Recording - April 2015 - 50

Things get messy
What's this "supposed to"? Well, all the polar patterns I showed
you were plotted at a middle frequency, 1 kHz. At other frequencies, particularly high ones, the patterns change.
Look at Figure 8; it's the polar pattern of the same omni microphone as
in Figure 7, but this picture shows what the mic does at 8 kHz. It's still omnidirectional, sort of, but the response from the back (180°) is down 6 dB.

Figs. 4-7: Polar pattern examples: cardioid (4), hypercardioid (5), figure-8 (6), and omni (7).
"heart"). At 0°, which is the front of the microphone, the curve
is over the circle labeled "0". That's the reference point.
The other circles indicate different levels; for example, take
a look at the 90° mark on the outside. At this angle the curve
is just inside the "5" circle, meaning the mic's response is a
little more that 5 decibels (dB) lower at 90° than it is at 0° (the
front). This means that an instrument whose sound hits the
microphone at a right angle to the way it's pointed (90°) will
generate a signal that's 6 dB lower than it would be if the
instrument was in front of the mic.
Now look at the 180° marking on the outside of the circle.
At this angle (sound coming from the back of the microphone)
the curve is well inside the circle marked "20", at a point corresponding to 25 dB. This means that sound sources directly
behind the mic will generate a signal 25 dB lower than they
would if they were in front. Engineers call this the microphone's null, or its angle of maximum rejection. Another name
for a cardioid is a unidirectional mic, and you can see why.
Mics with cardioid patterns are probably the most common
in studios, but they're certainly not the only ones used. Figure
5 shows another sort of unidirectional mic, called a hypercardioid (a Neumann KM185) Now the rejection at the sides
is greater; at 90° the signal produced is 9dB less than it
would be if the source was in front of the mic. That's 3 dB better rejection at the side than a cardioid mic, but the hypercardioid has its nulls at about 135° and 225°; at 180° there's
a lobe of non-rejection-a source at the back of the mic will
only be rejected by about 10 dB, rather than 25 dB.
Choosing a hypercardioid over a cardioid is a tradeoff; its
rejection of sounds from the back is poorer, but it rejects better from the side.
Another pickup pattern used in studios is the figure-8, or
bidirectional pattern (Figure 6; this is a Neumann U 87, set
to bidirectional mode). Looking at it, you can see how it was
named; it picks up equally well from the front and the back;
the nulls at the sides, on a good microphone, are very deep.
If you go way back to A Hard Day's Night, you'll see John
and Paul singing into two sides of a figure-8 microphone.
One oddity is that the front and back lobes have opposite
polarity; when the air pressure goes up, the signal voltage will
go up if the source is in front of the microphone, but down if
the source is in the back. We'll ignore that for now.
Most ribbon microphones are inherently figure-8, though it's
possible to engineer them with other patterns-one of the mics
I brought in my reserve kit was a Beyer hypercardioid ribbon.
Last but not least, omnidirectional microphones pick up signal from all directions equally, or at least they're supposed to
(Figure 7, a Neumann KM 183).
50

RECORDING April 2015

Part of what's happening is that the body of the mic casts an acoustic
shadow at higher frequencies, where the sounds' wavelengths are
short and comparable to the size of the microphone. This is why many
mics are designed as side-address: the capsule is mounted at right
angles to the mic body, so shadowing happens only at the angle the
body is pointing. Other factors that mess up the response kick in when
very short, high-frequency soundwaves hit the mic at an angle.
Polar patterns are often drawn like Figure 9, which shows the
behavior of the Microtech Gefell M930 mic I used on Will Soll's
voice and tenor banjo. The solid line shows the pickup pattern at
1 kHz; the dotted line shows the pattern at 8 kHz, while the dotand-dash line shows what's happening at 16 kHz.
Is this because these mics are crummy? Emphatically not; Neumann
and Microtech Gefell mics are among the most carefully designed
and built in the world. The deviations from perfection are, instead,
caused by the laws of physics-and those are hard to dodge.
A manifestation of those pesky laws is that smaller-diameter capsules have directional patterns that are more uniform with frequency. This is why measurement microphones often have tiny capsules
(as small as 1/4" in diameter). In recent years a few manufacturers
like Earthworks have introduced studio mics with similarly small
capsules, in search of similarly uniform pickup patterns.

Figs. 8-9: Frequency dependence of polar patterns. An omni
pattern at 8 kHz shows directionality in Figure 8, and Figure 9
shows how one mic's polar pattern changes with frequency.
Next month we'll look at microphones' frequency responses;
they make a huge difference in the way different mics sound, as
you'd imagine. Responses can be specified in several ways-some
useful, some not. Stay tuned!
Paul J. Stamler (stamler@recordingmag.com) is a recording musician, engineer, radio show host, educator, and collector of vintage
recordings, living in St. Louis. Figures 4-8 courtesy Neumann;
Figure 9 courtesy Microtech Gefell.

Recording - April 2015

Table of Contents for the Digital Edition of Recording - April 2015