Recording - March 2017 - 52

the past year, I've been called in to work on a number of rooms to ﬁx
BY BRUCE BLACK
low-frequency deﬁciency problems. The rooms' owners reported that
frequency response measurements on a realtime analyzer (RTA) indicated all was well, but to the ear there was a
huge gap in the LF response that can only be described as The Great Bass Suck-Out.
After examining the rooms, taking measurements on these rooms and pondering the data, I found the answer,
which sounds like the opening from Star Trek or something...

p6KOGVJGƂPCNHTQPVKGTq
In order to make sense of this, it's important to fully understand
what our current tool is telling us, and more importantly, what it
is not.
A realtime analyzer displays the spectrum of whatever sound
is taken in at any given moment. In the case of a room, it is the
total of the sound collected by the microphone. That would be the
initial direct sound, plus all the reverberant sound still bouncing
around the room. Fair enough. However,
because the direct sound has the power of
the speaker behind it, and the reverberant
sound has been attenuated by ongoing
reflections, absorption, diffusion (if you're
lucky), and so forth, the direct sound is the
strongest and is the dominant player on the
frequency response display.
As hinted at above, the effects of the
room aren't able to fully influence the direct sound. After the direct sound arrives
at the ear or mic, everything else arrives
later in the form of early reflections, later
reflections, and reverberation. In my experience, most of us overlook the effects of
time... and like so many other things in
life, time can take its toll.

reflections can cause problems-you have the direct sound
and a delayed, slightly attenuated sound combining at your
ear or mic capsule, potentially causing phase cancellations
or additions.
This is vitally important, because when you look at the data,
specifically the decay curves below, you see that a broadband
low frequency dip appears as time progresses, punctuated only
by the room resonances. From around 110 Hz down to below
10 Hz, the low frequency response drops
20 dB on average, as if there was a
highpass filter in the audio chain. Do you
think you can hear that? I'd say it'd be
hard to miss!
So it's real, and not some oddity or artifact of the data... but what is it, exactly?

In my experience,
most of us overlook
the effects of time...
and like some other
things in life, time
can take its toll.

6KOGVTCXGN
This is where and how time comes into play. Think of a direct
sound like a dog clicker or a clock tick-a quick impulse with
a really fast attack and decay. With each millisecond of time,
that impulse travels 1.126 ft. (in standardized air pressure,
temperature, etc).
Eventually that impulse will come to a wall and reflect. The
time it takes is determined by the distance. This is why early
52

RECORDING MARCH 2017

5QOGUNGWVJKPI

When trying to figure out acoustical
problems, I start from this: everything
counts. So I have to look at the room and
see just what is in there.
Here are a couple of tasty tidbits as a
starting point. Cancellations from phase
interference cause very sharp dips, not
broad filtering like this, so that's not the cause. Also, in 80 milliseconds, sound travels around 90 feet, so in a 15 x 8 x 12 foot
room, the sound has reflected perhaps as many as ten times to
travel that far, attenuating each time, so it's getting pretty weak.
As I'm surveying the room, I see that it is a drywalled room,
likely built by a residential builder. So that means the wall wasn't
carefully framed with the studs exactly the same distance apart.
It's a single layer of 1⁄2 in. drywall, 5/8 in. if we're lucky. The
drywall tradesperson used enough screws to hold each panel in
place, hopefully to code, but not particularly rigid. And if the

Table of Contents for the Digital Edition of Recording - March 2017