Cardioid Carrying Member
The March’03 Tech’s Files
2003 by Eddie Ciletti
Management is discussed is some detail elsewhere on this site, not
just for Surround Mixing, but a more global zoom-out of four low-frequency
perception variables — the Equal Loudness Curve for humans (1), the Loudness
Button on consumer gear (2), Acoustics (3) and the Proximity Effect of
Directional Microphones (4). The latter is this month’s topic, but first
a quick review.
THROWING A CURVE
The Equal Loudness Curve (1) details
the ear’s unequal sensitivity to the audio spectrum at various levels and
insensitivity to Bass at low monitoring levels in particular. A digital
Sound Pressure Level (SPL) meter —about $50 at Radio Shack — is the first
step toward achieving consistent monitoring levels. With it, I learned
that my comfort level is around 70dB-SPL, 15dB lower than the 85dB-SPL
reference level specified for surround and other sound-for-picture facilities.
Depending on the version of Loudness Curve being interpreted, suffice to
say that for me, a Low Frequency (LF) boost is in order of not less than
6dB and possibly 9dB to 12dB. If not addressed in the monitoring system,
it might translate into that much more bottom in the mix.
NOTE: A $50 SPL meter is not
expected to be an Absolute reference to NIST standards — www.nist.gov
— but it can provide the end user with a Relative indication
of monitoring level. Choosing and maintaining a target SPL is part of the
ear training process — how things sound at that level as well as what type
of sounds translate best to the consumer environment.
The Loudness Button (2) on consumer
gear boosts bottom. On some car stereos there is no button although the
curve is not always part of the equation — at the maximum "Loudness Control"
setting, for example, there is no curve. In my own private Utopia this
would be calibrate-able to the 85dB reference. Back in the real world where
there is no standard, the Loudness Curve is gradually applied somewhere
between 50% and 75% of the control’s range — more of the curve is added
as the control level is decreased. Now the cranial light bulb should pop
on and be white hot. Dig it! The curve is a moving target, so the relative
level of the sound source affects the amount of EQ applied to it. A super
hot, ultra-normalized CD gets more of the Loudness curve than one with
more dynamic range and less dynamics processing.
THE CENTRAL SCRUTINIZER
Acoustics (3) and the laws of physics
dictate that smaller control rooms will have more low frequency issues,
problems that can be tamed to a degree but not eliminated. My theory is
that the 1-2-3 combo — of Loudness Curve, inconsistent monitoring levels
and unresolved acoustic issues — conspire to reduce the perceived Low Frequency
Energy (LFE). Even before the listener becomes aware of this deficiency,
along comes the directional microphone. Used up close and personal (to
maximize isolation), Cardioid and Figure-of-8 Polar Patterns have a Proximity
Effect — a noticeable rise in LFE as the mic is moved closer to the source
— a bump that make things sound "right" and "warm" at least for the moment.
Furthermore, a multi-channel recording
— where many directional mics have been used at close range — may suffer
from a low-frequency "muck" build up. Better to fix the muck than compensate
by boosting in the 2kHz to 5kHz range where the ear is most sensitive.
Heightened awareness of these four variables— any one of which might be
your Achilles Heal — should make them easier to tame, hence this month’s
continuing zoom-zoom-zoom on the boom-boom-boom. This is My Personal Theory
currently under investigation. Care to join me?
MAPS and CHARTS
If you’ve never paid much attention
to the Frequency Response of a Microphone, now’s your chance. More than
any single piece of audio gear, mic and speaker (transducer) idiosyncrasies
are the most easily identifiable by ear. You should be intimately familiar
with at least one of the seven microphones represented in Fig-1
through Fig-6. There are all sorts of bumps and lumps — electronics
by comparison are almost indistinguishably flat. The charts were downloaded
from the web, then resized for equal amplitude (vertical) and frequency
(horizontal). The editors were especially kind this month, allowing all
six pix to be published. If you need a closer look, they are already posted
at www.tangible-technology.com by clicking on "recent articles."
To generate their published "flat"
response curves, microphones are typically measured in the lab at a distance
of 1-meter (39.37 inches) from the source. For directional mics, any distance
closer to the source yields increasingly more "bottom," hence the proximity
Effect. (Omni mics are not affected by Proximity.) Knowing that their products
will be used up close and personal, directional microphone designers may
incorporate a bass roll-off feature either by switch or by default, the
latter can be parsed from the 1-meter response although detailed Proximity
Curves are preferred. Three of the mics investigated here have published
proximity response, detailing that all the action is at 1-foot or less
—reinforcing an oft-suggested phrase to "move the mic first " rather than
touch the EQ.
Manufacturers are not consistent
in their presentation of data, both in graphic and text form, although
the Internet is much easier to update than reams of product literature.
I suggest gathering as much info on the mics you use the most — even going
so far as to request the proximity curves — because this knowledge will
assist your intuitive sense. On the net, I learned that the roll-off of
the AKG C-12VR is 6dB/octave at 100 Hz and 12 dB/octave at 130 Hz) while
the Neumann TLM-170 roll-off circuit "attenuates the frequency response
below 100 Hz to suppress undesired structure borne noise." Another example
is the Earthworks SR-77, which is "flat" at 15cm (6-inches) so obviously
the response at 1-meter shows a substantial roll-off. All are helpful bits
of information, but the addition of proximity curves can help identify
a problem and determine whether the best solution is the built-in filter
or external EQ.
I had originally planned to start
in familiar territory with the Shure SM-57 but instead chose the Beta-57
because it provided proximity curves at 2-feet, 2-inches, 1-inch and 1/8-inch
<< For Figure-1 use image
Designed as a vocal mic, these most
important curves reflect the typical response when used under "normal conditions,
the "eating" style favored by rock performers at 2-inches or less. (I love
it when I see vocalists actually work the mic, but that doesn’t happen
often.) At 2-feet the response is down almost 3dB at 200Hz, but at 2-inches
the response between 100Hz and 200Hz increases 7dB to 10dB or about equal
to the hyped response in the presence region (5kHz). An inch closer adds
almost 5dB of warmth, that’s the Inverse Square law working its magic!
If not already instinctively ingrained, this information should be helpful
to recording and live sound engineers alike.
<< For Figure-2, use image
The now vintage Sennheiser e-609
(a new version with midrange bump just released at NAMM 2003) and the venerable
MD-421 do their chart dance together in Figure-2. At the top left,
the e-609’s proximity effect at 5-cm (1.96 inches) yields a 10dB boost
centered around 125Hz. Below, the MD-421 shows the results of its five-position
Music-Voice switch from "flat" to 16dB roll-off at 100Hz. If like me the
MD-421 is often your kick drum mic, that 10dB upper-midrange bump is most
likely the reason why.
<< For Figure-3, use image
Figure-3 details the on- and
off-axis response of an Electrovoice RE-20 (top and bottom, respectively).
The off-axis (leakage) response greatly affects the perceived sound of
a mic. The RE-20 rejection response is particularly smooth. DPA (formerly
B&K) provided the widest selection of graphs for their mics including
a highly detailed look at just the Proximity Response of their 4011 Cardioid
PLEASE NOTE: Figure-4 is two graphs
in one — to the left are the effects of Proximity from 1-meter to .1-meter
in five steps. To the right of 1kHz I have appended the graph with the
effect of cable length on maximum output level (148dB-SPL). This is not
the response of the mic under normal use even with these cable lengths.
The DPA-4001’s low frequency response
seems to be optimized at a distance of 12-inches where it is flat down
to 20 Hz. Just under 4-inches from the source, the boost at 50 Hz (and
below) is about 12dB. Wow!
<< For Figure-4, use image
<< For Figure-5, use image
Next are the large-diaphragm condensers.
Figure-5 shows the response for a Neumann TLM-170. The roll-off
is about 6dB per octave starting at 160Hz. Figure-6 show the response
of three AKG CK-12 capsules as re-skinned by Walker Microphones in Canada
(519-654-0070) mounted on a C-12 (vacuum tube electronics) body. The CK-12
chart is useful for several reasons. First, it is not a "publicity photo"
but rather the raw output from a measurement system. All of the other response
charts are smoothed either purposefully via system options or by a graphic
artist. Smoothing is fine for depicting the overall character of the mic.
Figure-6 shows subtle variations in three capsules, barely obvious
except in the bottom two octaves. In this case the chart was used to determine
the best match for a stereo pair. ALL microphones have production tolerances.
<< For Figure-6, use image
THE LOW DOWN
Studying these charts made each mic
more tangible — mics I have used and abused ad infinitum. From this research
I learned that many directional mics are designed to be "flat" not at one
meter but at unique distances, predetermined by their "typical" application.
I hope this article shed some light on the power of Proximity — knowing
where "flat" is and from there, realizing the dramatic difference plus
or minus four inches can make.
It was equally interesting — surprising
in some cases — to discover each mic’s unique characteristics at the opposite
end of the spectrum. We all know how different models respond to similar
applications, we all have favorites and now the "why" is a little more
apparent. Like the proverbial squeaky wheel, it’s the odd track that gets
our scrutiny, so being familiar with microphone characteristics goes a
long way toward reinforcing an engineer’s intuition at the moment of capture.
In our business there are more Relatives
than Absolutes. Whether tracking, overdubbing or mixing, the primary job
is to make each channel relate to the whole almost regardless
of the monitoring system being used. If the process of making the puzzle
pieces fit becomes a struggle — particularly when the bass needs more room
— it’s the monitoring system that gets the scrutiny. Then we want absolute
confirmation of the truth, but in the end, it’s all relative. Isn’t that
something Stephen St. Croix would say? I meant to say that the final judgement
rests with how it plays in Peoria.
Eddie researches, repairs and consults
from the afternoon to well past midnight. From dawn till noon he can be
found "engineering" small wooden railroad systems, applying conflict management
techniques to toddlers as well as attending to their "Input / Output" detail.