this problem, Figure 2 shows the fast cure for any conventional microphone - omni or unidirectional. By making the distance d₁ very small compared to d₂ (and damping the reflecting surface, if possible), the reflected sound will be much lower in volume. Unidirectional mics also help achieve this since the reflected sound will hit them off-axis (sound pick-up rejection).

Figure 3 presents another solution, which may seem a bit strange at first. It is, however, a very good mic position for a fast recording set-up. With the mic placed on a hard, reflective floor, it's perfect for distant miking of a guitar amp; close to a wall you may pick up the entire room sound, eg. for reverberation purposes. Put the mic on a table and you'll get a perfect recording of a speaker sitting at it. By making the sound travel paths approximately equal, all delays of the picked up sound will be very short, hardly deteriorating the sound at all.

Pressure Zone Microphones

When PZM microphones first came out, the theory behind them sparked off some discussion. Reality proved that PZM devices were a perfect way of reducing boundary reflection problems.

The PZM integrates the primary boundary in its design and the plate, common to all PZM designs, represents the primary boundary. The transducer (a small omnidirectional condenser) does not sense the sound reflections off this surface, because it is either flush-mounted to the surface or within the so-called pressure zone ie. very close to the surface.

Now, what is meant by the primary boundary? It is simply the first reflecting surface, the piano lid, the nearest wall, the floor, whatever. It is important to realise that reflections from other, more distant, surfaces do reach the PZM. Actually, these reflections are very important to obtain a feel of the room size. If they are missing (due to overdamping), the sound is not only very dry, but also unpleasant to hear. Our brain is absolutely unaccustomed to a totally dry kind of sound. Entering an anechoic chamber (no reverb at all) will cause nausea to some people. That's partly why some people dislike the announcers on BBC Radio 4 who have very dry voices. In the USA most radio announcers on FM stations have a small amount of reverb added to their voice to create a more psychoacoustically pleasing sound for the listener.

PZMs have a hemispherical directional characteristic. Their performance can be improved, especially in the bass range, by mounting on a larger surface - a wall, the floor, or a plexiglass plate. The latter is often used when recording an acoustic guitar, especially classical, where the guitar player sits about one metre in front of this plate.

PZM is actually a registered trademark of Crown who first introduced the mic, but nowadays almost all microphone manufacturers make such models. It should be noted that they all sound fairly different. Some, like the Schoeps, feature a distinguished treble lift (up to 10dB).

Another interesting feature of PZMs is their long reach. They can be used at quite some distance from the sound source. On the other hand, one should bear in mind the wide acceptance angle which means undiminished pick-up of signals from the side. This wide acceptance angle together with the long reach presented a serious problem during the broadcasting of the Oscar award ceremony in 1982. While PZMs mounted on the floor kept the stage completely free of distracting mics and stands, the PZMs also picked up wonderfully the conversation of stage hands standing at the side of the stage!

PZMs are fine for miking up a grand piano. Usually two are employed. Placements include attaching it to the lid, thus extending the boundary; or about one inch above the piano strings, which will reduce sound spillage and feedback in a live concert. For maximum gain-before-feedback, mount the mics on the lid and then close it.

Famous Last Words

This, the last instalment of Using Microphones, brings to an end a series on microphone placement and choice, and recording techniques covering all major instruments one might come across when recording pop, rock or jazz. Looking through all these articles, it seems to me that I haven't left out any important issues - with the one exception of what to do if there are two microphones and just one input.

Figure 4 shows how to connect the two mics in series and Figure 5 shows parallel wiring, which is technically better. The four resistors should equal the value of the mixer's input impedance. Try 4K7 (4.7KQ), but be sure you get the polarity right.