Preprint No. 2531
Presented at the 83rd AES Convention
1987 October 16-19
Arthur M. Noxon
Acoustic Sciences Corp.
Eugene, OR 97402 U.S.A.
Sampling room design is of current interest. This is due to the
influx of sample processors into the studio. The goal is to catch
samples that are musically realistic. The question arises as to
how short the sampling time window can be, what controls it and
what is in it. Independent of the sampling problem, a studio mic
technique was developed over three years ago. It is characterized
by a strong liveness quality being added to an otherwise quite dry
sound. Over the last year, sampling rooms have been designed and
built using this acoustic technique, providing very satisfactory
results. This paper presents the design strategy and acoustic signatures
of recording rooms that have this “quick” sound quality
and presents a case for its suitability as a sampling room.
In the beginning, there is only one. Soon, the knowing few step
forward, and eventually come the hordes. This is also the lifeline
of each acoustical moment. We have the direct signal, soon followed
by a set of early reflections, trailed by the multitudes in reverberation.
0.0 0.0 Introduction
The direct signal is received as it is sent. Both the early reflections
and the reverberation will have distinct characteristics that are
a function of the reflecting surfaces that support them. Each of
these two reflection groups can be weak or strong. They may have
temper, or spectral band pass characteristics. They each will also
have a temporal or time-wise signature that describes their density
and distribution of discrete signals including the decay rate. A
flow diagram can be drawn outlining the multiple signal path options
between the sound source and receiver. This outline is loaded with
vocabulary that describes the quality of the sound options.
0.1 Early Reflections
There are two sides to the control room window. The recording studio
produces the signal which then is mixed in the control room. Recording
studios are very carefully set up to produce the proper composite
balance between the direct, strong early and weak late reflections
(1). The goal is to get a natural, realistic and full instrument
sound onto a track suitable for the mix.
Control room standards require the engineer to hear the set of
early reflections produced in the studio without distraction by
the set of early reflections that belong to the control room (2).
Its ambience is allowed after an initial time delay of 30ms.
0.1A and 0.1B show, at the risk of oversimplification, a comparison
of the ETC (Energy Time Curve) of these two rooms, and that the
handling early reflections is their major difference. The recording
studio (A) emphasizes early reflections. The control room (B) suppresses
0.2 Recording Studio Rooms
Large recording studios with strong diffusive surfaces and fast
decay rates are prescribed for accurate instrumental recording.
They provide plenty of early reflections. The vocal booth and drum
booth are other types of source rooms in the studio that have opposite
natures. They are small and usually quite dead. In spite of themselves,
they are often used as isolation rooms for instrumental recording.
|0.3 The Sampling Room
The acoustic environs appropriate to a sampling room are at present
ill defined. The traditional vocal booth approach is over damped,
too dry. There are not enough early reflections to collect sufficient
signals to develop a realistic sample. The musician also needs to
hear the full sound of the instrument; such feedback is necessary
in order to fine-tune voicing detail.
The acoustically bright and large studio produces reflections that
are strong, easy to play to, but mainly too time delayed and initially
too sparse. If sampling occurs in a small bright room, the early
reflections may be soon enough, but risk being too strong and too
colored with small room resonances. Such small room mic work is
extremely position dependent; setup and repeatability are difficult
and time consuming.
When the concept of diffusion is introduced, the time frame of
10ms for early reflection signals is the basis (3). Triple tonguing
trumpet players in a concert hall produce audible dynamic transients
whose duration is between 10ms and 20ms (4). Strong, dense and early
reflections are necessary to accurately track musical transitions.
The apparent goal is to establish a small sampling room that has
very fast decay rates, as does the small vocal room and drum booth,
yet it must have a measure of very early, neutral and diffuse ambience,
reminiscent of the larger recording studios. Rapid decay with rapid
diffusion may well define the timewise signature appropriate to
the sampling room. This will have the quality of being acoustically
“quick,” i.e. live yet dry at the same time (5).
|1.0 The Quick Sound Gobo, the “Acoustic Island”
A gobo technique was developed over three years ago that reduced
the sluggish presence of the large recording studio ambience, while
increasing the density of early reflections. A sense of liveness
is developed in the signal. The original “Acoustic Island”
gobo technique remains in use in numerous studios and is present
|1.1 The Setup
The Acoustic Island utilizes a grouping of cylindrical sound absorbers
called TUBE TRAPS.* A description of them will develop an understanding
as to the reason for their use. An interior air volume (C) is surrounded
by a dense fiberglass wall (R). This is a lumped parameter design
whose acoustic RC time constant helps to access the low frequencies
(7,8). About one half of the surface of this patented (9) trap is
covered by a “limp mass” (L); it reflects mid and high
frequency sound yet passes the lows for absorption.
The setup for the Acoustic Island gobo is in the form of a horseshoe
pattern. Two 3 foot sound trap cylinders are connected together
to form a column. Typically, seven columns are placed on a 2 foot
radius centered about the mic. The reflectors of the traps are directed
This gobo system performs two acoustic functions at the same time.
The absorptive side of each trap faces the room, intercepting the
sound of the room. This acoustic shadow zone feature develops 5dB
isolation from room ambience. The second feature is by the reflectors.
The direct signal is immediately followed by a dense fill of diffuse
signals, strong in the first 10ms, which provides a boost of 4.7dB
in the nearfield ambience. This immediate, dense and diffuse backfill
is the voice of the QUICK SOUND FIELD (QSF)* effect.
*TUBE TRAPS and QSF are both registered trademarks of Acoustic
|1.2 Gobo Testing
A typical vocal gobo was set up in a lightly treated, gyp board
sound-testing room (8 x 14 x 18.5’). A “hot spot”
speaker simulated the voice and 1/4” mic was positioned 2
feet away. The setup was 9 feet out from one corner with a patch
of carpet below and some 1” fiberglass batt overhead.
1.2A and 1.2B (see below) are the ETC and waterfall taken in the
room without a gobo. The ETC is 40ms, and the TEF (Time Energy Frequency)
time ranges from 1 to 33ms. The ETC shows very few signals in the
first 10ms compared to the second 10ms period. The direct to reverb
energy ratio is 8.1dB with an early decay rate of 0.2 sec. Not the
first 20ms has sparsely distributed returns. TEF waterfall (B) shows
the room holding energy up through 5k, but notice the rapid shift
from the full spectrum direct signal to the half spectrum set of
1.2C and 1.2D (see below) show the gobo setup but with the reflectors
positioned to the outside. The room reverberant field is weakened,
dir/rev ratio is 13.1dB, as the direct sound is absorbed by the
traps. TEF waterfall and ETC both show some increased density of
early reflections, due to the impedence discontinuity of the absorbers.
1.2E and 1.2F (see below) show the correct setup, reflectors toward
the mic. Note the ETC, tremendous early reflection backfill. Direct/reverb
ratio is 8.85dB, with an early decay rate of 0.05 sec. Reflections
from the gobo immediately follow the direct signal for 10ms. The
ETC has the classic QUICK SOUND FIELD signature, an immediate and
strong backfill of diffuse energy lies just behind the direct signal.
This feature establishes the “quick” quality of sound,
giving it a lifelike, snappy presence.
(left) shows the EFC (Energy Frequency Curve) of the early diffuse
reflections, the backfill off of the reflections of the traps. A
frequency sweep was taken at 5 1/3ms, only 3ms following the direct
signal. This reflection is also visible in the waterfall of 1.2H.
Frequency is linear in both. The neutral, broadband early diffuse
reflections are clearly present.
A-F Acoustic Island Gobo Signatures
2 The QUICK SOUND FIELD Room
Success with the Acoustic Island gobo in the larger studio spaces
led to an extension of the principles into the smaller, dedicated
sound rooms, such as a vocal booth, drum room, broadcast voice-over
rooms and the like, including the sampling room.
2.1 The Basics
The acoustic devices utilized are the half and quarter round versions
of the full round sound traps used in the gobo system. These segmented
traps each have a reflector covering the central 1/3 to 1/2 of the
surface of the trap. They are easily mounted in any position: horizontal,
vertical or upside down. Their stiffness is due to a built-up-beam
integral to the mechanically damped backboard structure.
The curved reflector in each trap serves to scatter midrange and
high frequency sound. The lower frequencies are absorbed by the
entire trap’s surface. The lows are scattered not directly
by the trap but by the process of diffraction as they rebound off
the thin reflective wall strips between each trap. Dispersion of
sound here is a two stage process.
Two types of walls have been built. A bare gyp boardroom can be
outfitted with a set of traps on 18” centers. Another approach,
initially used, is a freestanding isolation booth. It uses lead-backed
traps with “tongue and groove” Plexiglas strips in between.
That combination produces an STC (Sound Transmission Coefficient)
of 32dB yet provides 30% visual contact with the outside.
© 2009 Acoustic Sciences
Corporation. All Rights Reserved.
TO PAGE 2