|
Acoustical Shrinkage of Large Rooms into Small Rooms
by
Stephen Harner
Honey, I shrunk the room!
The physics of sound is well understood and the equations that
describe the propagation of sound can be derived from the
fundamental nature of the molecules of air. The equations of sound
can then be used to calculate various parameters for a given set
of boundary conditions, i.e. volume, surface area, etc., of a room.
Unfortunately, there aren't many widely accepted parameters that
acousticians agree will consistently improve the acoustics of small
rooms. However, there are now several widely accepted parameters
that can be adjusted to improve concert hall acoustics. In the
following sections, I will discuss the six most widely accepted
parameters for concert halls (large rooms), and then apply them to
small rooms by extrapolation, or so called acoustical shrinkage.
LOUDNESS
Loudness, as applied to concert halls, really means the relative
perceived increase in loudness due to the reflected sound field
over the loudness that naturally comes directly from the source.
In a large concert hall, not being able to hear the quiet passages,
(i.e. soft violin notes) would be bad for the overall sound quality.
Since small rooms require listening in close proximity to the source,
loudness is not really a factor.
Initial Time Delay Gap - ITDiG
The Initial Time Delay Gap, which I call ITDiG (it dig, you dig?),
is the time between the arrival of the direct sound and the arrival
of the first reflected sound.
Beranek says that the ITDiG should be less than 20 milliseconds for
optimum quality of concert hall sound, while longer values have a
negative effect.
Small rooms will generally always have an ITDiG less than 20
milliseconds, and typically will be around 2 milliseconds, unless
the sound source is placed within 1 foot of the wall (ignoring
the low level edge diffraction effects).
Advocates of live end dead end theory say that early reflections
"smear" or "blur" the direct sound, and that eliminating the early
reflections improves the sound. This involves treating the front
half of the room with acoustic absorbers. Others say to use a
mirror moved around on the wall to locate the speaker reflections as
viewed from the listening position, then treat these spots with
acoustic absorbers. Both of these methods are akin to lengthening the
ITDiG.
Does too short an ITDiG make for bad sound, or does it improve
the sound by adding width and fullness (like reverberation)?
Very short ITDiGs, less than 1 millisecond, do affect the way we
localize sound sources. Hafter's Auditory Perception Lab has
shown that when the left ear is fed an impulse incrementally
increasing from 0.1 to 1.0 milliseconds before or after the right ear,
the brain hears the sound as coming from further and further left or
right, depending on which ear is fed the first impulse. Certainly,
if the 2 impulses smeared each other, then the brain would not be
able to discern them. The fact that the ear/brain auditory system can
discern impulse signals spaced apart by only 0.1 milliseconds is
almost unbelievable!
If the second sound came from the same direction as the direct sound,
and at the same time, then some phase cancellation would occur
(generally referred to as comb filtering) causing peaks and valleys
in the frequency response. In fact, this can be heard
when 2 speakers are placed in close proximity and fed the same signal.
But the direction and timing of reflected sounds are typically
different from the direct sound, and there is no conclusive evidence
that people actually hear comb filtering when speakers are placed
close to walls. It is widely accepted that the ear/brain auditory
system can key in on the direct sound, or first sound, then fuse
or compensate for, any subsequent similar sound without detrimental
affect to the overall sound.
For small rooms, in light of the localization effects, I agree with
Ludwig that it would seem prudent to ensure at least a 1 millisecond
ITDiG by placing the sound source at least 2 feet away from the walls.
However, placing an acoustic absorber on the wall to absorb the
first reflection may actually hurt more than it helps because
most absorbers commonly used are not wideband, but rather have very
little absorption in the low frequencies, thereby creating a very
unbalanced first reflection spectrum.
Bass Ratio - BRat
Bass Ratio, what I call BRat, is a measure of the level of lower
frequency reverberation versus the level of higher frequency
reverberation. Beranek defines it as the average Reverb time
at 125Hz and 250Hz, divided by the average Reverb time
at 500Hz and 1000Hz (see below for definition of Reverb
time). Although it is defined as a time measurement, it is perceived
as a level (or gain versus loss). For example, a room with a high BRat
sounds warm or bassy, while a low BRat sounds harsh or bright.
The main factor affecting BRat is absorption. Thin paneling and drywall
absorb bass by flexing and vibrating. Most acoustical absorbers absorb
treble and midrange, but little bass (just like carpet).
Beranek's recommended range is from 1.1 to 1.45,
for concert halls with lower reverberation times.
Note that most studies of concert hall acoustics use symphonic music
as the main subject, which in general has much lower levels of low
frequency sounds than does modern music (i.e. rock and roll). This
would require a higher BRat for good sounding symphonic music and
a lower BRat for good sounding modern music.
Small rooms should sound natural and balanced, and since
most rooms inherently have more high frequency absorption than low,
a slight emphasis in the bass is desireable, which agrees with the
range of 1.1 to 1.45. However, overuse of foam type absorbers
is all too common, resulting in too high BRat with too much bass.
Since most live end dead end rooms use foam on the dead end,
they sound unnatural - half live, half dead (frequency-wise).
Inter-Aural Cross Correlation (subtracted from one) - I-ACk
Inter Aural Cross Correlation subtracted from one, what I call I-ACk,
is a measure of Apparent Source Width, or more specifically
I-ACk=1-IACC. A sound source directly in front of you without
any reflections will have zero I-ACk. A sound source fed to your
left ear, while blocking your right ear, will have I-ACk equal to one.
Lateral reflections from the side walls increase I-ACk, because
sounds coming from left or right will arrive at the ears differently
due to that big fat head in the way. The exception being if the source
is dead center in front of you, and you are positioned exactly half
way between two identical walls, but even then tiny movements of
your head will increase I-ACk. Of course, stereo speakers playing
sounds panned left or right will also increase I-ACk.
Since live end dead end theory reduces early lateral reflections,
favoring later reflections from behind, it also reduces I-ACk.
This has a negative effect on acoustical quality. For small rooms
you should make the walls as live as possible, with any
acoustic absorption used only on the floor or ceiling.
Reverberation - Reverb
Everybody loves a little ambience, and some love it a lot. Reverberation
Time, or simply Reverb, is the time it takes a sound to drop in level by
60dB. It is the most obvious difference between indoor and outdoor sound.
Talking in a dead room or an anechoic chamber is immediately
disconcerting. Music played thusly is universally considered abnormal
and some may say it is actually unpleasant.
So there is no question that rooms should have at least some Reverb,
but how much is best? Beranek and many other acousticians agree that
about 2 seconds is best for most symphonic music in concert halls.
They also agree that different types of music require different Reverb
times.
Beranek, among others, goes on to say that the early decay time (EDT) is
more important than the later decay time, and defines EDT as the time
it takes a sound to drop in level by 10 dB then multiplied by 6 to
normalize it. There are some concert halls, like the Meyerson/McDermott
in Dallas that have very different early and late decay times.
The three most widely praised concert halls in the world, Symphony Hall
in Boston, Concertgebouw in Amsterdam, and Musikvereinssaal in Vienna,
have Reverb times close to 2 seconds, with some, but not much, variation
in their early versus late decay times.
The consensus is that smaller rooms should have less reverberation
than larger rooms. They typically do anyway because reflections
are attenuated each time they bounce off a surface, and sound will have
more bounces per second in smaller rooms. Thus the density of reflections
will be higher in a small room but sound will decay more rapidly.
The most important factor in determining the optimum Reverb time is
the type of music. Modern music is faster and more impulsive than
classical music. This requires less Reverb time.
Acoustical absorbers may be used to reduce
the Reverb time, but most absorbers work on higher
frequencies with very little effect on lower frequencies.
Of course there are specialized bass traps
that absorb narrow bands of lower frequencies, but their
absorption is proportional to their area, so covering a large area
(and a wide band of frequencies) will require a lot of them. Contrary
to popular opinion, putting one or two bass traps in the corner won't
do much.
The right amount of reverberation in a small room
is whatever sounds good to the listener, which may vary from one
person to the next (along with the type of music). Beware of
excessively dead rooms with Reverb times much less than 0.5 seconds
which will have a negative effect on sound quality. A possible
exception would be a control room for monitoring a recording, where
you might want to hear exactly what is being recorded without any
pleasant reverberation added. Although, in that case, headphones
would be a better solution.
Surface Diffusivity - SurDif
Surface Diffusivity, which I call SurDif, is a property of highly
irregular walls. Flat walls have no SurDif. Diffusion is not
directly related to absorption, but absorption tends to limit the
effects of diffusion. That is, dead surfaces cannot have good
diffusion because they absorb most of the sound.
The degree of diffusivity is proportional to the amount of surface
irregularities, but is inversely proportional to the absorption.
Unfortunately, there is no objective way to measure it.
Beranek says that diffusivity is an architectural feature that
must not be underestimated. The three most widely praised
concert halls in the world have excellent surface diffusivity.
Diffusion is always sighted as very important to good sound. Haan
and Fricke say that it is the definitive quantity for concert hall
sound, and the following is how they defined the Surface Diffusivity
Index:
High Diffusivity (1.0) - coffered or checker-designed with deep recesses
or deep beams [greater than 4 in.], or random diffusing elements over
the full area [greater than 2 in. in depth], and all of the area must
not embody any sound absorbing materials.
Medium Diffusivity (0.5) - angled array of broken surfaces, or ornamentally
decorative treatment applied with shallow recesses [less then 2 in. in
depth], or flat concrete surfaces behind a semi-acoustically transparent
screen with mostly reflective materials.
Low Diffusivity (0.0) - large separate paneling, or smoothly curved surface,
or large flat and smooth surface, or semi-acoustically transparent mesh
screen, or heavy absorptive treatment applied.
SurDif is the most important factor in small room acoustics,
so the higher the diffusivity, the better. There are several
diffusors available on the market, but they tend to be quite
expensive. Some of them use mathematical formulas to help ensure
random frequency reflection distribution. Fortunately, there are
some household objects that make great diffusors, like bookshelves!
Summary
Irregular surfaces, such as bookshelves, brick fireplaces, cabinets, or
manufactured diffusors are the panacea for small room acoustics. They will
help your room sound as marvelous as it can. Moderate use of acoustical
absorbers may help in some rooms, but as far as placing the absorbers
following live end dead end theory, forget about it!
The following ranks the parameters discussed above according to relative
importance for room acoustics. However, equally important as perhaps all
of these parameters combined, is standing waves, which has been covered
extensively by many other authors. I do have a somewhat unique solution
for the problems caused by standing waves that I may present in
the future.
Harner's Subjective Preference rankings for small rooms
1. SurDif - 40%
2. Reverb - 25%
3. I-ACk - 20%
4. BRat - 10%
5. ITDiG - 5%
6. Loudness - 0%
References
1. Ludwig, Art, Room Acoustics, etc.,
http://www.silcom.com/~aludwig/
2. Beranek, Leo, Concert and Opera Halls - How They Sound,
1996 by Acoustical Society of America
3. Everest, F. Alton, Master Handbook of Acoustics - Fourth Edition
2001 by The McGraw-Hill Companies, Inc.
4. Rettinger, Michael, Handbook of Architectural Acoustics and Noise Control
1988 by Blue Ridge Summit
5. Ando, Yoichi, Concert Hall Acoustics, 1985
6. Haan, C. H. and Fricke, F. R., Surface diffusivity as a measure of the
acoustic quality of concert halls, 1993
7. Hafter, Ervin, Auditory Perception Lab,
http://ear.berkeley.edu/auditory_lab
|
