4:54 PM
Page i
Second edition
3:11 PM
Page ii
3:11 PM
Page iii
Second edition
John Eargle
Focal Press is an imprint of Elsevier
3:11 PM
Page iv
Focal Press
An imprint of Elsevier
Linacre House, Jordan Hill, Oxford OX2 8DP
30 Corporate Drive, Burlington MA 01803
First published 2005
Copyright © 2005, John Eargle. All rights reserved
The right of John Eargle to be identified as the author of this work
has been asserted in accordance with the Copyright, Designs and
Patents Act 1988
No part of this publication may be reproduced in any material form (including
photocopying or storing in any medium by electronic means and whether
or not transiently or incidentally to some other use of this publication) without
the written permission of the copyright holder except in accordance with the
provisions of the Copyright, Designs and Patents Act 1988 or under the terms of
a licence issued by the Copyright Licensing Agency Ltd, 90 Tottenham Court Road,
London, England W1T 4LP. Applications for the copyright holder’s written
permission to reproduce any part of this publication should be addressed
to the publisher
Permissions may be sought directly from Elsevier’s Science and Technology Rights
Department in Oxford, UK: phone: (+44) (0) 1865 843830; fax: (+44) (0) 1865 853333;
e-mail: permissions@elsevier.co.uk. You may also complete your request on-line via the
Elsevier homepage (www.elsevier.com), by selecting ‘Customer Support’
and then ‘Obtaining Permissions’
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloguing in Publication Data
A catalogue record for this book is available from the Library of Congress
02405 1961 2
For information on all Focal Press publications visit our website at:
Typeset by Newgen Imaging Systems (P) Ltd, Chennai, India
Printed and bound in the United States
Working together to grow
libraries in developing countries
www.elsevier.com | www.bookaid.org | www.sabre.org
4:55 PM
Page v
Preface to the First Edition
Preface to the Second Edition
Symbols Used in This Book
1 A Short History of the Microphone
2 Basic Sound Transmission and Operational Forces on
3 The Pressure Microphone
4 The Pressure Gradient Microphone
5 First-Order Directional Microphones
6 High Directionality Microphones
7 Microphone Measurements, Standards, and
8 Electrical Considerations and Electronic Interface
9 Overview of Wireless Microphone Technology
10 Microphone Accessories
11 Basic Stereophonic Recording Techniques
12 Stereo Microphones
13 Classical Stereo Recording Techniques and
14 Studio Recording Techniques
15 Surround Sound Microphone Technology
16 Surround Recording Case Studies
3:11 PM
Page vi
17 A Survey of Microphones in Broadcast and
18 Fundamentals of Speech and Music
19 Overview of Microphone Arrays and Adaptive
Care and Maintenance of Microphones
Classic Microphones: The Author’s View
References and Bibliography
3:11 PM
Page vii
Most sound engineers will agree that the microphone is the most
important element in any audio chain, and certainly the dazzling array
of current models, including many that are half-a-century old, attests to
that fact. My love affair with the microphone began when I was in my
teens and got my hands on a home-type disc recorder. Its crystal microphone was primitive, but I was nonetheless hooked. The sound bug had
bitten me, and, years of music schooling not withstanding, it was
inevitable that I would one day end up a recording engineer.
About thirty years ago I began teaching recording technology at various summer educational programs, notably those at the Eastman School
of Music and later at the Aspen Music Festival and Peabody Conservatory.
I made an effort at that time to learn the fundamentals of microphone performance and basic design parameters, and my Microphone Handbook,
published in 1981, was a big step forward in producing a text for some of
the earliest collegiate programs in recording technology. This new book
from Focal Press presents the technology in greater depth and detail and,
equally important, expands on contemporary usage and applications.
The Microphone Book is organized so that both advanced students
in engineering and design and young people targeting a career in audio
can learn from it. Chapter 1 presents a short history of the microphone.
While Chapters 2 through 6 present some mathematically intensive
material, their clear graphics will be understandable to those with little
technical background. Chapters 7 through 10 deal with practical matters
such as standards, the microphone-studio electronic interface, and all
types of accessories.
Chapters 11 through 17 cover the major applications areas, with
emphasis on the creative aspects of music recording in stereo and surround sound, broadcast/communications, and speech/music reinforcement. Chapter 18 presents an overview of advanced development in
microphone arrays, and Chapter 19 presents helpful hints on microphone
maintenance and checkout. The book ends with a comprehensive microphone bibliography and index.
3:11 PM
Page viii
Preface to the First Edition
I owe much to Leo Beranek’s masterful 1954 Acoustics text, A. E.
Robertson’s little known work, Microphones, which was written for the
BBC in 1951, as well as the American Institute of Physics’ Handbook of
Condenser Microphones. As always, Harry Olson’s books came to my
aid with their encyclopedic coverage of everything audio.
Beyond these four major sources, any writer on microphones must
rely on technical journals and on-going discussions with both users and
manufacturers in the field. I would like to single out for special thanks
the following persons for their extended technical dialogue: Norbert
Sobol (AKG Acoustics), Jörg Wuttke (Schoeps GmbH), David Josephson
(Josephson Engineering), Keishi Imanaga (Sanken Microphones), and in
earlier days Steve Temmer and Hugh Allen of Gotham Audio. Numerous
manufacturers have given permission for the use of photographs and
drawings, and they are credited with each usage in the book.
John Eargle
April 2001
3:11 PM
Page ix
The second edition of The Microphone Book follows the same broad
subject outline as the first edition. Most of the fundamental chapters
have been updated to reflect new models and electronic technology,
while those chapters dealing with applications have been significantly
broadened in their coverage.
The rapid growth of surround sound technology merits a new chapter of its own, dealing not only with traditional techniques but also with
late developments in virtual imaging and the creation of imaging that
conveys parallax in the holographic sense.
Likewise, the chapter on microphone arrays has been expanded to
include discussions of adaptive systems as they involve communications
and useful data reduction in music applications.
Finally, at the suggestion of many, a chapter on classic microphones
has been included. Gathering information on nearly thirty models was a
far more difficult task than one would ever have thought, and it was
truly a labor of love.
John Eargle
Los Angeles, June 2004
3:11 PM
Page x
radius of diaphragm (mm); acceleration (m/s2)
ampere (unit of electrical current)
audio frequency
speed of sound (334 m/s at normal temperature)
temperature (degrees celsius)
distance (m)
relative level (decibel)
A-weighted sound pressure level
signal voltage level (re 0.775 volt rms)
critical distance (m)
directivity index (dB)
voltage (volt dc)
signal voltage (volt rms)
frequency in hertz (s–1)
high frequency
frequency (hertz, cycles per second)
acoustical intensity (W/m2)
dc electrical current, ampere (Q/s)
signal current (ampere rms)
mechanical moment of inertia (kg m2)
complex algebraic operator, equal to 1
wave number (2/)
mass, kilogram (SI base unit)
temperature (degrees kelvin, SI base unit)
low frequency
sound pressure level (dB re 20 Pa)
reverberant sound pressure level (dB re 20 Pa)
noise sound pressure level (dB re 20 Pa)
meter (SI base unit)
mid frequency
millimeter (m 10–3)
micrometer or micron (m 10–6)
3:11 PM
Page xi
Symbols Used in This Book
p; p(t)
R, R
T, t
u; u(t)
U, U(t)
microphone system sensitivity, mV/Pa
capacitor microphone base diaphragm sensitivity, V/Pa
force, newton (kg, m/s2)
rms sound pressure (N/m2)
power (watt)
electrical charge (coulombs, SI base unit)
directivity factor
distance from sound source (m)
electrical resistance (ohm)
room constant (m3 or ft3)
random efficiency of microphone (also REE)
radio frequency
relative humidity (%)
second (SI base unit)
surface area (m2)
torque (N m)
time (s)
magnetic flux density (tesla)
reverberation time (seconds)
diaphragm tension (newton/meter)
atmospheric pressure; equal to mm of mercury (mmHg), or
133.322 Pa (Note: 760 torr is equal to normal atmospheric
pressure at 0C)
air particle rms velocity (m/s)
air volume rms velocity (m3/s)
air particle displacement (m/s)
mechanical, acoustical, or electrical reactance ()
electrical voltage (voltage or potential)
mechanical, acoustical or electrical resistance ()
average absorption coefficient (dimensionless)
wavelength of sound in air (m)
phase, phase shift (degrees or radians)
dependent variable in polar coordinates
density of air (1.18 kg/m3)
specific acoustical impedance of air (415 SI rayls)
angle (degrees or radians), independent variable in polar
2f (angular frequency in radians/s)
surface mass density (kg/m2)
3:11 PM
Page xii
2:34 PM
Page 1
The microphone pervades our daily lives through the sound we hear on
radio, television and recordings, paging in public spaces, and of course
in two-way communications via telephone. In this chapter we will touch
upon some of the highlights of more than 125 years of microphone
development, observing in particular how most of the first 50 of these
years were without the benefits of electronic amplification. The requirements of telephony, radio broadcast, general communications, and
recording are also discussed, leading to some conjecture on future
As children, many of us were fascinated with strings stretched between
the ends of a pair of tin cans or wax paper cups, with their ability to convey speech over a limited distance. This was a purely mechano-acoustical
arrangement in which vibrations generated at one end were transmitted
along the string to actuate vibrations at the other end.
In 1876, Alexander Graham Bell received US patent 174,465 on the
scheme shown in Figure 1–1. Here, the mechanical string was, in a sense,
replaced by a wire that conducted electrical direct current, with audio
signals generated and received via a moving armature transmitter and its
associated receiver. Like the mechanical version, the system was reciprocal. Transmission was possible in either direction; however, the
patent also illustrates the acoustical advantage of a horn to increase the
driving pressure at the sending end and a complementary inverted horn
to reinforce output pressure at the ear at the receiving end. Bell’s further
experiments with the transmitting device resulted in the liquid transmitter,
2:34 PM
Page 2
The beginnings of
telephony; Bell’s
original patent.
Bell’s liquid transmitter
exhibited at the
Philadelphia Centennial
Exposition of 1876.
shown in Figure 1–2, which was demonstrated at the Philadelphia
Centennial Exposition of 1876. Here, the variable contact principle
provided a more effective method of electrical signal modulation than
that afforded by the moving armature.
The variable contact principle was extended by Berliner in a patent
application in 1877 in which a steel ball was placed against a stretched
metal diaphragm, as shown in Figure 1–3. Further work in this area was
done by Blake (patents 250, 126 through 250, 129, issued in 1881), who
used a platinum bead impressed against a hard carbon disc as the variable resistance element, as shown in Figure 1–4. The measured response
of the Blake device spanned some 50 decibels over the frequency range
2:34 PM
Page 3
1: A Short History of the Microphone
Berliner’s variable
contact microphone.
from 380 Hz to 2000 Hz, and thus fell far short of the desired response.
However, it provided a more efficient method of modulating telephone
signals than earlier designs and became a standard in the Bell system for
some years.
Another interim step in the development of loose contact modulation of direct current was developed in 1878 by Hughes and is shown in
Figure 1–5. In this embodiment, very slight changes in the curvature of
the thin wood plate diaphragm, caused by impinging sound waves, gave
rise to a fairly large fluctuation in contact resistance between the carbon
rod and the two mounting points. This microphone was used by Clement
Ader (Scientific American, 1881) in his pioneering two-channel transmissions from the stage of the Paris Opera to a neighboring space. It was
Hughes, incidentally, who first used the term microphone, as applied to
electroacoustical devices.
The ultimate solution to telephone transmitters came with the development of loose carbon granule elements as typified by Blake’s transmitter
of 1888, shown in Figure 1–6. Along with the moving armature receiver,
the loose carbon granule transmitter, or microphone, has dominated
telephony up to the present. Quite a testimony to the inventiveness and
resourcefulness of engineers working nearly 130 years ago.
2:34 PM
Page 4
The carbon granule transmitter and moving armature receiver complemented each other nicely. The limitations in bandwidth and dynamic
range have never been a problem in telephony, and the rather high distortion generated in the combined systems actually improved speech
intelligibility by emphasizing higher frequencies. Even after the invention
of electronic amplification (the de Forest audion vacuum tube, 1907),
these earlier devices remained in favor.
When commercial broadcasting began in the early 1920s, there was
a clear requirement for better microphones as well as loudspeakers.
Western Electric, the manufacturing branch of the Bell Telephone system,
was quick to respond to these needs, developing both electrostatic
(capacitor) microphones as well as electrodynamic (moving conductor)
microphones. The capacitor, or condenser, microphone used a fixed electrical charge on the plates of a capacitor, one of which was a moving
diaphragm and the other a fixed back plate. Sound waves caused a slight
variation in capacitance, which in turn was translated into a variation in
the voltage across the plates. An early Western Electric capacitor microphone, developed by Wente in 1917, is shown in Figure 1–7.
While first employed as a driving element for loudspeakers and
headphones, the moving coil and its close relative, the ribbon, eventually
found their place in microphone design during the mid-twenties. Moving
coil and ribbon microphones operate on the same principle; the electrical conducting element is place in a transverse magnetic field, and its
motion generated by sound vibrations induces a voltage across the conducting element. Under the direction of Harry Olson, Radio Corporation
of America (RCA) was responsible for development and beneficial
exploitation of the ribbon microphone during the 1930s and 1940s.
Blake’s carbon disc
Hughes’ carbon rod
2:34 PM
Page 5
1: A Short History of the Microphone
Blake’s variable
contact microphone.
Beginning as far back as the 1920s, a number of smaller American companies, such as Shure Brothers and Electro-Voice, began to make significant contributions to microphone engineering and design. General
applications, such as paging and sound reinforcement, required ingenious and economical solutions to many problems. Commercial development of capacitor microphones was more or less ruled out, due to the
requirements for a cumbersome polarizing supply, so these companies
concentrated primarily on moving coil designs.
The work of Bauer (1941) was significant in producing the Shure
Unidyne directional (cardioid pattern) design based on a single moving
coil element. Wiggins (1954) developed the Electro-Voice “Variable-D”
single moving coil element, which provided low handling noise with
excellent directional response.
Other companies designed crystal microphones for low cost,
moderate-quality paging applications. These designs were based on the
principle of piezoelectricity (from the Greek piezien, meaning pressure),
which describes the property of many crystalline structures to develop a
voltage across opposite facets when the material is bent or otherwise
deformed. The early advantage of the piezos was a relatively high output
signal, but eventually the coming of small, high energy magnet materials
ruled them out.
For many years the penalty carried by the capacitor microphone was its
requirement for external polarization voltage. In the early sixties, Sessler
and West of Bell Telephone Laboratories described a capacitor microphone which used a permanently polarized dielectric material between
The Wente capacitor
microphone of 1917.
2:34 PM
Page 6
the movable diaphragm and the backplate of the microphone. Early
materials exhibited significant losses in sensitivity over time, but these
problems have been overcome. Further improvements have come in
miniaturization, enabling electret microphones to be designed for a wide
variety of close-in applications, such as tie-tack use, hidden on-stage
pickup, and many other uses. Today’s small electret microphone requires
only a miniature self-contained 5-to-9 volt battery to power its equally
miniature preamplifier. It is a testimony to the electret and its long-term
stability and excellent technical performance that the Brüel & Kjaer series
of studio microphones designed in the 1980s used electret technology.
The microphone is the first stage in the complex and extended technical
chain between live performance and sound reproduction in the home
or motion picture theater. Little wonder then that so much attention
has been paid to the quality and technical performance of these fine
Capacitor microphones have dominated studio recording since the
late 1940s when the first German and Austrian capacitor microphones
came on the scene. As with any mature technology, progress comes
slowly, and the best models available today have a useful dynamic range
that exceeds that of a 20-bit digital recorder. With regard to directional
performance, many newer microphones exhibit off-axis response
integrity that far exceeds the best of earlier models.
At the beginning of the 21st century, it is interesting to observe
the great nostalgia that many recording engineers have for the earlier
vacuum tube capacitor models, especially the Neumann and AKG classic
microphones of the 1950s. All of this reminds us that technology is so
often tempered with subjective judgment to good effect.
The microphone per se is so highly developed that it is often difficult to
see where specific improvements in the basic mechanisms are needed.
Certainly in the area of increased directionality, via second and higher
order designs, is there additional development engineering to be done.
There may never be a direct-converting, high-quality digital microphone
as such, but it is clear that digital signal processing will certainly play an
important part in active microphone array development.
New usage concepts include microphones in conferencing systems,
with their requirements for combining and gating of elements; and
microphones in large arrays, where highly directional, steerable pickup
patterns can be realized. These are among the many subjects that will be
discussed in later chapters.
2:35 PM
Page 7
All modern microphones benefit from electrical amplification and thus
are designed primarily to sample a sound field rather than take power
from it. In order to understand how microphones work from the physical and engineering points of view, we must understand the basics of
sound transmission in air. We base our discussion on sinusoidal wave
generation, since sine waves can be considered the building blocks of
most audible sound phenomena. Sound transmission in both plane and
spherical waves will be discussed, both in free and enclosed spaces.
Power relationships and the concept of the decibel are developed. Finally,
the effects of microphone dimensions on the behavior of sound pickup
are discussed.
Figure 2–1 illustrates the generation of a sine wave. The vertical component of a rotating vector is plotted along the time axis, as shown at A.
At each 360 of rotation, the wave structure, or waveform, begins anew.
The amplitude of the sine wave reaches a crest, or maximum value,
above the zero reference baseline, and the period is the time required for
the execution of one cycle. The term frequency represents the number of
cycles executed in a given period of time. Normally we speak of frequency in hertz (Hz), representing cycles per second.
2:35 PM
Page 8
Generation of a sine
wave signal (A); phase
relationships between two
sine waves (B).
For sine waves radiating outward in a physical medium such as air,
the baseline in Figure 2–1 represents the static atmospheric pressure,
and the sound waves are represented by the alternating plus and minus
values of pressure about the static pressure. The period then corresponds
to wavelength, the distance between successive iterations of the basic
The speed of sound transmission in air is approximately equal to
344 meters per second (m/s), and the relations among speed (m/s), wavelength (m), and frequency (1/s) are:
c (speed) f (frequency) (wavelength)
f c/
For example, at a frequency of 1000 Hz, the wavelength of sound in air
will be 344/1000 0.344 m (about 13 inches).
Another fundamental relationship between two waveforms of the
same frequency is their relative phase (), the shift of one period relative
to another along the time axis as shown in Figure 2–1B. Phase is normally measured in degrees of rotation (or in radians in certain mathematical operations). If two sound waves of the same amplitude and
2:35 PM
Page 9
2: Basic Sound Transmission and Operational Forces on Microphones
Wavelength of sound in air
versus frequency; in meters
(A); in feet (B).
frequency are shifted by 180 they will cancel, since they will be in an
inverse (or anti-phase) relationship relative to the zero baseline at all
times. If they are of different amplitudes, then their combination will not
directly cancel.
The data shown in Figure 2–2 gives the value of wavelength in air
when frequency is known. (By way of terminology, velocity and speed
are often used interchangeably. In this book, speed will refer to the rate
of sound propagation over distance, while velocity will refer to the
specifics of localized air particle and air volume movement.)
For most recording activities indoors, we can assume that normal temperatures prevail and that the effective speed of sound propagation will
be as given above. There is a relatively small dependence of sound propagation on temperature, as given by the following equation:
Speed 331.4 0.607 C m/s
where C is the temperature in degrees celsius.
2:35 PM
Page 10
In any kind of physical system in which work is done, there are two
quantities, one intensive, the other extensive, whose product determines
the power, or rate at which work is done in the system. One may intuitively think of the intensive variable as the driving agent and the extensive variable as the driven agent. Table 2–1 may make this clearer.
Power is stated in watts (W), or joules/second. The joule is the unit
of work or energy, and joules per second is the rate at which work is
done, or energy expended. This similarity among systems makes it easy
to transform power from one physical domain to another, as we will see
in later chapters.
Intensity (I) is defined as power per unit area (W/m2), or the rate of
energy flow per unit area. Figure 2–3 shows a sound source at the left
radiating an acoustical signal of intensity I0 uniformly into free space.
We will examine only a small solid angle of radiation. At a distance of
10 m that small solid angle is radiating through a square with an area of
1 m2, and only a small portion of I0 will pass through that area. At a distance of 20 m the area of the square that accommodates the original solid
angle is now 4 m2, and it is now clear that the intensity at a distance of
20 m will be one-fourth what it was at 10 m. This of course is a necessary consequence of the law of conservation of energy.
TABLE 2–1 Intensive and extensive variables
Intensive variable
Extensive variable
voltage (e)
force (f )
current (i)
velocity (u)
watts (e i)
watts (f u)
torque (T)
angular velocity (
watts (T )
pressure (p)
volume velocity (U)
watts (p U)
Sound intensity variation
with distance over a fixed
solid angle.
2:35 PM
Page 11
2: Basic Sound Transmission and Operational Forces on Microphones
The relationship of intensity and distance in a free sound field is
know as the inverse square law: as intensity is measured between distances of r and 2r, the intensity changes from 1/I0 to 1/4I0.
The intensity at any distance r from the source is given by:
I W/4r2
The effective sound pressure in pascals at that distance will be:
p I0c
where 0c is the specific acoustical impedance of air (405 SI rayls).
For example, consider a point source of sound radiating a power of
one watt uniformly. At a distance of 1 meter the intensity will be:
I 1/4(1)2 1/4 0.08 W/m2
The effective sound pressure at that distance will be:
p (0.08)405 5.69 Pa
The relation between air particle velocity (u) and particle displacement
(x) is given by:
u(t) j (t)
where 2f and x(t) is the maximum particle displacement value. The
complex operator j produces a positive phase shift of 90.
Some microphones, notably those operating on the capacitive or
piezoelectric principle, will produce constant output when placed in a
constant amplitude sound field. In this case u(t) will vary proportional to
Other microphones, notably those operating on the magnetic induction principle, will produce a constant output when placed in a constant
velocity sound field. In this case, x(t) will vary inversely proportional to
We do not normally measure acoustical intensity; rather, we measure
sound pressure level. One cycle of a varying sinusoidal pressure might look
like that shown in Figure 2–4A. The peak value of this signal is shown as
unity; the root-mean-square value (rms) is shown as 0.707, and the average value of the waveform is shown as 0.637. A square wave of unity
value, shown at B, has peak, rms, and average values that all are unity. The
rms, or effective, value of a pressure waveform corresponds directly to the
power that is delivered or expended in a given acoustical system.
The unit of pressure is the pascal (Pa) and is equal to one newton/m2.
(The newton (N) is a unit of force that one very rarely comes across in
2:35 PM
Page 12
Sine (A) and square (B)
waves: definitions of peak,
rms and average values.
daily life and is equal to about 9.8 pounds of force.) Pressures encountered in acoustics normally vary from a low value of 20 Pa (micropascals) up to normal maximum values in the range of 100 Pa. There is a
great inconvenience in dealing directly with such a large range of numbers, and years ago the decibel (dB) scale was devised to simplify things.
The dB was originally intended to provide a convenient scale for looking
at a wide range of power values. As such, it is defined as:
Level (dB) 10 log (W/W0)
where W0 represents a reference power, say, 1 watt, and the logarithm is
taken to the base 10. (The term level is universally applied to values
expressed in decibels.) With one watt as a reference, we can say that
20 watts represents a level of 13 dB:
Level (dB) 10 log (20/1) 13 dB
Likewise, the level in dB of a 1 milliwatt signal, relative to one watt, is:
Level (dB) 10 log (0.001/1) 30 dB
From basic electrical relationships, we know that power is proportional to the square of voltage. As an analog to this, we can infer that
acoustical power is proportional to the square of acoustical pressure. We
can therefore rewrite the definition of the decibel in acoustics as:
Level (dB) 10 log (p/p0)2 20 log (p/p0)
In sound pressure level calculations, the reference value, or p0, is established as 0.00002 Pa, or 20 micropascals (20 Pa).
Consider a sound pressure of one Pa. Its level in dB is:
dB 20 log (1/0.00002) 94 dB
This is an important relationship. Throughout this book, the value of 94
dB LP will appear time and again as a standard reference level in microphone design and specification. (LP is the standard terminology for sound
pressure level.)
Figure 2–5 presents a comparison of a number of acoustical sources
and the respective levels at reference distances.
2:35 PM
Page 13
2: Basic Sound Transmission and Operational Forces on Microphones
Sound pressure levels of
various sound sources.
The graph in Figure 2–6 shows the relationship between pressure in
Pa and LP. The nomograph shown in Figure 2–7 shows the loss in dB
between any two reference distances from a point source in the free field.
Referring once again to equation (2.4), we will now calculate the
sound pressure level of one acoustical watt measured at a distance of 1 m
from a spherically radiating source:
LP 20 log (5.69/0.00002) 109 dB
It can be appreciated that one acoustical watt produces a considerable sound pressure level. From the nomograph of Figure 2–7, we can
see that one acoustical watt, radiated uniformly and measured at a
2:35 PM
Page 14
Relationship between
sound pressure and sound
pressure level.
Inverse square sound pressure level relationships as a function of distance from the source; to determine the level difference
between sound pressures at two distances, located the two distances and then read the dB difference between them; for
example, determine the level difference between distances 50 m and 125 m from a sound source; above 50 read a level of
34 dB; above 125 read a level of 42 dB; taking the difference gives 8 dB.
distance of 10 m (33 feet), will produce LP 89 dB. How “loud” is a
signal of 89 dB LP? It is approximately the level of someone shouting in
your face!
A free field exists only under specific test conditions. Outdoor conditions
may approximate it. Indoors, we normally observe the interaction of a
2:35 PM
Page 15
2: Basic Sound Transmission and Operational Forces on Microphones
direct field and a reverberant field as we move away from a sound
source. This is shown pictorially in Figure 2–8A. The reverberant field
consists of the ensemble of reflections in the enclosed space, and reverberation time is considered to be that time required for the reverberant
field to diminish 60 dB after the direct sound source has stopped.
There are a number of ways of defining this, but the simplest is given
by the following equation:
Reverberation time (s) 0.16 V
where V is the room volume in m3, S is the interior surface area in m2,
and is the average absorption coefficient of the boundary surfaces.
The distance from a sound source to a point in the space where both
direct and reverberant fields are equal is called critical distance (DC).
In live spaces critical distance is given by the following equation:
The reverberant field.
Illustration of reflections
in an enclosed space
compared to direct sound
at a variable distance from
the sound source (A);
interaction of direct and
reverberant fields in a live
space (B); interaction of
direct and reverberant
fields in a damped
space (C).
2:35 PM
Page 16
DC 0.14QS
where Q is the directivity factor of the source. We will discuss this topic
in further detail in Chapter 17.
In a live acoustical space, may be in the range of 0.2, indicating
that, on average, only 20% of the incident sound power striking the
boundaries of the room will be absorbed; the remaining 80% will reflect
from those surfaces, strike other surfaces, and be reflected again. The
process will continue until the sound is effectively damped out.
Figures 2–8B and C show, respectively, the observed effect on sound
pressure level caused by the interaction of direct, reflected, and reverberant fields in live and damped spaces.
Normally, microphones are used in the direct field or in the transition region between direct and reverberant fields. In some classical
recording operations, a pair of microphones may be located well within
the reverberant field and subtly added to the main microphone array for
increased ambience.
For wave motion in a free plane wave field, time varying values of sound
pressure will be in phase with the air particle velocity, as shown in
Figure 2–9. This satisfies the conditions described in Table 2.1, in which
the product of pressure and air volume velocity define acoustical power.
(Volume velocity may be defined here as the product of particle velocity
and the area over which that particle velocity is observed.)
If a microphone is designed to respond to sound pressure, the conditions shown in Figure 2–9A are sufficient to ensure accurate reading of
the acoustical sound field.
Most directional microphones are designed to be sensitive to the air
pressure difference, or gradient, existing between two points along a
given pickup axis separated by some distance l. It is in fact this sensitivity that enables these microphones to produce their directional pickup
characteristics. Figure 2–9B shows the phase relationships at work here.
The pressure gradient [dp/dl] is in phase with the particle displacement
[x(t)]. However, the particle displacement and particle velocity [dx/dt]
are at a 90 phase relationship.
These concepts will become clearer in later chapters in which we
discuss the specific pickup patterns of directional microphones.
Relatively close to a radiating sound source, the waves will be more or
less spherical. This is especially true at low frequencies, where the difference in wavefront curvature for successive wave crests will be quite
pronounced. As our observation point approaches the source, the phase
2:35 PM
Page 17
2: Basic Sound Transmission and Operational Forces on Microphones
Wave considerations
in microphone
performance: relationship
between sound pressure
and particle velocity (A);
relationship among air
particle velocity, air particle
displacement, and pressure
gradient (B); relationship
between pressure and
pressure gradient (C) (Data
presentation after
Robertson, 1963).
angle between pressure and particle velocity will gradually shift from zero
(in the far field) to 90, as shown in Figure 2–10A. This will cause an
increase in particle velocity with increasing phase shift, as shown at B.
As we will see in a later detailed discussion of pressure gradient
microphones, this phenomenon is responsible for what is called proximity effect, the tendency of directional microphones to increase their LF
(low frequency) output at close operating distances.
Figure 2–11 shows the effects of both inverse square losses and HF losses
due to air absorption. Values of relative humidity (RH) of 20% and 80%
are shown here. Typical losses for 50% RH would be roughly halfway
between the plotted values shown.
For most studio recording operations HF losses may be ignored.
However, if an organ recording were to be made at a distance of 12 m in
a large space and under very dry atmospheric conditions, the HF losses
could be significant, requiring an additional HF boost during the recording process.
2:35 PM
Page 18
Spherical sound waves:
phase angle between
pressure and particle
velocity in a spherical
wave at low frequencies; r
is the observation distance
and is the wavelength of
the signal (A); increase in
pressure gradient in a
spherical wave at low
frequencies (B).
Effects of both inverse
square relationships and
HF air losses (20% and
80% RH).
2:35 PM
Page 19
2: Basic Sound Transmission and Operational Forces on Microphones
Microphones are normally fairly small so that they will have minimal effect
on the sound field they are sampling. There is a limit, however, and it is difficult to manufacture studio quality microphones smaller than about 12 mm
(0.5 in) in diameter. As microphones operate at higher frequencies, there are
bound to be certain aberrations in directional response as the dimensions of
the microphone case become a significant portion of the sound wavelength.
Diffraction refers to the bending of sound waves as they encounter objects
whose dimensions are a significant portion of a wavelength.
Many measurements of off-axis microphone response have been
made over the years, and even more theoretical graphs have been developed. We will now present some of these.
Theoretical polar
response for a microphone
mounted at the end of a
tube. (Data presentation
after Beranek, 1954.)
2:35 PM
Page 20
Figure 2–12 shows polar response diagrams for a circular diaphragm
at the end of a long tube, a condition that describes many microphones. In
the diagrams, ka 2a/, where a is the radius of the diaphragm. Thus,
ka represents the diaphragm circumference divided by wavelength. DI
stands for directivity index; it is a value, expressed in decibels, indicating
the ratio of on-axis pickup relative to the total pickup integrated over all
directions. Figure 2–13 shows the same set of measurements for a microphone which is effectively open to the air equally on both sides. It represents the action a ribbon microphone, with its characteristic “figure-eight”
angular response.
Figure 2–14 shows families of on- and off-axis frequency response
curves for microphones mounted on the indicated surfaces of a cylinder
and a sphere. Normally, a limit for the HF response of a microphone
would be a diameter/ ratio of about one.
In addition to diffraction effects, there are related response aberrations due to the angle at which sound impinges on the microphone’s
diaphragm. Figure 2–15A shows a plane wave impinging at an off-axis
oblique angle on a microphone diaphragm subtended diameter which is
one-fourth of the sound wavelength. It can be seen that the center portion of the diaphragm is sampling the full value of the waveform, while
adjacent portions are sampling a slightly lesser value. Essentially, the
diaphragm will respond accurately, but with some small diminution of
output for the off-axis pickup angle shown here.
Theoretical polar response
for a free microphone
diaphragm open on both
sides. (Data presentation
after Beranek, 1954.)
2:35 PM
Page 21
2: Basic Sound Transmission and Operational Forces on Microphones
On and off-axis frequency
response for microphones
mounted on the end of a
cylinder and a sphere. (Data
after Muller et al., 1938.)
Plane sound waves
impinging on a microphone
diaphragm at an oblique
angle. Microphone
diaphragm subtended
diameter equal to /4 (A);
microphone diaphragm
subtended diameter equal
to (B). (Data after
Robertson, 1963.)
The condition shown in Figure 2–15B is for an off-axis sound wavelength which is equal to the subtended diameter of the microphone
diaphragm. Here, the diaphragm samples the entire wavelength, which
will result in near cancellation in response over the face of the
2:37 PM
Page 22
An ideal pressure microphone responds only to sound pressure, with no
regard for the directional bearing of the sound source; as such, the
microphone receives sound through a single active opening. In reality, a
pressure microphone exhibits some directionality along its main axis at
short wavelengths, due principally to diffraction effects. As we saw in
Figure 2–12, only as the received wavelength approaches the circumference of the microphone diaphragm does the microphone begin to depart
significantly from omnidirectional response. For many studio-quality
pressure microphones, this becomes significant above about 8 kHz.
The earliest microphones were of the pressure type, and the capacitor pressure microphone is widely used today in music recording as well
as in instrumentation and measurement applications. By way of terminology, the capacitor microphone is familiarly referred to in the sound
industry as the condenser microphone, making use of the earlier electrical term for capacitor. We use the modern designation capacitor throughout this book.
We begin with a study of the capacitor pressure microphone, analyzing it in physical and electrical detail. We then move on to a similar
analysis of the dynamic pressure microphone. Other transducers that
have been used in pressure microphone design, such as the piezoelectric
effect and the loose contact (carbon granule) effect, are also discussed.
The RF (radio frequency) signal conversion principle use in some capacitor microphones is discussed in Chapter 8.
Figure 3–1 shows section and front views of a modern studio capacitor
pressure microphone capsule along with its associated electrical circuitry.
It is very similar to Wente’s original 1917 model (see Figure 1–7), except
that the modern design shown here is about one-third the diameter of the
2:37 PM
Page 23
3: The Pressure Microphone
Wente model. General dimensions and relative signal values for a typical
12.7 mm (0.5 in) diameter capacitor pressure microphone are:
1. Dimension h is about 20 mm (0.0008 in).
2. Peak-to-peak diaphragm displacement for a 1 Pa rms sine wave signal (94 dB LP) is about 1 108 m.
3. The static capacitance of the capsule is about 35 pF.
4. For a dc polarizing voltage on the diaphragm of 60 V, the signal
voltage generated by the 1 Pa acoustical signal is about 12 Vrms at
the capacitor’s terminals before preamplification.
Table 3–1 indicates the peak diaphragm displacement as a function
of pressure level. For the sake of example, let us multiply the microphone’s
dimensions by 1 million. The microphone now is 12 km in diameter, and
Details of a 12 mm diameter
capacitor microphone:
section view (A); front
view (B); simplified
electrical circuit (C).
TABLE 3–1 Peak diaphragm displacement relative to pressure level
Level (dB LP)
Peak-to-peak displacement (m)
2:37 PM
Page 24
the distance from the backplate to the diaphragm is 20 m. For LP of
94 dB, the peak-to-peak displacement of the diaphragm is about 10 mm.
Using the same model, the peak diaphragm displacement at 134 dB LP is
about 1 m, representing a 5% variation in spacing between the backplate and diaphragm. This exercise aptly demonstrates the microscopic
nature of the capacitor microphone’s diaphragm motion at normal
sound levels.
The very small holes in the backplate (Figure 3–1B) are evenly distributed on a uniform grid. During the actual back and forth motion of
the diaphragm, the air captured in the holes provides damping of the
diaphragm’s motion at its principal resonance, which is normally in the
range of 8–12 kHz. The diaphragm is usually made of a thin plastic
material, typically Mylar, on which has been deposited a molecular-thin
layer of metal, often gold. Aluminum, nickel, and titanium also have
been used as diaphragm materials.
Operating in parallel with the tension of the diaphragm itself is the
added stiffness provided by the captured air behind the diaphragm. Both
of these are necessary to maintain the high resonance frequency of the
diaphragm assembly. A very small capillary tube connects the interior air
mass to the outside, providing a slow leakage path so that static atmospheric pressure will equalize itself on both sides of the diaphragm under
all atmospheric conditions. The polarizing circuitry is shown at C.
One early notable capacitor microphone model, the Altec-Lansing
21C, used a very thin 50 m (0.002 in) gold sputtered glass plate in
place of the normal flexible diaphragm. The inherent stiffness of the glass
plate required no additional tension to attain a suitably high-resonance
While we customarily speak of a capacitor microphone’s diaphragm as
having a single degree of freedom, the actual motion over most of the
frequency range resembles that of an ideal drum head, as shown in
exaggerated form in Figure 3–2A. It is intuitively clearer and mathematically simpler to think of the diaphragm as a rigid piston. If we take
the center of the diaphragm as our reference displacement value, it is
obvious that the outer portions of the diaphragm exhibit less displacement. Wong and Embleton (1994) show that the effective area of an
equivalent rigid piston is one that has an area one-third that of the
actual diaphragm.
As the circular edge-supported membrane vibrates at higher frequencies, its motion no longer is simple. Motions influenced by radial
and tangential modes in the diaphragm, such as those shown in
Figures 3–2B and 3–2C begin to appear, and the microphone’s amplitude
3:40 PM
Page 25
3: The Pressure Microphone
Complex diaphragm
vibration due to radial and
tangential modes: normal
motion (A); mode 11 (B);
mode 02 (C); Neumann
center-clamped assembly
(D); rectangular assembly as
used by Pearl Microphone
Laboratories (E). (Photo
courtesy Neumann/USA.)
response becomes erratic. As a general rule, radial modes become predominant only above the normal frequency range of the microphone.
Figure 3–2D shows details of Neumann’s back-to-back diaphragm
assembly with its characteristic center and rim clamping arrangement.
Obviously, this diaphragm has a completely different set of motions than
those of the simple rim supported diaphragm. Figure 3–2E shows a rectangular diaphragm structure, with yet another set of motions at high
frequencies. A diaphragm of this type is normally about 12 mm (0.5 in)
wide and about 38 mm (1.5 in) high.
An important design variation is used by Sennheiser Electronics, and
that is the practice of not increasing the diaphragm tension to maintain
flat response at high frequencies. Instead, flat response is obtained by
electrical boosting of the diaphragm’s output at HF to whatever degree
is required.
2:37 PM
Page 26
The capacitor microphone operates on the static relationship
where Q is the electrical charge (coulombs) on the plates of a capacitor,
C is the capacitance (farads), and E is the applied dc voltage. In the
capacitor microphone, the backplate and diaphragm represent the two
plates of the capacitor.
Figure 3–3 shows a capacitor with variable spacing between the two
plates. The capacitor at A is charged by applying a reference dc voltage, E,
across it; and a charge, Q, is produced on the plates. Now, if the plates are
separated slightly, as shown at B, the capacitance is reduced. The charge
remains fixed, however, and this causes the voltage, E, to rise. This comes
as a consequence of satisfying equation (3.1). Alternatively, if we move the
plates closer together, as shown at C, the capacitance increases and the
voltage drops.
In the externally charged capacitor microphone, the applied polarizing voltage is fed through a very high resistance, as in the network shown
in Figure 3–1C. The series resistance of 1 G (109 ohms) ensures that,
once the charge is in place, normal audio frequency variations in the
diaphragm’s displacement do not alter the charge but are manifest as a
variation in voltage, which is the inverse of the change in capacitance.
For our analysis here, we assume that negligible power is transferred
between the capacitor and the resistor.
It is important that the value of the polarizing voltage be well below
the break-down voltage of the air dielectric in the capacitor capsule. This
value is approximately equal to 3000 V/mm under normal atmospheric
conditions. Studio-grade capacitor microphones typically are polarized
in the range of 48–65 volts, while instrumentation microphones may
operate in the range of 200 V. A slight electrostatic attraction between
the diaphragm and backplate is caused by the polarizing voltage, which
causes the diaphragm to be drawn very slightly toward the backplate.
Relationships in a charged
capacitor: a fixed charge is
placed on the plates of a
capacitor (A); reducing the
capacitance causes the
voltage to increase (B);
increasing the capacitance
causes the voltage to
decrease (C).
2:37 PM
Page 27
3: The Pressure Microphone
Nonlinearity of
capacitance versus
displacement (A); for a very
small displacement, the
relationship is virtually
linear (B).
For a fixed charge, the basic polarizing voltage versus capacitance
relationship is not truly a linear one. When viewed over a wide range of
capacitance change, the relation is hyperbolic, as shown in Figure 3–4A.
However, over the normally small displacement range, as shown at B,
the relationship is very nearly linear. Recall from the section above on
Analysis of the Capacitor Pressure Microphone that normal diaphragm
motion is extremely small; only for signals in the range of 130 dB LP
and greater is the diaphragm displacement in the 5% range relative to its
distance from the backplate, and this represents a change in position of
only 1 part in 20.
Shielding the microphone is very important since the high electrical
impedance of the polarizing network makes the system susceptible to
electrostatic interference. Normally, a grounded metal mesh screen separates the diaphragm assembly from the outside and provides the necessary shielding.
The value of capacitance depends on, among other things, the displacement of one plate relative to the other. Therefore, for uniform (flat) electrical output from the capacitor element, the diaphragm displacement
should be independent of frequency. In a plane progressive wave, pressure and particle velocity are in phase with each other, as discussed in
Chapter 2. The relationship between air particle displacement and velocity is given by the integral of u(t) with respect to time:
x(t) u(t)dt
where x(t) is the instantaneous particle displacement and u(t) is the
instantaneous particle velocity. For a sinusoidal signal, u(t) is of the form
ejt; therefore,
x(t) u(t)dt (j/)e
This equation describes a response that is inversely proportional to
frequency (i.e., rolls off at 6 dB per octave with increasing frequency).
2:37 PM
Page 28
The response is further modified by the j complex operator, which
shifts the relative phase by 90.
Remember that the diaphragm is under considerable mechanical tension, with additional stiffness provided by the small air chamber behind
it. When a pressure wave impinges on the diaphragm, it encounters a
mechanical responsiveness proportional to j/S, where S is the mechanical stiffness of the diaphragm (N/m). The j/S term describes a response
directly proportional to frequency (i.e., rises 6 dB per octave with increasing frequency). The response is further modified by the positive j complex
operator, which shifts the relative phase by 90. We see the two effects are
complementary and their combined effect is shown as:
(j/)ejt (j/S) ejt/S
where ejt/S now is of the same form as u(t).
Therefore, the response is flat and the net phase shift is 0 in the
range over which the diaphragm is stiffness controlled.
The entire process is shown graphically in Figure 3–5. At A, the free
field relationships among air particle velocity, pressure, and displacement
are shown. At B, the combination of air particle displacement and
diaphragm resonance is shown.
Between the ranges of diaphragm stiffness and mass control is the
region of resonance; and for most studio microphones, this ranges from
about 8–12 kHz. In most microphones, the resonance is fairly well
damped so as not to cause an appreciable rise in response; perhaps no
more than about 2 or 3 dB. Beyond resonance, the overall response of the
microphone begins to roll off at a rate of 12 dB per octave; however, the
on-axis response of the microphone tends to remain fairly flat over about
half an octave above resonance, due to diffraction effects similar to those
shown in Figure 2–14.
The output of the capacitor pressure microphone remains very flat
down to the lower frequencies, limited only by the movement of air
through the atmosphere pressure adjusting effect of the capillary tube
and the variations in charge on the capacitor plates through the very high
Forces on the diaphragm:
relations among pressure,
air particle velocity, and
air particle displacement
(A); response of a
diaphragm (B).
2:37 PM
Page 29
3: The Pressure Microphone
biasing resistor (see Figure 3–1). In most studio-quality capacitor pressure microphones, LF rolloff effects are usually seen only in the range of
10 Hz or lower.
For normal recording, broadcasting, and sound reinforcement applications using capacitor microphones, changes in response due to temperature and barometric pressure variations generally may be ignored.
However, in many instrumentation applications, the variations may be
significant enough to warrant recalibration of the measurement system.
The primary effect of temperature increase is a reduction of
diaphragm tension, which causes an increase in sensitivity and a decrease
in bandwidth. The net effect on sensitivity is quite small, roughly of the
order of 0.005 dB/C. (See equations (3.7) and (3.8).)
A decrease in barometric pressure has a slight effect on the LF and
MF (medium frequency) sensitivity of the microphone, but the decrease
in air pressure at the diaphragm at f0 produces less damping of the
diaphragm motion at resonance, causing an increase in the response at f0.
The effects of both temperature and atmospheric pressure variations
on the response of an instrumentation microphone are shown in
Figures 3–6A and 3–6B, respectively.
Although the capacitor pressure microphone is of relatively simple construction, it consists of many individual acoustical masses, compliances,
and resistances. Figure 3–7 shows a complete circuit with definitions of
all elements. The transformer represents the conversion from the acoustical domain into the electrical domain.
At LF and MF the circuit can be simplified as shown in Figure 3–8.
At HF, the circuit can be simplified as shown in Figure 3–9A, and the
response curves in Figure 3–9B show the effect of varying the damping
on the diaphragm by altering the value of RAS.
Figure 3–10A shows details of the microphone preamplifier and polarizing system for a vacuum tube design. Much of the complexity here is
determined by the external powering source for polarization and the
necessity for correct dc biasing of the vacuum tube.
Figure 3–10B shows a similar arrangement for a modern FET (fieldeffect transistor) solid-stage design. Note here the presence of a 10 dB
attenuating (padding) capacitor around the capsule; the switchable
capacitor shunts the signal generated by the capsule so that the system
2:37 PM
Page 30
Effects of temperature
variations on the sensitivity
of a capacitor pressure
microphone (A); effect of
atmospheric pressure
variations on the response
of a capacitor pressure
microphone (B). (Data after
Brüel and Kjaer, 1977.)
Equivalent electroacoustical
circuit for a capacitor
pressure microphone
(impedance analogy).
(Data after Beranek,
2:37 PM
Page 31
3: The Pressure Microphone
LF and MF response,
simplified equivalent
circuit. (Data after
Beranek, 1954.)
HF response, simplified
equivalent circuit (A);
response (B). (Data at A
after Beranek, 1954.)
can operate at higher sound levels without causing electrical overload in
the following preamplification stage.
The shunt capacitor is normally chosen to have a value of about 0.4
that of the capsule itself. The parallel combination of the two capacitors
then has a value of about one-third that of the capsule itself, resulting in
an attenuation of output of about 10 dB. The action of the shunt capacitor is not like that of a simple voltage divider; rather, a slight signal nonlinearity is caused by the circuit.
Let x represent a variable corresponding to the normal signal incident on the capsule. Then xC represents the varying value of capsule
capacitance without the shunt. The parallel combination of the capsule
and the shunt has a net capacitance equal to
Cnet x 0.4C2
0.4C Cx
which is of the form
Cnet Ax
B Cx
where A 0.4C2, B 0.4C, and C is the capacitance of the capsule.
2:37 PM
Page 32
Details of capacitor
microphone preamplifiers:
vacuum tube type (A);
solid-state type (B).
Equation 3.5 can be expanded to
Cnet ABx ACx2
B2 (Cx)2
In this form, we see that the simple relationship Cx, which represents the
variable capacitance of the unattenuated capsule, has been transformed
into a combination that includes squared terms in both numerator and
denominator, indicating the presence of second harmonic distortion.
The effect of this is small, considering the minute signal variations
represented by x. In any event, the distortion introduced by the shunt
capacitance is negligible under the high sound pressure operating circumstances that would normally call for its use.
Signal output at the capsule can also be attenuated by dropping the
dc bias voltage padding in this manner. The Neumann models TLM170
and TLM50 achieve their capsule padding in this manner. The TLM50
is padded by changing the value of dc bias voltage on the capsule from
60 to 23 V. This causes a slight alteration in the microphone’s HF
response due to the reduced electrostatic attraction between diaphragm
2:37 PM
Page 33
3: The Pressure Microphone
and backplate, which causes the diaphragm to move slightly farther
away from the backplate.
Regarding the output stage of the microphone’s built-in preamplifier,
the typical electrical output impedance is in the range of 50–200 ohms.
The output is normally balanced, either with a transformer or by balanced solid-state output circuitry, to ensure interference-free operation
over long transmission paths downstream. The electrical load normally
seen by the microphone preamplifier is in the range of 1500–3000 ohms,
which approximates essentially an unloaded condition if the driving
source has a low impedance. Microphone wiring runs with low capacitance cable may extend up to 200 m with negligible effect on response.
As we saw in Figure 2–14, a free progressive plane wave arriving along
the primary axis of a cylinder or a sphere shows a considerable rise in
HF response. For the cylinder, the maximum on-axis rise will be 10 dB
relative to the response at 90; by comparison, the sphere produces only
a 6 dB on-axis rise. In both cases, the response at 90 is very nearly flat
and, in fact, approximates the integrated response of the microphone in
a random sound field.
In the design of pressure microphones for both instrumentation and
recording the diffuse random field. The smaller the diameter of the obstacle, the higher in frequency the divergence between random and on-axis
response, as shown in Figure 3–11. Here, the microphones have been
designed for essentially flat on-axis response.
Typical on-axis and
random response of 12 mm
(05 in) and 25 mm (1 in)
pressure microphones.
2:37 PM
Page 34
Figure 3–12 shows how a Brüel & Kjaer 4000 series omnidirectional
microphone can be modified in HF response by changing the protective
grid. The data shown at A is for the normal version of the microphone,
with flat on-axis response (6 dB at 40 kHz) and rolled-off random incidence response. When the normal grid is replaced with an alternate
grid that produces less diaphragm damping, the response in panel B is
produced. Here, the random incidence response is flat to 15 kHz and the
on-axis response exhibits a HF peak of about 6 dB.
The same microphone, with its normal grid replaced by a nose cone
baffle providing indirect access to the diaphragm, produces the response
shown at C for all directions of sound incidence.
The Neumann M50 microphone, designed during the mid-1900s,
consists of a 12 mm diameter capacitor element mounted on a sphere
Microphones designed
for specific pickup
characteristic: flat on-axis
response with rolled-off
random incidence response
(A); flat random incidence
with peaked response
on-axis (B); use of a special
nose cone baffle to produce
random incidence response
in all pickup directions (C).
(Data after Brüel and Kjaer,
2:37 PM
Page 35
3: The Pressure Microphone
Neumann M50 data,
including sketch of
capacitor capsule mounting
(A); polar response (B); and
on-axis frequency response
(C). (Figure courtesy
approximately 40 mm in diameter, approximating the spherical response
shown in Figure 2–14 for a diameter/wavelength corresponding to about
8200 Hz. Note the very close agreement between the on-axis response of
the M50 (Figure 3–13) and the data shown in Figure 2–14, when the
normalized unity value on the frequency axis is equated to 8200 Hz.
The M50 exhibits a flat diffuse field response but depends on the
sphere to give added HF presence to signals arriving on-axis. This microphone has never lost its popularity with classical recording engineers and
often is used in the transition zone between the direct and reverberant
fields for added presence at high frequencies.
The Sennheiser MKH20, a flat on-axis pressure microphone, can be
converted to essentially flat random incidence response by engaging an
internal electrical HF shelving boost. It also can be given preferential onaxis directivity by the addition of a small rubber ring mounted at the end
of the microphone. Details are shown in Figure 3–14. (The Sennheiser
MKH20 employs a radio frequency signal conversion system, which we
will discuss in Chapter 8. In terms of directional response, RF capacitor
microphones have the same characteristics as the other capacitor models
discussed in this chapter.)
Overall, the effects of diffraction and off-axis wave interference at
the diaphragm of a capacitor pressure microphone placed at the end of
a 21 mm diameter body produces the polar response versus frequency as
shown in Figure 3–15.
The capacitor pressure microphone represents a careful balance of technical attributes. It reached an advanced level of overall performance
2:37 PM
Page 36
Sennheiser MKH20 data:
response with and without
baffle ring (A); response
with and without baffle
ring, electrically boosted at
HF (B); profiles of
microphone with and
without baffle ring (C).
(Data courtesy Sennheiser.)
Set of polar curves for
a pressure microphone
placed at the end of a
21 mm diameter cylinder,
exhibiting narrowing of
response patterns at HF.
(Data after Boré, 1989.)
during the middle 1980s, and today’s many models invariably represent
different “operating points” among the variables of low noise, distortion
at high levels, bandwidth extension, and polar pattern control.
With the coming of digital recording, the self-noise of a capacitor
capsule and its associated preamplifier have come under close scrutiny,
since the effective noise floor of the Compact Disc and other higher
2:37 PM
Page 37
3: The Pressure Microphone
density formats is vastly lower than that of the long-playing stereo disc.
Microphone noise levels that were acceptable in the predigital era may
not be acceptable today. Over the years, we have seen the weighted noise
floors of studio-quality capacitor microphones drop by 10–12 dB.
The self-noise floor of a capacitor microphone is normally weighted
according to the A curve shown in Figure 3–16 and stated as an equivalent acoustical rating. For example, a noise rating of 10 dB(A) indicates
that the noise floor of the microphone is approximately equivalent to
that which a theoretically perfect microphone would pick up if it were
operating in an actual acoustical space that had a residual noise rating of
10 dB(A), or its equivalent value of NC 10.
These matters are discussed in detail in Chapter 7, which is devoted
to microphone measurements; here, we intend only to detail the spectral
nature of the noise itself.
Figure 3–17 shows the composite one-third octave noise spectra
of a preamplifier with input capacitances equivalent to 25 mm (1in)
and 12 mm (0.5 in) instrumentation microphones. Note that as the
diaphragm diameter decreases by one half the noise floor rises approximately 6 dB. However, as a tradeoff of performance attributes, each
Standard weighting
curves for acoustical noise
One-third octave noise
spectra at the output of a
microphone preamplifier
with input loaded by
capacitors equivalent to
25 mm and 12 mm
diameter capsules.
2:37 PM
Page 38
succeeding halving of diameter produces a microphone with twice the
HF bandwidth capability of the preceding one. Fortunately, the ear’s
sensitivity to low frequencies is diminished at low levels, as indicated by
the weighting curves shown in Figure 3–16.
The curves show a rise below about 1 kHz of approximately 10 dB
per decade with decreasing frequency. This rise is usually referred to as
1/f noise (inversely proportional to frequency) and is primarily a property of the electronics. The flattening of the spectrum and eventual slight
upturn at HF largely is the result of the capacitor element itself and the
reflection of its mechano-acoustical resistance into the electrical domain.
Newer electronics have reduced l/f noise considerably; however, this
is a minor improvement when we consider the relative insensitivity of the
ear to low frequencies at low levels. Typical microphone preamplifier
output noise spectra are shown in Figure 3–17.
Electret materials have been known for more than a century, but only in
the last 35 years or so has the electret had an impact on capacitor microphone design, bringing excellent performance to very low-cost models.
The electret is a prepolarized material, normally polytetrafluoroethylene,
which has been given a permanent electrostatic charge through placement in a strong electric field and under heat. As the heat is withdrawn,
the electric charge remains. The material is virtually the electrostatic
equivalent of a permanent magnet, and if a capacitor backplate is coated
with one of the newer electret materials, the resulting microphone will
have the same performance characteristics as a standard capacitor capsule with an effective polarizing value of about 100 V. Figure 3–18A
shows a section view of a capacitor capsule using an electret backplate.
Alternatively, an electret diaphragm can be used, as shown in
Figure 3–18B. One drawback here is that the metallized electret foil has
somewhat greater mass per unit area than typical non-electret diaphragm
materials, possibly compromising HF response.
Figure 3–19A shows a typical electrical implementation of the electret; note the simplicity of the design. A photograph of a small electret
microphone is shown in Figure 3–19B.
The electret capsule:
electret backplate (A);
electret coated
diaphragm (B).
2:37 PM
Page 39
3: The Pressure Microphone
Complete electrical circuit
for an electret microphone
(A); photo of a small
electret capsule (B). (Photo
courtesy of AKG
Pressure capacitor microphones are generally used for all instrumentation and measurement purposes, and as such they differ slightly in design
from those microphones intended for recording and broadcast use.
Figure 3–20 presents a cutaway view of a Brüel & Kjaer instrumentation
capsule showing a perforated backplate somewhat smaller in diameter
than the diaphragm, exposing a back air cavity normally not present in
microphones designed for the studio.
Instrumentation microphones are used in the loudspeaker industry,
architectural acoustics, and as a primary tools for acoustical response
and noise measurements in industrial workplaces.
The most ubiquitous use of these microphones is in the sound level
meter (SLM), shown in Figure 3–21A. The SLM is the primary instrument for sound survey and noise measurement in our everyday environment. (For routine home sound system setup and level checking, there
are a number of low cost sound survey meters, most of them in the $40
range and available from electronics stores.)
The instrumentation microphone is designed for very uniform
response, unit to unit, and is provided with accessories to facilitate calibration in the field for variations in temperature and barometric pressure. Polarizing voltages are normally in the 200 V range, which allows
somewhat greater separation between diaphragm and backplate, ensuring lower distortion at higher operating levels.
Instrumentation microphones are available in many sizes to cover
the frequency range from well below 1 Hz to 200 kHz and beyond,
depending on the application. There are models that respond to seismic
air movements and have a self-noise floor of 2 dB(A); other models are
designed to work at sound pressure levels in the range of 180 dB LP and
above. Figure 3–21B shows the sound pressure level operating ranges of
several studio and instrumentation microphones.
2:37 PM
Page 40
Cutaway view of an
microphone. (Figure
courtesy of Brüel and
Kjaer, 1977.)
A modern digital sound level meter (A); operating ranges of selected studio and
instrumentation microphones (B). (Photo courtesy of Brüel and Kjaer, 1977.)
2:37 PM
Page 41
3: The Pressure Microphone
Kinsler et al. (1982) present the following simplified equation for the
base sensitivity of a capacitor capsule:
volt per pascal
E0 polarizing voltage (V)
h nominal distance from diaphragm to backplate (m)
a radius of diaphragm (m)
T0 effective diaphragm tension (N/m)
Using the following constants for a 25 mm (1 in) diameter instrumentation microphone:
a 8.9 10–3 m
T0 2000 N/m
E0 200 volts
h 25 106 m
m (surface mass density of diaphragm) 0.0445 kg/m2
we can calculate approximate values of microphone sensitivity and
upper frequency limit:
0.04 V/Pa
Using a more accurate equation, Zukerwar (1994) arrived at a closer
value of 0.03 V/Pa. The simplified equation gives us only a reasonable
Moving on to an equation for the upper frequency limit (fH) of the
microphone, we have (Wong and Embleton, 1994):
fH fH 2.4
6.28(8.9 103)
9150 Hz
Above fH, the response of the microphone is governed by the degree of
damping at the resonance frequency and diffraction effects. The response
on axis can extend fairly smoothly up to about 20 kHz.
2:37 PM
Page 42
The LF bandwidth extends to very low frequencies, limited only by
the RC (resistance-capacitance) time constant established by the low
capacitance of the pressure element and the high value of the series resistance. For the 25 mm diameter microphone discussed here, these values
are C 15 1012 farads and R 109 . Calculating the rolloff
frequency, fL:
fL 1/(2RC) 1/6.28(109 15 1012)
fL 1/(6.28 15 103) 10.6 Hz
The diaphragm rms displacement in meters is given by:
Displacement pa2
For the 25 mm diameter diaphragm under discussion here, the rms displacement is:
Displacement (8.9 103)2 5 1011 m
It is evident that the diaphragm displacement in the instrumentation
microphone discussed here is considerably less than in the studio microphone discussed above.
The dynamic microphone is based on principles of electricity and magnetics
dating from the 19th century. Because of their relatively low output, it was
not until the coming of electrical amplification that these microphones
found their place in commerce and early radio broadcasting. Today, we
find relatively few dynamic pressure microphones in widespread use, their
role having been taken over for the most part by the electret.
The dynamic microphone is also referred to as the electrodynamic,
electromagnetic, or moving coil microphone. It is based on the principle
of magnetic induction in which a conductor, or wire, moving across a
magnetic field has induced along it a voltage proportional to the strength
of the magnetic field, the velocity of the motion, and the length of the
conductor crossing the magnetic field. The governing equation is:
e(t) Blu(t)
where e(t) is the instantaneous output voltage (V), B is the magnetic flux
density (T), l is the length of the conductor (m), and u(t) is the instantaneous velocity of the conductor (m/s). Since B and l are constant, the output voltage is directly proportional to the instantaneous velocity of the
The basic principle of magnetic induction is shown in Figure 3–22,
which shows the vector relationship among flux density, conductor
orientation, and conductor velocity. In a microphone, the normal
2:37 PM
Page 43
3: The Pressure Microphone
Basic principle of magnetic
Section view (A) and front
view (B) of a moving coil
microphone assembly
(protection grid removed
in front view).
realization is in the form of a multi-turn coil of wire placed in a radial
magnetic field. Section and front views of a typical dynamic pressure
microphone are shown in Figure 3–23. In this form, the action is the
reciprocal of the modern dynamic loudspeaker.
In a plane progressive sound field, air particle velocity and pressure are
in phase. Therefore, for flat electrical output across the conductor, the conductor must execute constant velocity across the entire frequency band in
a constant pressure sound field. Since the coil/diaphragm assembly consists
of mechanical mass and compliance, it will exhibit mechanical resonance;
and this is normally designed to be near the geometric mean of the
intended frequency response of the microphone. The geometric mean
between two quantities along the same numerical scale is defined as:
Geometric mean lower quantity higher quantity
For a typical response extending from 40 Hz to about 16 kHz, the design
resonance is:
f0 40 16,000 800 Hz
The microphone is thus designed to be resistance controlled over its useful frequency range through the application of external damping of the
diaphragm’s motion.
Figure 3–24 shows the response of an undamped diaphragm (curve 1),
along with the effects of increased damping (curves 2 through 5). Note
2:37 PM
Page 44
Basic diaphragm
response control, the effect
of externally damping the
that, as the damping increases, the midband output becomes flatter, but at
the expense of midband sensitivity. To achieve satisfactory frequency
response while maintaining a useful overall sensitivity, several design techniques are employed. The diaphragm is damped by placing layers of silk
or thin felt in the region of the gap that separates the coil from the back of
the air chamber. In general, the MF resonance peak is reduced by about
25–35 dB, producing the response shown in curve 5. To add more
midrange damping in this manner would result in unsatisfactory output
The next step is to address the LF response falloff, and this is accomplished by inserting a long, narrow tube into the back air chamber exiting
to the outside air. The tube dimensions are chosen so that the air mass in
the tube will resonate with the compliance provided by the internal air
chamber itself. This Helmholz resonance is chosen to be at some frequency
in the 40–100 Hz range. A final step is to compensate for the HF falloff,
and this is done by creating a small resonant chamber just inside the
diaphragm, tuned in the range 8–12 kHz. These added LF and HF resonances boost the microphone’s output in their respective frequency
The design process of damping the primary diaphragm resonance,
while adding both LF and HF peaks in the response can best be seen
through analysis of the equivalent circuit (after Beranek, 1954), as
shown in Figure 3–25.
At LF the response is largely determined by the resonance set up by
the bass equalization tube and the back air chamber, and the equivalent
circuit in this frequency range is shown in Figure 3–26A.
At MF the response is governed by the damping provided by the silk
layers just below the coil, and the equivalent circuit in this frequency
range is shown in Figure 3–26B.
At HF the response is determined largely by the resonance set up by
the mass of the diaphragm and the compliance of the front air chamber
2:37 PM
Page 45
3: The Pressure Microphone
Simplified equivalent
circuit of the dynamic
pressure microphone.
(Data after Beranek, 1954.)
directly below it. The equivalent circuit in the frequency range is shown
in Figure 3–26C, and the net overall response is shown in Figure 3–26D.
We can see that there is much design leeway here, and some pressure
dynamic microphones have been designed with more resonant chambers
than we have shown here. With care in manufacturing, response tolerances for dynamic pressure microphones can be maintained within a
range of 2.5 dB from 50 Hz to about 15 kHz, as measured on axis.
We can appreciate that the design of a good dynamic pressure microphone results from a combination of physics, ingenuity, and years of
design practice. Such questions as where to place damping materials and
where to assign the LF and HF resonances require experience, and the
many variables provide for tailored response for a given microphone
application. See Souther (1953) for a practical approach to the design of
a dynamic pressure microphone for general application. See also Beranek
(1954) and Kinsler et al. (1982).
The microphone diaphragm may be made of duralumin, a stiff, lightweight alloy of aluminum, or one of many stable plastic materials that
can be molded to tight tolerances in thin cross-section. The dome normally moves as a unit, with mechanical compliance provided by reticulated treatment in the outer detail of the dome. Small neodymium
magnets are very common in modern designs. Typical physical values for
2:37 PM
Page 46
Designing for low,
medium, and high
frequency regimes: LF
resonance (A); MF
damping (B); HF resonance
(C); net response (D). (Data
at A, B, and C after
Beranek, 1954.)
a dynamic pressure microphone are:
B 1.5 T (tesla)
l (length of wire in gap) 10 m
Diaphragm/coil radius 9 mm
Diaphragm displacement for 94 dB LP at 1 kHz 2 102 m rms
2:37 PM
Page 47
3: The Pressure Microphone
Electrical circuit showing
hum-bucking coil in a
dynamic microphone.
Mass of moving system 0.6 grams
Electrical impedance 150–125 Sensitivity range 1.5–2.5 mV/Pa
Many dynamic microphones, both omnidirectional and directional,
incorporate a “hum-bucking” coil to cancel the effect of stray power line
induction at 50/60 Hz. This is a low-resistance coil in series with the
voice coil and located externally to the magnet structure of the microphone. It is wound reversely and located along the same axis as the voice
coil. It is designed to produce the same induced signal that the voice coil
would pick up in the stray alternating magnetic field; thus the two
induced signals will cancel. The hum-bucking coil is shown in the
schematic of Figure 3–27.
Numerous other transducing principles have been used in the design of
pressure microphones, including the following:
Loose contact principle. These methods were discussed in Chapter 1
since they were key items in the early history of microphone technology. Only the loose carbon particle microphone has remained a viable
design in telephony over the years, and those readers interested in its
ongoing development are referred to the Bell Telephone Laboratories
(1975) book, History of Engineering and Science in the Bell System.
Magnetic armature techniques. This technology was described in
Chapter 1 as an essential part of Bell’s original telephone patent, but
the microphone application is not current today. Olson (1957) discusses it in detail. The system is bilateral and can be employed as
either a microphone or a loudspeaker.
Electronic techniques. Olson (1957) describes a microphone in
which an electrode in a vacuum tube is mechanically actuated by an
external diaphragm, thus varying the gain of the vacuum tube. The
technique is entirely experimental and has not been commercialized.
2:37 PM
Page 48
Details of an optical
microphone. (Data after
Magnetostriction. Certain magnetic materials undergo a change in
dimension along one axis under the influence of a magnetic field.
The technique has applications in mechanical position sensing.
Optical techniques. Sennheiser has demonstrated an optical microphone in which a light beam shines on a diaphragm with a modulated
beam reflecting onto a photodiode. Details are shown in Figure 3–28.
The device can be made quite small and has no electronic components
at the diaphragm (US Patent 5,771,091 dated 1998).
Thermal techniques. Olson (1957) describes a microphone consisting of a fine wire heated by direct current. Air currents due to sound
propagation around the wire change its resistance proportionally. If
the wire can be biased by direct current air flow, then minute
changes in wire resistance due to sound pressure in the direction of
the air flow can be detected. The technique is very promising in the
area of acoustical intensity measurement.
During the 1930s to the 1960s, the piezoelectric, or crystal, microphone
was a significant factor in the market for home recording and small-scale
paging and sound reinforcement activities. Today, low-cost dynamic and
electret microphones have all but displaced the piezoelectric microphone.
(The term piezo comes from the Greek piezen, meaning pressure.)
Certain crystalline materials, such as Rochelle salts (potassium
sodium tartrate), ADP (ammonium dihydrogen phosphate), and lithium
sulfite, exhibit the property of developing a voltage along opposite sides
when they are flexed or otherwise subjected to deformation. The effect is
bilateral, and the elements can also be used as HF loudspeakers.
2:37 PM
Page 49
3: The Pressure Microphone
Details of a piezoelectric
bimorph element (A);
section view of a
microphone (B).
The crystal structures must be cut and arrayed along the appropriate crystalline axes to produce the desired output voltage. Most piezoelectric microphones are assembled from so-called bimorph piezo
structures in which adjacent crystal elements are cemented to each other
in an opposite sense so that a push-pull output result. Figure 3–29A
shows a typical bimorph structure, consisting of two crystal elements
cemented together with metal foil on each conducting surface. The element is secured on three corners, and the free corner is driven by the
diaphragm through a connecting member.
Figure 3–29B shows a section view of a typical crystal microphone.
The diaphragm motion is connected to the free corner of the bimorph,
and the output voltage is fed downstream to a preamplifier. Since the
output voltage from the bimorph is proportional to signal displacement,
the diaphragm is normally tuned to a high frequency, with external
mechanical damping, as shown in the figure.
Typical crystal microphones exhibit output sensitivities in the midband of about 10 mV/Pa and, as such, can be used to drive high-impedance preamplifiers directly. The cable length between microphone and
preamplifier is limited because of shunt capacitance HF losses in the
In the areas of hydrophones (sensors used for underwater sonar signal transmission and detection), mechanical positioning, and very high
acoustical level measurements, piezo electric elements are useful, in that
their mechanical impedance characteristics are better suited for these
applications than for normal audio use.
Those readers wanting more information on piezoelectric microphones are referred to Beranek (1954) and Robertson (1963).
2:38 PM
Page 50
The pressure gradient microphone, also known as the velocity microphone,
senses sound pressure at two very closely spaced points corresponding to
the front and back of the diaphragm. The diaphragm is thus driven by the
difference, or gradient, between the two pressures. Its most common form,
the dynamic ribbon (diaphragm) microphone, was developed to a high
degree of technical and commercial success by Harry Olson of RCA during
the 1930s and 1940s and dominated the fields of broadcasting and recording in the US through the mid-1950s. The BBC Engineering Division in the
United Kingdom was also responsible for significant development of the
ribbon, extending its useful response to 15 kHz at the advent of FM radio
transmission in the mid-1950s.
The use of the term velocity derives from the fact that the pressure
gradient, at least at long wavelengths, is virtually proportional to air particle velocity in the vicinity of the diaphragm or ribbon. However, the
more common term is pressure gradient, or simply gradient, microphone. The basic pressure gradient microphone has a “figure-8”-shaped
pickup pattern and its pickup pattern is often referred to as bidirectional.
By way of review, the basic action of the pressure microphone is shown
in Figure 4-1A and B. The mechanical view shown at A is equivalent to
the physical circuit shown at B. The significance of the plus sign in the
circle is that a positive pressure at the microphone will produce a positive voltage at its output. The circle itself indicates that the microphone
is equally sensitive in all directions.
2:38 PM
Page 51
4: The Pressure Gradient Microphone
A mechanical view of the gradient microphone is shown at C. It is
open on both sides of the diaphragm and is symmetrical in form. The
equivalent physical circuit is shown at D. Here, there are two circles, one
positive and the other negative, which are separated by the distance d.
The significance of the negative sign is that sound pressure entering the
back side of the microphone will be in polarity opposition to that entering from the front. The distance d represents the equivalent spacing
around the microphone from front to back access to the diaphragm.
Definition of the pressure
gradient. Mechanical view
of pressure microphone (A);
physical circuit of pressure
microphone (B); mechanical
view of pressure gradient
microphone (C); physical
circuit of pressure gradient
microphone (D); 0 sound
incidence on pressure
microphone (E); sound
incidence on pressure
gradient microphone (F).
2:38 PM
Page 52
Sound incident on the pressure microphone along the 0 axis will
enter only from the front, as shown at E, while sound incident on the gradient microphone along the 0 axis will enter by the front opening as well
as through the back opening delayed by the path d. This is shown at F.
The gradient microphone has the capability of determining the propagation angle of a wave. This can be seen in the upper part of Figure 4–2A.
A portion of a plane wave is shown along with locations of the two microphone openings placed in line with the wave propagation. These two points
relate directly to the points indicated on the sine wave shown below and
represent the two openings of the gradient microphone. Now, if the two
openings are positioned as shown in Figure 4–2B, there will be no pressure
gradient between them and thus no signal at the microphone’s output.
The pressure gradient increases with frequency, as shown in
Figure 4–3. Here, we show the gradients that exist at LF, MF, and HF
when the pressure sensing points are in line with wave propagation. For
long wavelengths, the gradient is directly proportional to frequency and
is of the form:
gejt K1 jpejt
where gejt represents the instantaneous value of the gradient, K1 is an
arbitrary constant, and pejt is the instantaneous value of pressure. If the
air particle velocity is uniform with respect to frequency, the K1j multiplier in the right side of equation (4.1) indicates that the value of the
pressure gradient will be proportional to frequency, rising at a rate of
6 dB/octave, and that the phase of the pressure gradient will be advanced
90 with respect to pressure. This can in fact be seen through a study of
the graphical data shown in Figure 4–3.
For a pressure gradient microphone to exhibit flat response in a uniform pressure field, there must be some form of equalization present that
Pressure gradient measured
longitudinally in a
progressive plane wave is
equal to p2–p1 (A); when
placed transversely in the
progressive plane wave the
gradient is zero (B).
2:38 PM
Page 53
4: The Pressure Gradient Microphone
Pressure gradient at LF,
MF, and HF.
exactly counteracts the HF rise of the pressure gradient. In the case of the
dynamic gradient microphone the electrical output is:
E ejt B1Uejt
where E is the rms output voltage of the microphone, B is the flux density in the magnetic gap (tesla), l is the length of the moving conductor in
the magnetic field (m), and U is the rms velocity of the conductor (m/s).
This output voltage would tend to rise 6 dB/octave at HF if the
microphone were placed in a sound field that exhibited rising HF air particle velocity. The required equalization can be done electrically, but is
accomplished very simply by using a mass-controlled ribbon or diaphragm
of the form K2/j. The combination of the pressure gradient and the
mechanical HF rolloff is then:
Net response(K1 pjejt)/(K2/j)(K1/K2)pejt
which is now proportional to the rms pressure and is flat over the
frequency band.
With the capacitor form of the gradient microphone, flat electrical output in a flat sound pressure field depends on constant displacement of the
diaphragm over the entire frequency range. Since the driving pressure gradient rises 6 dB/octave at HF, it will produce a constant particle displacement with frequency at the capacitor diaphragm. Therefore, if the capacitor
diaphragm’s mechanical damping is large compared to its compliance and
mass reactance, then the required equalization will be attained.
Figure 4–4A shows the basic “figure-8” response of the gradient
microphone. It is important to note that the response is maximum at
2:38 PM
Page 54
0 and 180, but that the polarity of the signal is negative (reversed) in
the back hemisphere relative to the front hemisphere. Response at
45 is 0.707, which is equivalent to 3 dB. Response at 90 is
effectively zero. The directional response equation in polar coordinates is:
cosine (4.4)
where represents the magnitude of the response and represents the
polar angle. Figure 4–4B shows the response plotted in decibels.
The frequency range over which the desired polar response is maintained depends on the size of the microphone and the effective distance
between the two sensing points, or distance from the front of the
diaphragm around the assembly to the back of the diaphragm. The first
null in response takes place when the received frequency has a wavelength that is equal to the distance, d, between the two openings of the
The pressure gradient, as a function of frequency, is shown in
Figure 4–5A. The response rises 6 dB per octave, falling to a null in
response at the frequency whose wavelength is equal to d, the path
around the microphone. The theoretical equalized microphone output is
shown at B. However, due to diffraction effects, the actual response at
HF may resemble that shown by the dashed line in the figure.
For extended frequency response, we would like d to be as small as
possible. If d/ 1/4, the drop in response will only be 1 dB (Robertson,
1963). Solving this equation for a frequency of 10 kHz gives a value of
d 8.5 mm (0.33 in). However advantageous this might be in terms of
frequency response, the very short path length would produce a relatively small pressure gradient, and the resultant output sensitivity of the
microphone would be low. Fortunately, diffraction effects work to our
advantage, maintaining excellent on-axis frequency response even when
a substantially longer front-to-back path is used. This will be discussed
in the following section.
Directional response in
polar coordinates of the
gradient microphone;
sound pressure gradient
response (A); response level
in decibels (B).
2:38 PM
Page 55
4: The Pressure Gradient Microphone
Pressure gradient force
versus frequency,
unequalized (A) and
equalized (B), for a typical
ribbon microphone (typical
on-axis HF response shown
by dashed curve).
Mechanical and electrical views of a ribbon gradient microphone are
shown in Figure 4–6. The ribbon itself is normally made of corrugated
aluminum foil and is about 64 mm (2.5 in) in length and 6.4 mm (0.25 in)
in width. It is suspended in the magnetic gap with provisions for adjusting its tension. In newer designs the thickness of the ribbon is often in
the range of 0.6 microns (2.5 105 in), and the length in the range of
2:38 PM
Page 56
25 mm (1 in). The mass of the ribbon, including its associated air load, is
about 0.001 to 0.002 grams, and the subsonic resonance frequency of
the ribbon is in the range of 10 to 25 Hz. Because of the small gap
between the ribbon and polepieces, there may be considerable viscous
damping on the ribbon in the region of resonance. Flux density in the
gap is normally in the range of 0.5 to 1 T (tesla).
Early ribbon microphones were quite large, with heavy, relatively
inefficient magnets. The fragility of the ribbon microphone was well
known to all who used them; some ribbons tended to sag slightly over
time, and a good puff of wind could actually deform them. Because of
their flexibility, ribbons may, like a string, vibrate at harmonic multiples
of their fundamental frequency (Robertson, 1963). This is not often
encountered in well-damped designs, however.
A closer look at the workings of a ribbon microphone will often reveal
a careful use of precisely designed and shaped screens, made of fine metal
mesh or of silk, whose purposes are to make fine adjustments in frequency
response. As shown in Figure 4–7A, a metal mesh screen has been mounted
around the polepiece assembly in order to compensate for LF losses due to
the high damping on the ribbon in the region of resonance.
The fine mesh of the structure has been chosen to provide a fairly
uniform acoustical impedance over the operating frequency range of the
microphone. The vector diagram at B shows its effect at LF. The loss
introduced by the screen over the path d results in a lower value of
pressure (p2) at the rear of the ribbon. This actually produces a larger
A ribbon gradient
microphone; front view (A);
top section view (B);
perspective view (C);
electrical circuit (typical)
associated with the ribbon
microphone (D). (Data at
B and C after Beranek,
2:38 PM
Page 57
4: The Pressure Gradient Microphone
Correcting LF response
in a gradient microphone
through acoustical
attenuation between front
and back. Microphone with
screen (A); vector diagram
at LF (B); vector diagram at
HF (C).
gradient (p) at LF than would exist without the screen (p), and the net
effect is a compensating boost at LF.
At MF and HF, the screen’s action is as shown at C. Here, the effect
of the screen is to cause a slight decrease in the gradient (p), which may
be a small price to pay for correcting the LF response. This design is used
to good effect in the Coles model 4038 microphone and is described in
detail by Shorter and Harwood (1955).
There are four main factors affecting the base sensitivity of the ribbon:
magnetic flux density in the gap, length and mass of the ribbon, and
the length of the front-to-back path. Aluminum foil, since it possesses the
best combination of low mass and low electrical resistivity, is perhaps the
ideal material for the application. The aspects of internal dimensioning
are all interrelated, and any attempt to optimize one will likely result in
a compromise elsewhere. For example, doubling the length of the ribbon
2:38 PM
Page 58
will raise the sensitivity by 6 dB but will adversely affect vertical directional response at high frequencies. Doubling the path length around the
polepieces would also provide a 6-dB increase in sensitivity but would
compromise the HF performance of the microphone because of the overall increase in microphone size.
The area where many modern improvements have taken place is in
magnetics. The use of higher energy magnets, along with new magnetic
circuit materials and topology, has made it possible to achieve enough
improvement in base sensitivity to allow further beneficial tradeoffs,
including shorter ribbons, with their better behaved polar response and
HF performance. In most magnetic designs, the choice and dimensioning
of materials is made so that the flux through the polepieces is effectively
saturated, and thus uniform throughout the length of the gap.
Because of the extremely low electrical output from the ribbon, a
step-up transformer is normally installed in the microphone case. A
transformer turns ratio of about 20:1 can be used to match the very low
impedance of the ribbon to a 300 ohm load. The open circuit sensitivity
of a typical ribbon microphone, operating without an output transformer, is normally in the range of 0.02 mV/Pa. The addition of a stepup transformer raises the system open circuit sensitivity to the range of
The design simplicity of the ribbon microphone produces a predictable,
uniform pattern up to the frequency where d (5/8). Olson (1957)
presents the theoretical polar data shown in Figure 4-8. Note that as the
wavelength approaches the internal dimensions of the microphone, the
pattern flattens slightly on-axis, eventually collapsing to produce zero
on-axis response, with polar response of four equal off-axis lobes.
Figure 4–9 shows on-axis response curves for typical commercial ribbon gradient microphones. The curve shown at A is from RCA product
literature for the model 44-B ribbon microphone. The data shown at B
Polar response of a
theoretical ribbon gradient
microphone as a function
of and d. (Olson, 1957.)
2:38 PM
Page 59
4: The Pressure Gradient Microphone
Typical frequency response
curves for ribbon gradient
microphones RCA 44-B
(A); Coles 4038 (B).
is for the Coles model 4038 microphone. Note that the Coles exhibits
slightly flatter and more extended HF response.
Figure 4–10A shows details of the beyerdynamic Model M 130 ribbon microphone. It is perhaps the smallest, and flattest, ribbon microphone available today, with the entire ribbon assembly contained within
a sphere 38.5 mm in diameter. Details of the ribbon structure are shown
at B, and the microphone’s response at 0 and 180 is shown at C.
The more recent AEA model R-84 has much in common with the
earlier RCA designs in that it has a large format ribbon. This model is
shown in Figure 8–11A, with front and back response curves at B. Note
that the vertical scale is expanded and that the response, though
extended to 20 kHz, has very much the same general contour as the
response curve of the RCA 44.
Figure 4–12A shows the equivalent circuit for a ribbon gradient microphone using the impedance analogy (Beranek, 1954). Ignoring the effects
of the slits between the ribbon and polepieces and damping in the
system, we can simplify the circuit to the form shown at B. Here, the
5:39 PM
Page 60
Beyerdynamic M 130
ribbon microphone:
photo (A); details of ribbon
structure (B); frequency
response (C). (Data courtesy
of beyerdynamic.)
AEA R-84 ribbon
microphone: photo (A);
front and back response
curves (B). (Data courtesy
of AEA.)
2:39 PM
Page 61
4: The Pressure Gradient Microphone
Equivalent circuits for a
ribbon gradient
microphone; acoustical
circuit (A); simplified
electroacoustical circuit (B).
(Data after Beranek,
2:39 PM
Page 62
mobility analogy is used to give an equivalent electromechanical circuit
over the range from 50 Hz to about 1 kHz. As can be seen from the circuit, the microphone presents very nearly a pure mass reactance to the
acoustical driving signal.
The output voltage is given by:
S cos (
e0 u
where u is the component of air particle velocity perpendicular to the ribbon, d is the path length around the microphone, S is the effective area
of the ribbon, and 2MMA MMR represents the mass of the ribbon and
its associated air load on both sides.
A well-designed ribbon gradient microphone can handle fairly high
sound pressure levels in normal studio applications. The moving system
is however displacement limited at LF due to the constraints of its design
resonance frequency.
Figure 4–13 shows three forms of capacitor gradient microphone. The
single-ended form shown at A is most common, but its asymmetry does
cause a slight HF response difference between front and back. Some
designers will place a dummy perforated electrode, which is not polarized,
at the front to compensate for this, as shown at B. The push-pull design,
shown at C, doubles the output voltage for a given diaphragm excursion
relative to the other forms, but is complicated by the required biasing
As we discussed in the section entitled Definition and Description of
the Pressure Gradient, the capacitor gradient microphone operates in a
resistance-controlled mode. This requires that the diaphragm has an
undamped resonance in the midband and that considerable viscous air
damping be applied to it by the many small holes through the backplate
or backplates. There is a practical limit to the amount of damping that
is applied, however. Figure 4-14 shows the damping applied progressively, and a typical operating point is indicated at curve 3. Carrying the
damping further, in order to gain extended LF response, will result in a
very low sensitivity, degrading the self noise floor of the microphone system. Generally, in normal use the LF rolloff will be at least partially compensated by proximity effect (see next section).
Figure 4-15 shows on-axis frequency response curves for a typical high
quality capacitor gradient microphone, indicating the degree of LF rolloff
that may be tolerated. As in the case of the ribbon microphone, on-axis diffraction effects are used to advantage in maintaining extended HF response.
When judging the merits of various figure-8 microphones, most engineers will choose a ribbon over a capacitor. This may account for the fact
that there are relatively few capacitor models available. The natural LF
resonance of the ribbon, a vital ingredient in its design, provides
extended LF response, which most engineers (and artists) appreciate.
2:39 PM
Page 63
4: The Pressure Gradient Microphone
The capacitor gradient
microphone; single-ended
asymmetrical form (A);
single-ended symmetrical
form (B); push-pull
(electrically balanced)
form (C).
Effect of damping the
capacitor diaphragm.
Typical frequency
response curve for
capacitor gradient
2:39 PM
Page 64
All directional microphones exhibit proximity effect, which causes an
increase in LF output from the microphone when it is operated close to
a sound source. The effect comes about because, at close working distances, the front and back paths to the diaphragm may differ substantially in their relative distances from the sound source. The gradient
component acting on the diaphragm diminishes at lower frequencies,
while the inverse square component remains constant with frequency for
a given working distance. When this distance is small, as shown in
Figure 4-16A, it dominates at LF and causes the output to rise. The data
at B shows the relationship between the gradient and inverse square
forces on the diaphragm, and the equalized data shown at C indicates the
net output of the gradient microphone.
The magnitude of proximity LF boost for a gradient microphone is
given by:
Boost (dB) 20 log
1 (kr)2
where k 2p/l, r is the distance from the sound source (m), and l is the
signal wavelength (m). (For example, at an operating distance of 5.4 cm
(0.054 m) at 100 Hz, the value of kr is 2p(100)/344 1.8. Evaluating
equation (4.6) gives 20.16 dB.)
The rise at LF for several operating distances for a gradient microphone is shown in Figure 4-17. During the 1930s and 1940s, radio
crooners soon learned to love the famous old RCA ribbon microphones,
with their “well-cushioned” LF response that so often enhanced singers’
and announcers’ voices.
For symmetrical front and back performance all figure-8 gradient
microphones must be side addressed, at 90 to the axis of the microphone’s
Proximity effect in the
pressure gradient
microphone: sound
pressure levels proportional
to the square of the
distance from the source
(A); combination of inverse
square and
frequency-dependent forces
on the diaphragm (B);
electrical output of
microphone system due to
the forces shown at B (C).
2:39 PM
Page 65
4: The Pressure Gradient Microphone
Typical frequency versus
operating distance response
curves for a gradient
Axes of maximum
sensitivity for a variety
of microphone forms.
case, as shown in Figure 4-18. Also in this category would be most studio
large diaphragm microphones with adjustable patterns. As a general rule,
most small diaphragm models with interchangeable omni and cardioid
capsules will be end addressed, as shown in Figure 4-18.
2:40 PM
Page 66
The great majority of directional microphones in professional use today
are members of the first-order cardioid family. The term first-order refers
to the polar response equation and its inclusion of a cosine term to the
first power, or order. These microphones derive their directional patterns
from combinations of the pressure and gradient microphones discussed
in the previous two chapters. By comparison, a second-order microphone will exhibit a pattern dependent on the square of the cosine term.
Stated differently, a first-order microphone has response proportional to
the pressure gradient, whereas a second-order design has response that is
proportional to the gradient of the gradient.
The earliest directional microphones actually combined separate
pressure and gradient elements in a single housing, combining their outputs electrically to achieve the desired pattern. Today, most directional
microphones with a dedicated pattern have a single transducing element
and make use of calculated front-back delay paths to achieve the desired
pattern. We will discuss both approaches.
The addition of a gradient element and a pressure element is shown in
Figure 5–1. In polar coordinates, the omnidirectional pressure element is
assigned a value of unity, indicating that its response is uniform in all
directions. The gradient element is assigned the value cos , indicating its
bidirectional figure-8 pattern:
0.5 (1 cos ) 0.5 0.5 cos (5.1)
which is the polar equation for a cardioid pattern (so-named for its
“heart-like” shape). The geometric construction of the pattern is derived
2:40 PM
Page 67
5: First-Order Directional Microphones
Derivation of the cardioid
pattern by combination
of the pressure
component and the
gradient (bidirectional)
simply by adding the two elements at each bearing angle around the
microphone, taking account of the negative back lobe of the gradient
Figure 5–2 shows several of the family of polar curves that can be produced by varying the proportions of omnidirectional (pressure) and cosine
(gradient) components. Here, the general form of the polar equation is:
A B cos (5.2)
where A B 1.
Figure 5–3 shows graphs of the four main cardioid patterns normally
encountered today. The graphs are shown in both linear and logarithmic
(decibel) form.
The Subcardioid: Although other A and B ratios may used, this pattern is generally represented by the polar equation:
0.7 0.3 cos (5.3)
The pattern is shown in Figure 5–3A. The directional response is 3 dB
at angles of 90 and 10 dB at 180. Microphones of this fixed pattern
are of relatively recent development, and they have found great favor
with classical recording engineers. The subcardioid pattern is sometimes
referred to as a “forward oriented omni.”
The Cardioid: This pattern is the standard cardioid, and it is represented by the polar equation:
0.5 0.5 cos (5.4)
This pattern is shown in Figure 5–3B. Directional response is 6 dB at
90 and effectively zero at 180. It is the most widely directional pattern in general use today, its usefulness in the studio deriving principally
from its rejection of sound originating at the rear of the microphone.
The Supercardioid: This pattern is represented by the polar equation:
0.37 0.63 cos (5.5)
The pattern is shown in Figure 5–3C. Directional response is 12 dB at
90 and 11.7 dB at 180. This pattern exhibits the maximum frontal
2:40 PM
Page 68
Various combinations of
pressure and gradient
components and their
resultant patterns.
pickup, relative to total pickup, of the first-order cardioids, and as such
is useful for pickup over a wide frontal angle in the studio.
The hypercardioid: This pattern is represented by the polar equation:
0.25 0.75 cos (5.6)
The pattern is shown in Figure 5–3D. Directional response is 12 dB at
90 and 6 dB at 180. This pattern exhibits the greatest random efficiency, or “reach,” in the forward direction of all members of the firstorder cardioid family. In the reverberant field, this pattern will provide the
greatest rejection, relative to on-axis pickup, of reverberant sound.
2:40 PM
Page 69
5: First-Order Directional Microphones
Polar graphs, linear and
logarithmic, for the
following patterns:
subcardioid (A); cardioid
(B); supercardioid (C); and
hypercardioid (D).
2:40 PM
Page 70
Figure 5–4 shows in chart form the principal characteristics of the firstorder cardioids. While most of the descriptions are self evident, two of
them need additional comment:
Random Efficiency (RE): RE is a measure of the on-axis directivity
of the microphone, relative to its response to sounds originating from all
directions. An RE of 0.333, for example, indicates that the microphone
will respond to reverberant acoustical power arriving from all directions
with one-third the sensitivity of the same acoustical power arriving along
the major axis of the microphone. (The term random energy efficiency
(REE) is also used.)
Distance Factor (DF): Distance factor is a measure of the “reach” of
the microphone in a reverberant environment, relative to an omnidirectional microphone. For example, a microphone with a distance factor of
2 can be placed at twice the distance from a sound source in a reverberant environment, relative to the position of an omnidirectional microphone, and exhibit the same ratio of direct-to-reverberant sound pickup
as the omnidirectional microphone. These relationships are shown in
Figure 5–5 for several first-order patterns.
There are three distinctions in directional patterns that concern the
engineer. In many applications, especially in sound reinforcement, the
reach of the microphone is important for its ability to minimize feedback.
In the recording and broadcast studios, the forward acceptance angle is
important in defining the useful pickup range and how the performers
must be arrayed. Under similar operating conditions, the rejection of offaxis sound sources may be equally important.
These considerations lead to several important distinctions among
subcardioid, cardioid, supercardioid, and hypercardioid patterns. Note
for example that the response at 90 varies progressively from sub- to
2:40 PM
Page 71
5: First-Order Directional Microphones
Characteristics of the
family of first-order
cardioid microphones.
Illustration of distance
factor (DF) for the
first-order cardioid family.
2:40 PM
Page 72
hypercardioid over the range of 3 to 12 dB. The effective 6 dB
included frontal angle varies progressively from sub- to hypercardioid
over the range of 264 to 141, while the distance factor varies from 1.3
to 2. The back rejection angle varies from 180 for the cardioid to 110
for the hypercardioid. As a function of the B term in equation (5.2), we
have the following definitions:
Randon Efficiency (RE) 1 2B 4B2/3
Front-to-Total Ratio (FTR) REF/RE
REF 0.5 0.5B B2/6
Front-to-Back Ratio (FBR) REF/REB
REB 0.5 1.5B 7B2/6
Continuous values of these quantities from B 0 to B 1 are shown
in Figure 5–6A, B and C. The figures are labeled to show the maximum
values of the functions (Bauer, 1940 and Glover, 1940). It is from this
data that the accepted definitions of the supercardioid and hypercardioid
patterns were taken.
We can draw microphone patterns only in two-dimensional form, and
it is easy to forget that the patterns actually exist in three dimensions,
as suggested in Figure 5–7. For microphones that are end addressed,
these patterns exhibit consistent rotational symmetry, since the microphone case lies along the 180 angle relative to the major axis. Side
addressed microphones may exhibit slightly asymmetrical patterns,
inasmuch as the body of the microphone lies along a 90 angle from the
major axis.
Figure 5–8 shows details of the Western Electric/Altec 639 multi-pattern
microphone. It consists of a ribbon element and a moving coil pressure
element in a single housing. The two outputs are summed, and a switch
varies the ratio between them producing a family of polar patterns.
Because of the relatively large spacing between the elements, accurate
frequency response and polar integrity are difficult to maintain beyond
about 8 kHz. It can easily be seen that precise alignment of the two
elements is critical to their proper summation.
2:40 PM
Page 73
5: First-Order Directional Microphones
Graphical illustrations of
directivity factor, distance
factor, random efficiency,
front-to-total ratio, and
front-to-back ratio for the
first order family.
Working along slightly different lines, Olson (1957) describes a cardioid ribbon microphone, a variant in the 77-DX series, that effectively
divided the ribbon into two sections, as shown in Figure 5–9. The upper
section of the ribbon provided a gradient operation while the lower section provided pressure operation. Electrical summation of the pressure
and gradient effects took place in the ribbon itself.
2:40 PM
Page 74
“directivity balloons” for
the first-order family.
Section view of the
Western Electric model
639 dual-element variable
directional microphone.
(Figure courtesy of Altec.)
2:40 PM
Page 75
5: First-Order Directional Microphones
Section view of mechanism
of a ribbon cardioid
Since the cardioid pattern is a combination of a pressure element (which
has no proximity effect) and a gradient element (which does), we would
expect the cardioid pattern to exhibit somewhat less proximity effect
than the gradient element alone. This is in fact the case, and Figure 5–10
shows the proximity effect on-axis for a cardioid microphone as a function of operating distance. For comparison with the figure-8 pattern, see
Figures 4–15 and 4–16.
It is also necessary to consider proximity effect in a cardioid microphone as a function of the operating angle, and this is shown in
Figure 5–11 for an operating distance of 0.6 m (24 in). For on-axis operation, the proximity effect will be the same as shown in Figure 5–10 for
the same operating distance. Note that at 90 there is no proximity
effect. The reason is that at 90 there is no contribution from the gradient element; only the pressure element is operant, and it exhibits no
proximity effect. As we progress around the compass to 180 we observe
that, for sound sources in the far field, there will be no response.
However, as the sound source at 180 moves closer to the microphone,
there will be a pronounced increase in proximity effect at very low frequencies. The rapid increase in response with lowering frequency at 180
2:40 PM
Page 76
Proximity effect versus
operating distance for a
cardioid microphone on
Proximity effect versus
operating angle at a fixed
distance of 0.6 m (24 in).
Typical action of a
high-pass filter to reduce
proximity effect at close
working distances.
2:40 PM
Page 77
5: First-Order Directional Microphones
is due to the subtractive relationship between gradient and pressure
quantities at that angle.
Most directional microphones, dynamic or capacitor, have a built-in
switch providing an LF cut in response to partially counteract proximity
effect when the microphone is used at short operating distances, as
shown in Figure 5–12. This response modification is used primarily for
speech applications; for many music applications the proximity LF boost
may be considered a benefit.
A single diaphragm cardioid capacitor microphone is shown in
Figure 5–13; representations of sound incident at 0, 90, and 180 are
shown at A, and the corresponding physical circuits are shown at B. The
diaphragm is resistance controlled. Note that there is a side opening in
Basic principle of a single
diaphragm cardioid
microphone; mechanical
views (A); physical circuits
(B); cutaway view of a
capacitor single diaphragm
cardioid microphone
(C). (Photo courtesy of
Audio-Technica US.)
2:40 PM
Page 78
the microphone case that provides a delay path from the outside to the
back of the diaphragm. The net pressure gradient on the diaphragm is
the combination of the variable external delay, which is dependent on the
angle of sound incidence, and the fixed internal delay. The external
acoustical delay and the internal delay have been designed to be equal for
on-axis operation.
At 0, the external acoustical delay and fixed internal delay add to produce a fairly large pressure gradient. This gradient increases with frequency,
as we noted in the preceding chapter, and the effect on a resistancecontrolled capacitor diaphragm is a constant output over the frequency
range. At 90, the external delay is zero since sound arrives at the front
and rear openings of the microphone in-phase. Only the internal delay
contributes to the gradient and will be one-half the total delay for 0. For
sound arriving along the 180 axis, the internal and external delays will
be equal but in phase opposition, producing a zero gradient and no signal
output from the microphone. A cutaway view of a single diaphragm
cardioid capacitor microphone is shown at C.
A similar approach can be taken with the dynamic microphone
(Bauer, 1941), as shown in Figure 5–14. This basic design is shown at A.
With numerous refinements and additions, it has become the basis for
today’s dynamic “vocal microphones” widely used throughout the music
and communication industries. For flat LF response the diaphragm must
be mass controlled. This may result in problems of LF handling noise
and overload, and most directional dynamic microphones are designed
with substantial LF damping, relying on proximity effect at close working
distances to equalize the effective frequency response.
Basic principle of a single
diaphragm dynamic
cardioid microphone (A);
cutaway view of a typical
dynamic single diaphragm
cardioid microphone (B).
(Photo courtesy of
Audio-Technica US.)
2:40 PM
Page 79
5: First-Order Directional Microphones
As we have seen, the single diaphragm cardioid microphone has
much in common with the gradient microphone discussed in the previous
chapter. It is in fact a gradient microphone with the addition of a fixed
internal delay path behind the diaphragm. Changing the amount of internal delay enables the designer to produce other cardioid types. For example, by shortening the delay appropriately a supercardioid pattern, with
its characteristic null at 126, and a hypercardioid pattern, with its
characteristic null at 110, can be attained. A cutaway view of a
typical single diaphragm dynamic cardioid microphone is shown at B.
By providing a means for physically changing the internal delay path,
variable pattern single diaphragm microphones can be realized. In general, a microphone that offers a selection of patterns forces a trade-off in
terms of HF response and pattern integrity as compared with a microphone that has been optimized for a single pickup pattern.
Olson (1957) describes an RCA ribbon microphone in the 77-series
that offered a number of pickup patterns. The basic design is shown in
Figure 5–15A. A cowl was placed behind the ribbon, and an aperture at
the back of the cowl could be varied by a shutter to produce the family
of first-order patterns, as shown at B.
The AKG Model C-1000 capacitor microphone is normally operated
with a cardioid pattern; however, the grille that encloses the capsule can
be removed by the user, and an adapter slipped over the capsule assembly
that effectively alters path lengths around the microphone to produce a
hypercardioid pattern. This design is shown in Figure 5–16. An external
view of the microphone is shown at A; the normal internal configuration
is shown at B, and the modified configuration at C.
When the pattern change adapter is placed over the microphone, an
additional distributed gradient path is provided. Since it is applied to the
positive side of the diaphragm, it works to some degree in opposition to
the built-in gradient path on the negative side of the diaphragm. In so
doing, it increases the dipole, or figure-8, contribution to the microphone’s effective pattern thus producing a hypercardioid pattern.
A very elegant example of a variable pattern, single diaphragm
capacitor microphone is the Schoeps MK 6 capsule. Section views of the
internal spacing of elements for the three basic patterns are shown in
Figure 5–17. The ingenious mechanism fits within a cylindrical housing
only 20 mm in diameter. The operation of the microphone is explained
as follows:
Omnidirectional response. At A, both inner and outer moving assemblies are in their far left positions, effectively closing off the back
openings to the diaphragm and resulting in omnidirectional response.
2:40 PM
Page 80
A variable pattern single
ribbon microphone design
(A); functional views,
acoustical circuits and
resultant patterns (B).
(Figure from Olson, 1957.)
Cardioid response. At B, the inner moving assembly has been positioned to the right, exposing the rear opening. The delay path
through the rear of the microphone is equal to the path around the
microphone, so sounds arriving from the rear of the microphone will
cancel at the diaphragm.
Figure-8 response. At C, both inner and outer moving assemblies
have been positioned to the right, and a new rear opening, symmetrical with the front opening, has been exposed. This front-back symmetry will produce a figure-8 pattern. The left portion of the design
is nonfunctional mechanically; essentially, it provides a matching
2:40 PM
Page 81
5: First-Order Directional Microphones
Details of the AKG model
C-1000 cardioid capacitor
microphone; external view
(A); internal view, normal
operation (B); internal view,
hypercardioid operation
(C). (Figure courtesy of
AKG Acoustics.)
acoustical boundary condition on the front side of the microphone
which matches that on the back side when the microphone is in the
figure-8 mode.
By far the most common variable pattern microphone in use today is the
Braunmühl-Weber (1935) dual diaphragm capsule design. The basic
design is shown in Figure 5–18A. The backplate is positioned between
the two diaphragms; the backplate has shallow holes in each side as well
as holes that go through the plate. The purpose of both sets of holes
is to provide the required acoustical stiffness and damping on the
diaphragms for the various modes of operation of the microphone. An
alternate design as used by AKG Acoustics is shown at B; here, there are
two perforated backplates separated by a damping element.
As with all capacitor microphones operating as gradient devices, the
diaphragms have a midrange tuning and are heavily damped to work in
a resistance-controlled mode. The perforations connecting the two
diaphragms contribute to a high degree of resistive damping on the
diaphragm motion. The air between the two diaphragms acts as a capacitance, and a simplified analogous circuit is shown at C. The challenge in
the design is to ensure that the resistive damping is more dominant than
2:40 PM
Page 82
Details of the Schoeps
MK 6 single diaphragm
variable pattern capsule.
Omnidirectional response
(A); cardioid response (B);
figure-8 response (C).
(Figure after Schoeps and
Posthorn Recordings.)
the reactive stiffness and mass of the captured air. A photo of a typical
dual element capsule is shown at D.
In the example shown in Figure 5–19A, only the left diaphragm has
been polarized; the right diaphragm is at zero voltage. First, consider a
sound source at a direction of 90 relative to the diaphragms (B). Since
both diaphragms are equidistant from the sound source there will be
equal and opposite pressure vectors, S1 and S2, as shown at B. The pressures will push the diaphragms in and out against the stiffness of the
enclosed air.
2:40 PM
Page 83
5: First-Order Directional Microphones
Section view of the basic
Braunmühl-Weber dual
diaphragm capacitor
microphone (A); an
equivalent design used
by AKG Acoustics (B);
simplified equivalent circuit,
impedance analogy
(C); view of typical dual
backplate capsule (D).
Vector diagrams showing
operation of the dual
diaphragm assembly; for
sound incident at 0 (A);
for sound incident at 90
(B); for sound incident at
180 (C).
2:40 PM
Page 84
Electrical circuit for
combining the outputs of
the dual diaphragms and
producing the first-order
family of patterns.
For a source at 0, as shown at A, there will be the same set of vectors plus and additional set of vectors (s1 and s2) caused by the pressure
gradient effect at the microphone. These pressures will push the
diaphragms and enclosed air as a unit because of the interconnection
between the two sides of the backplate. The two sets of vectors will combine as shown in the figure, and the two vectors on the back side (180
relative to the signal direction) will, through careful control of damping
and stiffness, cancel completely. Only the left diaphragm will move, producing an electrical output. For sounds arriving at 180, only the right
diaphragm will move, as shown at C. Since the back diaphragm is not
polarized there will be no electrical output.
In effect, the assembly behaves as two back-to-back cardioid capacitor microphones sharing a common backplate. If both diaphragms are
connected as shown in Figure 5–20 and independently polarized, the
entire family of first-order patterns can be produced. The data shown in
Figure 5–21 illustrates how back-to-back cardioids can be added and
subtracted to produce the family of first-order cardioid patterns.
Wiggins (1954) describes a novel variation on the standard dynamic
directional microphone. As discussed above under Proximity Effect in
Cardioid Microphones, the standard dynamic directional microphone
2:40 PM
Page 85
5: First-Order Directional Microphones
Combinations of dual
back-to-back cardioids 1
and 2 and their resultant
polar patterns.
relies on a mass-controlled diaphragm in order to maintain the extended
LF response. Wiggins designed a dynamic microphone with a resistancecontrolled diaphragm and changed the normal single back path into a set
of three back paths to cover LF, MF, and HF actions separately, hence the
term Variable-D (standing for variable distance). The intent of the design
was to produce a wide-band directional dynamic microphone that exhibited better LF response and greater resistance to handling noise and
mechanical shock than is typical of mass-controlled dynamic cardioids.
2:41 PM
Page 86
Schematic view of
Electro-Voice Variable-D
microphone (A); cutaway
view of a typical Variable-D
microphone (B). (Diagram
courtesy of Electro-Voice.)
If a resistance-controlled diaphragm were to be used with a single
back path, the response would roll off at 6 dB/octave at LF. By successively lengthening the back path for lower frequencies in selective bandpass regions, the necessary pressure gradient can be maintained, in each
frequency region, to produce flat output.
A schematic view of the microphone is shown in Figure 5–22A, and
the three back path distances (LF, MF, and HF) are clearly indicated. A
cutaway view of a typical Variable-D microphone is shown at B.
An equivalent circuit, shown in Figure 5–23, indicates the complexity of the design in terms of acoustical filtering. The microphone works
by establishing a pressure gradient that is effectively uniform over a large
frequency range (about 200 Hz to 2 kHz). Since the driving force on the
diaphragm is constant over that range, the diaphragm assembly must be
resistance controlled for flat response. At higher frequencies flat on-axis
response is maintained by diffraction effects and careful attention to resonances in the region of the diaphragm. The shunt path in the equivalent
circuit (M6, RM6 and CM6) maintains flat LF response.
Figure 5–24 shows vector diagrams for LF action (A), MF action (B),
and at an intermediate frequency between the two (C). In these diagrams,
2:41 PM
Page 87
5: First-Order Directional Microphones
Equivalent circuit for
Variable-D microphone.
(Data after Robertson,
Vector diagrams showing
operation of Variable-D
microphone. (Data after
Robertson, 1963.)
k is the wavelength constant, 2p/l; dL, dM, and dI are, respectively, the LF,
MF, and internal delay paths around and through the microphone. The
quantities k(dL dI) and k(dM dI) represent, respectively, the phase
shift in radians at LF and MF operation. Note that the value of the vector
p remains constant, indicating that the gradient is effectively independent of frequency over the range of Variable-D operation.
A persistent problem in the design of single diaphragm directional microphones is the maintenance of target pattern control at very low and very
high frequencies. Typical response is shown in Figure 5–25, where it is
clear that polar pattern integrity is compromised at both LF and HF. The
primary reason for this is the fall-off of the gradient component in the
microphone’s operation at the frequency extremes, as discussed in
Chapter 4 under The Capacitor Gradient Microphone. It is true that
more internal damping of the diaphragm would solve this problem, but
at a considerable loss of microphone sensitivity. One method for getting
around this problem is to design a two-way microphone with one section
for LF and the other for HF.
2:41 PM
Page 88
Typical example of
on- and off-axis frequency
response curves for
single-diaphragm cardioid
Details of the Sanken
CU-41 two-way
microphone. Disassembled
view (A); polar response
(B); on- and off-axis
frequency response curves
(C). (Figure courtesy of
Weingartner (1966) describes such a two-way dynamic cardioid
design. In this design approach, the LF section can be optimized in its
damping to provide the necessary LF pattern control, while the HF section
can be optimized in terms of size and polar performance. It is essential
that the combining networks used to “splice” the two sections together
be carefully designed.
Perhaps the best known capacitor two-way design is the Japanese
Sanken CU-41, which is shown in Figure 5–26. The LF to HF crossover
2:41 PM
Page 89
5: First-Order Directional Microphones
frequency is 1 kHz, where the wavelength is about 0.3 m (12 in). Since
the internal component spacing is small compared to this wavelength,
accurate combining can be attained at all normal usage angles with minimum pattern lobing. In optimizing the HF section of this microphone,
the 90 off-axis performance, the cardioid target value of 6 dB, has
been maintained out to 12.5 kHz, as can be seen in the family of off-axis
curves. Not many cardioid microphones routinely do this. The LF
parameters, independently adjusted, result in excellent pattern control to
well below 100 Hz.
Photo of Polarflex array
showing basic pattern
orientations. (Photo
courtesy Schoeps.)
Polarflex signal flow
diagram. (Data after
In its Polarflex system, Schoeps has introduced a method of varying firstorder microphone patterns over a wide range. As we saw earlier, omni
and figure-8 patterns can be combined to create the entire family of firstorder cardioid patterns. Polarflex makes use of separate omni and figure8 elements and allows the user to combine them as desired over three
variable frequency ranges.
As an example, a user can “design” a microphone that is essentially
omnidirectional at LF, cardioid at MF, and supercardioid at HF. The
transition frequencies between these regimes of operation can be selected
by the user. Such a microphone would be useful in recording a large
performing group, orchestral or choral, by allowing the engineer to operate at a greater than normal distance while retaining the desired presence
2:41 PM
Page 90
Front (A) and rear (B)
views of DSP-4P processor.
(Photos courtesy Schoeps.)
at MF and HF. Another application is in the studio, where the directional
characteristics and proximity effect of a vocal microphone can be
adjusted for individual applications.
Figure 5–27 shows the basic array. Note that there are both omni
and figure-8 elements closely positioned one atop the other. This basic
array constitutes in effect a single microphone, and for stereo pickup a
pair of such arrays will be required. The elements of the array are
processed via the circuit shown in Figure 5–28. The front and rear views
of the control module are shown in Figure 5–29.
2:42 PM
Page 91
For many applications it is necessary to use a microphone with directional
properties exceeding those of the first-order cardioid family. In the film/
video industry, for example, dialog pickup on the shooting set is usually
done by way of an overhead boom microphone which must be clearly
out of the picture and may be two or more meters away from the actors.
Adequate pickup may depend on a highly directional microphone to
ensure both speech intelligibility and a subjective sense of intimacy.
Sports events and other activities with high ambient noise levels may
require highly directional microphones for noise immunity, and recording
in highly reverberant spaces may require such microphones to enhance
musical clarity. Field recording of natural events such as bird calls and the
like may call for operation at great distances; here, high directionality
microphones may be essential in attaining a usable recording.
High directionality microphones generally fall into three categories:
1. Interference-type microphones. These designs achieve high directionality by providing progressive wave interference of high frequency
sound arriving off-axis, thus favoring sound arriving on-axis.
2. Focusing of sound by means of reflectors and acoustical lenses.
These designs are analogous to optical methods familiar to us all.
3. Second and higher-order designs. These microphones make use of
multiple gradient elements to produce high directionality.
Olson (1957) describes a microphone consisting of a number of clustered parallel tubes that differ in length by a fixed amount, as shown in
3:43 PM
Page 92
Figure 6–1A. A microphone transducer is located at the far end, where
all sound pickup is combined. For sound waves arriving at 0 incidence,
it is clear that the signal at the transducer will be, to a first approximation, the in-phase sum of all contributions. For sound waves arriving
from some arbitrary off-axis angle , the sum will exhibit a degree of
phase cancellation of the contributions due to the differences in the
individual path lengths. Olson derives an equation that gives the net
sum at the transducer in terms of the received signal wavelength (), the
overall length of the array (l), and the angle of sound incidence ():
sin (l l cos )
R (l l cos )
Olson’s multiple tube
directional microphone;
physical view (A); plot of
DI as a function of overall
tube length and frequency
(B); photo of a multiple
tube directional
microphone (C). (Data
after Olson, 1957.)
2:42 PM
Page 93
6: High Directionality Microphones
A single tube line
microphone (A); vector
diagram for on- and
off-axis operation (B).
The approximate directivity index (DI) for the array of tubes is shown in
Figure 6–1B as a function of array length and received signal wavelength.
For example, at 1 kHz the wavelength is approximately 0.3 m (12 in).
Thus, for an array length of 0.3 m the DI is about 5 dB. A photo of one
embodiment of the design is shown as C.
To a large extent, the complexity of the multi-tube array can be
replaced by the far simpler design shown in Figure 6–2A. Here, we can
imagine that the individual tubes have been superimposed on one another,
forming a single tube with multiple slot openings along its length. Such a
microphone as this is often termed a line microphone. In order to work,
the acoustical impedance at each opening must be adjusted so that sound,
once having entered the tube, does not readily exit at the next opening,
but rather propagates along the tube in both directions. Some designs
make use of external protruding baffles, or fins, which provide HF
resonant cavities that help to maintain off-axis HF response.
Because of its obvious physical appearance, the single-tube line microphone is commonly referred to as a “rifle” or “shotgun” microphone, and
it is in this form that most commercial high-directionality microphones are
designed today. As a general rule, the line section of the microphone is
2:42 PM
Page 94
added to a standard hypercardioid microphone, and it acts to increase the
DI of the hypercardioid above a frequency inversely proportional to the
length of the line.
A vector representation of the operation of the line microphone is
shown in Figure 6–2B. For on-axis operation (
0) the individual signal vectors are in phase and add to create a reference response vector, p.
For sounds arriving slightly off-axis the individual vectors will be slightly
displaced from each other in phase angle, and the vector summation will
exhibit a reduction in level along with a lagging phase angle, relative to
on-axis arrival. For sounds arriving considerably off-axis the angular displacement of individual phasors will be greater, and the vector summation
will be even less.
These observations are borne out in measurements on commercial
microphones. Figure 6–3A shows the AKG Model C 468 line microphone,
which is available with one line section or with two, at the user’s choice.
The length of the microphone with a single section is 17.6 cm (7 in), and the
length with two sections is 31.7 cm (12.5 in). Corresponding DI plots are
shown at B. It can be seen that the DI data broadly follows that shown
in Figure 6–4A.
The audio engineer in search of a high-directionality line microphone must perform the relevant calculations to make sure that a given
microphone model is appropriate to the task at hand. The shorter models on the order of 30 cm (11 in) may produce the desired directionality
at 4 kHz and above, but to maintain this degree of directionality at, say,
700 Hz, would require a line microphone some two meters in length!
Not many line microphones have been designed to meet these requirements. One notable line microphone from the 1960s, the ElectroVoice
Model 643 “Cardiline,” had a length of 2.2 m (86 in). Many commercial
line microphones have an operating length of little more than 0.2 m
(8 in). Obviously, an operating point somewhere between these extremes
will yield the best combination of directional performance and convenience of field operation and flexibility.
The DI of a line microphone at low frequencies is simply the DI of the
basic transducing mechanism, normally a hypercardioid element with its
characteristic DI of 6 dB. As frequency increases, there is a point at which
the line takes over and above which the DI tends to rise approximately
3 dB per doubling of frequency. That approximate frequency is given by
Gerlach (1989):
f0 c/2l
where c is the speed of sound and l is the length of the line section of the
microphone (c and l expressed in the same system of units). Figure 6–4A
presents a graph showing the rise in DI as a function of frequency for a
line microphone of any length l. The relation between f0 and l is given in
2:42 PM
Page 95
6: High Directionality Microphones
Views of the AKG C
468 line microphone (A);
DI plots for single and
double sections (B). (Figure
courtesy of AKG Acoustics.)
equation (6.1) and, for a given microphone, f0 may be read directly from
the data shown in Figure 6–4B. For example, assume that a line microphone has a line section that is 200 mm (8 in) long. From Figure 6–4B we
can read directly the value of 900 Hz for f0. Now, examining Figure
6–4A, we note that, if the microphone has a base hypercardioid element,
the effect of the line will only become apparent above about 3f0, or about
2700 Hz.
2:42 PM
Page 96
Directional performance
of a line microphone;
directivity index of a line
microphone as a function
of line length (A);
determining the value
of f0 as a function of line
length, l (B).
When microphone directional performance is presented directly by
way of polar plots, the DI may be estimated by noting the included angle
between the 6 dB beamwidth values and entering that included angle
into the nomograph shown in Figure 6–5. (Note: Beamwidth is defined
here as the angular width of microphone pickup over which the response
loss is no greater than 6 dB, relative to the on-axis value.)
Figure 6–6A shows a section view of a parabolic reflector microphone,
as described by Olson (1957). Polar response is shown in Figure 6–6B,
and DI is shown at C. It can be appreciated how cumbersome such
microphones are to transport and use. However, microphones of this
type are used in field nature recording activities as well as in certain
surveillance activities, because the directional beamwidth at very high
frequencies can be quite narrow.
2:42 PM
Page 97
6: High Directionality Microphones
A microphone based on the acoustic lens principle is shown in
Figure 6–7 (Olson, 1957). Here, the lens acts as a converging element,
focusing parallel rays of sound onto a transducer located at the focus of
the lens. It is clear that a microphone such as this is of little practical use,
and we show it primarily as an example of the acoustical lens principle.
Determining approximate
directivity index (DI)
when the nominal 6 dB
beamwidth of the
microphone is given; for a
given polar pattern, strike
off the included angle over
which the response is no less
than 6 dB relative to 0 dB
on-axis; then, using the
nomograph, read directly
the DI in dB.
As we have seen in previous chapters, the basic family of directional
microphones are referred to as first-order, inasmuch as the directional
response is proportional to the first power of a cosine term. A secondorder response pattern has directivity that is proportional to the square
of the cosine term. Stated differently, a first-order microphone has
response that is proportional to the pressure gradient, whereas a secondorder design has response proportional to the gradient of the gradient.
The principle of the second-order microphone is developed as shown
in Figure 6–8. At A, we show the physical circuit of a first-order cardioid
microphone as a basis for comparison. The directional response of this
microphone is given as 0.5(1 cos ), where represents the bearing angle of sound arriving at the microphone.
If two such first-order microphones are placed very close together
and their outputs subtracted from each other, we have their equivalent
physical circuit as shown at B, and the directional response will be
(0.5 cos )(1 cos ). The directional pattern is shown at C in both
linear and log (decibel) polar plots.
As we have seen in earlier chapters, the effective gradient distance
between front and back openings in a first-order microphone (D) is quite
small, perhaps no more than a centimeter or so. The requirement for an
additional gradient distance, D, calls for extreme measures if this distance is to be minimized. Effective second-order microphones can be
made to work out to perhaps 6 or 8 kHz, with diminished second-order
effect above that point.
The general directional equation for a second-order microphone is:
(A B cos )(A B cos )
where A B 1 and A B 1.
At present, second-order microphones are rarely used in the recording studio, and their major application is in the area of close-speaking,
noise canceling operations in difficult communications environments.
Proximity effect is a problem in that it rises at a rate of 12 dB/octave at
low frequencies, rendering these microphones very sensitive to wind
effects and close placement to sound sources. Woszczyk (1984) discusses
some studio applications in detail.
As examples of second-order response, we show pertinent data for
two designs: (0.5 0.5 cos )(0.5 0.5 cos ) (see Figure 6–9) and
cos2 (see Figure 6–10).
2:42 PM
Page 98
Details of a parabolic
microphone; section
view (A); polar response (B);
directivity versus
frequency (C).
Section view of an
acoustic lens directional
The design shown in Figure 6–9 resembles a cardioid pattern with a
slightly narrowed contour; its response at 90 is 12 dB, as compared
to 6 dB for a first-order cardioid. The physical circuit is shown at A,
and the actual realization is shown at B. Polar plots, both linear and log,
are shown at C.
2:42 PM
Page 99
6: High Directionality Microphones
Principle of the
second-order microphone;
physical circuit of a
first-order cardioid
microphone (A); physical
circuit for a second-order
cardioid (B); polar patterns
for the microphone shown
in B (C).
The design shown in Figure 6–10 resembles a figure-8 pattern with
a slightly narrowed contour, and its response at 45 is 6 dB, as compared to 3 dB for the first-order figure-8 pattern. The physical circuit
is shown at A, and the actual realization is shown at B. Polar plots, both
linear and log, are shown at C.
The design of higher-order microphones can be simplified to some degree
by sharing of elements, as shown in Figure 6–11. Here, the two sections
making up the second-order design consist of elements one and two, and
elements two and three. Element two is shared between both sections. In
general, this approach provides a closer overall spacing between elements of the microphone. Here, note that distance D is the same for both
gradient values.
Another design option is to limit the HF range of second-order
action, and letting normal HF pattern narrowing to take over at higher
frequencies, as shown in Figure 6–12. Here, a pair of cardioid elements
are connected for second-order operation up to about 7 kHz. Above that
2:42 PM
Page 100
Design data for the
second-order microphone,
(0.5 0.5
cos )(0.5 0.5cos ):
physical circuit (A);
mechanical circuit (B); polar
response (C).
frequency the contribution of the rear element is limited, and the front
element’s increasing directivity is allowed to take over response in the
upper octave. The baffle surrounding the front element can be sized and
shaped to optimize the transition from second-order response to beaming first-order response. Above about 12 kHz, the increased beaming of
the front element can be ignored.
The microphone shown in Figure 6–13 is intended to be used with a
handheld video camera with a zoom lens. The microphone “zooms”
electronically in synchronism with the lens system. The microphone
array consists of three first-order cardioid elements whose outputs are
combined in several ways to produce three distinct patterns, including
the intervening patterns. The three primary patterns are:
1. Wide angle: omnidirectional pickup (sum of elements 2 and 3). It is
produced when potentiometers R1 and R2 are in position 1.
2:42 PM
Page 101
6: High Directionality Microphones
Design data for the
second-order microphone,
cos2 : physical circuit
(A); mechanical circuit
(B); polar response (C).
Second-order microphone
with shared elements.
2. Medium angle: cardioid pickup (element 1 alone). It is produced
when potentiometers R1 and R2 are in position 2.
3. Narrow angle: second-order pickup (sum of elements 2 and 3 with
necessary equalization). It is produced when potentiometers R1and
R2 are in position 3.
There is an overall gain shift of 12 dB throughout the microphone’s
operating range to compensate for distance effects.
2:42 PM
Page 102
A combination of first- and
second-order microphone
Details of a “zoom”
microphone. (Data after
Ishigaki et al., 1980.)
The microphone shown in Figure 6–14A has the polar equation
. The resulting directional response is shown at B. Third-order
microphones have been used primarily in noise canceling applications,
where their immunity to distant sound sources is considerable (Beavers
and Brown, 1970).
2:43 PM
Page 103
6: High Directionality Microphones
Details of a third-order
microphone; physical circuit
(A), polar response, both
linear and log, (B).
The RE (random efficiency) of a microphone whose polar pattern is
symmetrical about its principal axis is given by:
RE 1 sin [f(
)]2 d
where is the response angle in radians and f(
) is the response value ()
at angle . If f(
) can be described by cosine relationships (as most standard patterns can), equation (6.4) leads to a definite integral that can be
easily solved.
2:43 PM
Page 104
TABLE 6–1 Microphone directivity data in tabular form (orders 0 through 4)
Pattern Equation
Efficiency (RE)
Factor (DF)
Index (DI)
1 (omni)
0 dB
cos (figure-8)
cos (2nd order figure-8)
cos3 (3rd order figure-8)
cos (4th order figure-8)
0.5 0.5 cos (cardioid)
0.25 0.75 cos (hypercardioid)
(0.5 0.5 cos )cos (2nd order cardioid)
(0.5 0.5 cos )cos2 (3rd order cardioid)
(0.5 0.5 cos )cos3 (4th order cardioid)
Data from Olson, 1972
Data for several orders of microphone patterns are given in the
Table 6–1.
The relationships among RE, DF, and DI are:
DI 10 log DF dB
2:43 PM
Page 105
In this chapter we discuss the performance parameters of microphones
that form the basis of specification documents and other microphone
literature. While some microphone standards are applied globally, others
are not, and this often makes it difficult to compare similar models from
different manufacturers. Some of the differences are regional and reflect
early design practice and usage. Specifically European manufacturers
developed specifications based on modern recording and broadcast practice using capacitor microphones, whereas traditional American practice
was based largely on standards developed in the early days of ribbon
and dynamic microphones designed originally for the US broadcasting
industry. Those readers who have a special interest in making microphone measurements are referred to the standards documents listed in
the bibliography.
1. Directional properties: Data may be given in polar form or as a set
of on- and off-axis normalized frequency response measurements.
2. Frequency response measurements: Normally presented along the
principal (0) axis as well as along 90 and other reference axes.
3. Output sensitivity: Often stated at 1 kHz and measured in the free
field. Close-talking and boundary layer microphones need additional qualification. Some manufacturers specify a load on the
microphone’s output.
4. Output source impedance.
2:43 PM
Page 106
5. Equivalent self-noise level.
6. Maximum operating sound pressure level for a stated percentage of
total harmonic distortion (THD).
Additionally, a complete listing of mechanical and physical characteristics
and any switchable performance features built into the microphone are
described in this chapter.
Frequency response data should always state the physical measuring
distance so that an assessment of proximity effect in directional microphones can be correctly made. If no reference is made, it can be assumed
that measurements are made at 1 meter. Data for professional microphones may be presented with tolerance limits, as shown in Figure 7–1.
Here, the data indicate that the microphone’s response falls within a
range of 2 dB above about 200 Hz (slightly greater below that frequency); however, there is no indication of the actual response of a sample
If the data can be presented with clarity, some manufacturers will
show proximity effects at distances other than the reference one meter,
as shown in Figure 7–2. This data is especially useful for vocal microphones that are intended for close-in applications.
Many manufacturers show response at two or more bearing angles
so that the variation in response for those off-axis angles can be clearly
seen, as shown in Figure 7–3. Here, the response for a cardioid is shown
on-axis and at the nominal null response angle of 180. For supercardioid and hypercardioid microphones, the response at the null angles of
110 and 135 may also be shown.
Taking advantage of normal microphone symmetry, polar plots may
be restricted to hemispherical representation, as shown in Figure 7–4.
For microphones that are end-addressed, it is clear that response will be
Amplitude response versus frequency with upper and lower limits for a capacitor vocal microphone; effect of LF cut is
also shown. (Figure courtesy of Neumann/USA.)
2:43 PM
Page 107
7: Microphone Measurements, Standards and Specifications
Proximity effect for a
dynamic vocal microphone
shown at several working
distances. (Figure courtesy of
Shure Inc.)
Amplitude response shown
at reference angles of 0
and 180 for a Variable-D
dynamic microphone.
(Figure courtesy of
Microphone polar groups,
hemispherical only, for
omni capacitor (A) and
cardioid capacitor
microphones (B).
(Figure courtesy of AKG
2:43 PM
Page 108
symmetrical about the axis throughout the frequency band. However, for
side-addressed microphones, there will be variations, especially at higher
frequencies where the asymmetry in diaphragm boundary conditions
becomes significant. Normally, this data is not shown, but there are
interests within the microphone user community to standardize the presentation of more complete directional information. As present, such
additional data is presented at the discretion of each manufacturer.
The principal method of presenting microphone sensitivity is to state the
output rms voltage (mV/Pa) when the microphone is placed in a 1 kHz
free progressive sound field at a pressure of 1 Pa rms (94 dB LP). A nominal microphone load impedance of 1000 ohms may be stated as well,
but the standard is normally referred to as the “open circuit” output
voltage of the microphone. Another way of stating this data is to give the
rms voltage output level in dB relative to one volt:
Output level (dBV) 20 log (rating in mVrms) 60 dB
Microphone power output specifications were developed during the
early days of broadcast transmission when the matched impedance concept was common. Here, the microphone is loaded with an impedance
equal to its own internal impedance, as shown in Figure 7–5A. When
unloaded, as shown at B, the output voltage is doubled.
The rating method is somewhat complicated, and we now give
an example: consider a dynamic microphone with rated impedance of
50 ohms and an open-circuit output sensitivity of 2.5 mV/Pa. In modern
specification sheets this voltage level may also be expressed as 52 dB
dBV (re 1 volt). The same microphone, if its loaded output power is
Microphone output
unloaded (A); loaded (B).
2:43 PM
Page 109
7: Microphone Measurements, Standards and Specifications
Microphone power output
given, would carry an output power rating of 45 dBm. This is solved as
When the microphone is given a matching load of 50 ohms, its output
voltage will be reduced by one-half, or 1.25 mV. The power in the load
will then be:
Power (1.25)2/50 3.125 108 W, or 3.125 105 mW
Solving for power level in dBm:
Level 10 log (3.125 105) 45 dBm
The nomograph in Figure 7–6 lets us solve this directly, as indicated by
the line that has been drawn over the nomograph. Here, we simply take
2:43 PM
Page 110
the unloaded output voltage level (re 1 mV) of 8 dB (60 52) and
locate that value at A. The nominal impedance of the microphone
(50 ohms) is located at B. A line is then drawn between the two points
and the microphone’s sensitivity, in dBm per pascal, is read directly at B.
Other rarely used variations of this method are:
1. Output in dBm per dyne per square centimeter (dBm measured in a
matched load impedance at 74 dB LP).
2. Output in dBm, EIA rating (dBm measured in a matched load
impedance at an acoustical level of 0 dB LP).
The reader can readily appreciate the simplicity and universality of
the modern open circuit output voltage rating method.
Virtually all of today’s professional microphones, capacitor or dynamic,
are what may be called “low impedance,” as opposed to the high impedance models of decades past. The range of impedance may be typically
from 50 to 200 ohms for capacitor microphones, or up to the range of
600 ohms for some dynamic models.
Since the traditional input load impedance seen by a microphone
today is in the range of 2000 to 5000 ohms, it is clear that the load
impedance is high enough that it has little measurable effect on the
microphone’s output voltage.
This being the case, the microphone’s output impedance rating is of
little practical consequence in modern systems layout. However, some
microphone preamplifiers have a control for adjusting the input circuitry
for specifically matching a wide range of microphone output impedances. (See Chapter 8 under the Stand-Alone Microphone Preamp.)
In designing capacitor microphones the engineer is free to set the reference output sensitivity to match the intended use of the microphone.
Table 7–1 gives normal sensitivity ranges.
The design criterion is simple; microphones intended for strong
sound sources will need less output sensitivity to drive a downstream
TABLE 7–1 Normal sensitivity ranges by use
Microphone usage
Normal sensitivity range
Close-in, hand-held
Normal studio use
Distant pickup
2–8 mV/Pa
7–20 mV/Pa
10–50 mV/Pa
2:43 PM
Page 111
7: Microphone Measurements, Standards and Specifications
preamplifier to normal output levels, while distant pickup via boundary
layer microphones or rifle microphones will need greater output sensitivity for the same purposes.
Today, the self-noise level of a capacitor microphone is expressed as an
equivalent acoustical noise level stated in dB(A). For example, a given
microphone may have a self noise rating of 13 dB(A). What this means is
that the microphone has a noise floor equivalent to the signal that would
be picked up by an ideal (noiseless) microphone if that microphone were
placed in an acoustical sound field of 13 dB(A). Modern studio grade
capacitor microphones generally have self-noise ratings in the range from
7 dB(A) to 14 or 15 dB(A). Tube models will have higher noise ratings,
many in the range from 17 dB(A) to 23 dB(A).
As a practical matter, the self-noise of a modern capacitor microphone
will be about 10 to 12 dB greater than the equivalent input noise (EIN)
of a good console or preamplifier input stage; thus, the self-noise of the
microphone will be dominant. With a dynamic microphone this is not
normally the case; the output voltage of the dynamic microphone may be
10 to 12 dB lower than that of a capacitor model so that the EIN of the
console will dominate. As a result of this, dynamic microphones do not
carry a self-noise rating; rather, their performance must be assessed relative
to the EIN of the following console preamplifier.
Some microphone specifications carry two self-noise ratings. One of
these is the traditional A-weighted curve and the other is a psychometric
curve that is more appropriate for microphones used in non-recording
measurement purposes. The two curves are shown in Figure 7–7.
Two self-noise weighting
curves for microphone
2:43 PM
Page 112
For studio quality microphones the reference distortion limit is established
as the acoustical signal level at 1 kHz which will produce no more than
0.5% THD (total harmonic distortion) at the microphone’s output.
Reference distortion amounts of 1% or 3% may be used in qualifying
dynamic microphones for general vocal and hand-held applications.
Microphone distortion measurements are very difficult to make inasmuch as acoustical levels in the range of 130 to 140 are required. These
levels are hard to generate without significant loudspeaker distortion. A
pistonphone (mechanical actuator) arrangement can be used with pressure
microphones where a good acoustical seal can be made, but it is useless
with any kind of gradient microphone.
It has been suggested (Peus 1997) that microphone distortion measurements can be made using a twin-tone method in which two swept frequencies, separated by a fixed frequency interval, such as 1000 Hz, be
applied to the microphone under test. Since the individual sweep tones
are separately generated, they can be maintained at fairly low distortion;
any difference tone generated by the diaphragm-preamplifier assembly
represents distortion and can be easily measured with a fixed 1000 Hz
filter, as shown in Figure 7–8. One problem with this method is that it is
difficult to establish a direct equivalence with standard THD techniques.
In many studio quality microphones, the distortion present at very
high levels results not from the nonlinearities of diaphragm motion but
rather from electrical overload of the amplifier stage immediately following the diaphragm. Accordingly, some manufacturers simulate microphone distortion by injecting an equivalent electrical signal, equal to
what the diaphragm motion would produce in a high sound field, and
then measure the resulting electrical distortion at the microphone’s output. This method assumes that the diaphragm assembly is not itself producing distortion, but rather that any measured distortion is purely the
result of electrical overload. We must rely on the manufacturers themselves to ensure that this indeed the case.
Twin-tone method of
measuring microphone
2:43 PM
Page 113
7: Microphone Measurements, Standards and Specifications
The useful dynamic range of a microphone is the interval in decibels
between the 0.5% THD level and the A-weighted noise floor of the microphone. Many studio quality capacitor microphones have total dynamic
ranges as high as 125 or 130, better by a significant amount than that of
a 20-bit digital recording system.
As a matter of quick reference, many microphone specification
sheets present nominal dynamic range ratings based on the difference
between the A-weighted noise floor and a studio reference level of 94 dB
LP. Thus, a microphone with a noise floor of 10 dB(A) would carry a
dynamic range rating of 84 dB by this rating method. No new information is contained in this rating, and its usefulness derives only from the
fact that 94 dB LP represents a normal maximum operating level in a
broadcasting studio. Many manufacturers ignore this rating altogether.
While all microphones may be susceptible to strong “hum” fields produced by stray ac magnetic fields at 50 or 60 Hz and its harmonics,
dynamic microphones are especially susceptible because of their construction in which a coil of wire is placed in a magnetic flux reinforcing
iron yoke structure. Examining current microphone literature shows that
there is no universally applied reference flux field for measuring microphone hum pickup. Several magnetic flux field reference standards may
be found in current literature, including: 1 oersted, 10 oersteds, and
1 milligauss. The choice of units themselves indicates a degree of confusion
between magnetic induction and magnetic intensity. The specification of
hum pickup is rarely given today, perhaps due to the use of hum-bucking
coils and better shielding in the design of dynamic microphones, as well
as the general move away from tube electronics and associated power
supplies with their high stray magnetic fields.
Pressure microphones are primarily calibrated in the laboratory using the
reciprocity principle. Here, use is made of the bilateral capability of the
capacitor element to act as either a sending transducer or a receiving
transducer, with equal efficiency in both directions. The general process
is shown in Figure 7–9.
At A, an unknown microphone and a bilateral microphone are
mounted in an assembly and jointly excited by a third transducer of
unknown characteristics. From this measurement set we can obtain only
the ratio of the sensitivities of the bilateral and unknown microphones.
2:44 PM
Page 114
The next step, shown at B, is to measure the output of the unknown
microphone by driving it with the bilateral microphone, acting as a small
loudspeaker. In this step we can determine the product of the two sensitivities, taking into account the electrical and acoustical equivalent circuits. The sensitivity of either microphone can then be determined by
algebraic manipulation of the ratio and product of the sensitivities, taken
on a frequency by frequency basis. As a secondary standard for microphone level calibration a pistonphone is normally used. The pistonphone
is a mechanical actuator that can be tightly coupled to the capsule assembly of a pressure microphone and produces a tone of fixed frequency and
pressure amplitude. Those interested in further details of microphone
calibration are referred to Wong and Embleton (1995).
The impulse response of microphones is rarely shown in the literature
because of the difficulties in achieving a consistent impulse source and
interpreting the results. Usually, a spark gap discharge is used, but it has
been shown that at high frequencies the spectrum is not consistent.
Actually, the best results over the normal audio passband may be
obtained using special loudspeaker mechanisms. Figure 7–10 shows the
The reciprocity process;
determining the ratio
(A) and product (B) of
microphone sensitivities.
Impulse responses (spark
gap) of capacitor and
dynamic microphones.
(Figure after Boré, 1989.)
2:44 PM
Page 115
7: Microphone Measurements, Standards and Specifications
spark gap response of both a capacitor and a dynamic microphone, and
it can clearly be seen that the capacitor is better behaved in its time
domain response.
Given future standardization of an adequate generating source, impulse
response may become an important microphone specification, inasmuch as
it presents more information regarding HF behavior of the microphone than
is given by the traditional frequency response amplitude curve.
On-axis frequency
response of Sennheiser
model MKH800
microphone for five
directional settings. (Figure
courtesy of Sennheiser
2:44 PM
Page 116
As higher sampling rates have been introduced into modern digital
recording, the need for microphones with extended on-axis frequency
response capability has increased. Figure 7–11 presents on-axis response
curves for the Sennheiser multipattern model MKH800 microphone
showing response out to 50 kHz – an example of what can be accommodated through careful design and amplitude equalization.
The primary source of microphone standards for normal audio engineering
applications is the International Electrotechnical Commission (IEC). IEC
document 268-4 (1972) specifically lists the characteristics of microphones to be included in specification literature, along with methods for
performing the measurements. IEC document 327 (1971) specifically
covers reciprocity calibration of 1-inch pressure microphones.
In addition, many countries have their own standards generating
groups which in many cases adopt IEC standards, often issuing them
under their own publication numbers. See Gayford (1994, Chapter 10) for
additional discussion of microphone standards.
As the new century gets under way we are seeing microphone standards working groups focus attention on matters of time-domain performance along with greater resolution and detail in presenting off-axis
coverage. The attempt here is to define, far better than before, those subjective aspects that characterize microphone performance in real-world
environments and applications. We eagerly await the outcome of these
2:45 PM
Page 117
In this chapter we examine details of electronic performance of
microphones and their interface with the following preamplifiers and
consoles. Subjects to be covered include: remote powering, microphone
output/preamp input circuitry, the stand-alone microphone preamp,
microphone cable characteristics and interference, and overall system
considerations. A final section deals with capacitor microphones operating
on the radio frequency (RF) transmission principle.
Most modern capacitor microphones operate on 48 V phantom powering
(also known as “simplex” powering) provided by the preamplifier or console input module. The basic circuitry for phantom powering is shown in
Figure 8–1. Here, a master 48 V dc supply provides positive voltage to
pins 2 and 3 through a pair of 6800 ohm resistors, while a ground return
path is provided through pin 1. The signal is carried on pins 2 and 3 and
is unaffected by the presence of identical dc voltages on pins 2 and 3.
Today, pin 2 is universally designated as the “hot” lead; that is, a positivegoing acoustical signal at the microphone will produce a positive-going
voltage at pin 2. Pin 1 provides both the dc ground return path for
phantom powering as well as the shield for the signal pair.
The circuit normally used with a transformerless microphone for
receiving power from the source is shown in Figure 8–2A. Here, a split
resistor combination is tapped for positive dc voltage which, along with
the ground return at pin 1, is delivered to the microphone’s electronics
and to the capsule for polarizing purposes.
2:45 PM
Page 118
Basic circuit for phantom
(simplex) powering.
If the microphone has an integral output transformer, then a centertapped secondary winding is used for receiving the positive voltage. This
method is shown in Figure 8–2B. Phantom powering can also be used
when there is neither a microphone output transformer or a console
input transformer, as shown in Figure 8–2C.
Phantom powering at 48 V is normally designated as P48. There are
also standards (IEC publication 268-15) for operation at nominal values
of 24 V (P24) and 12 V (P12). Voltage tolerances and current limits for
the three standards are shown in Table 8–1. The resistor values are
generally held to tolerances of 1%.
T-powering (also known as A-B powering) is rarely encountered
today. Circuit details are shown in Figure 8–3. It is normally designated
as T12. T-powering is a holdover from earlier years and still may be
encountered in motion picture work, where it is built into the many
Nagra tape recorders used in that field. Here, the audio signal leads are
at different dc voltages, and any residual hum or noise in the dc supply
will be reflected through the microphone’s output as noise.
2:45 PM
Page 119
8: Electrical Considerations and Electronic Interface
Microphone power input
circuitry using a resistive
voltage divider (A); using a
center-tapped transformer
secondary (B); using no
transformers in the
powering path (C).
Voltage tolerances and current limits
Supply voltage
Supply current
Feed resistors
12 1 V
max. 15 mA
680 24 4 V
max. 10 mA
1200 48 4 V
max 10 mA
Most microphone preamplifiers and console input sections have provision
for individual switching of phantom power on or off. When using dynamic
microphones it is good engineering practice to turn off the phantom powering, even though no current will flow through the microphone’s voice
coil should the phantom power be left on. However, if T12 power is
2:45 PM
Page 120
Basic circuit for
T-powering at 12 V.
inadvertently applied to a dynamic microphone, the 12 V dc will appear
across the microphone’s voice coil with noticeable deterioration of
response and possible damage.
Another important rule is not to turn phantom power on or off
when a microphone is bussed on and assigned to the monitor channels.
The ensuing loud “pop” could easily burn out a HF transducer in the
monitor loudspeaker systems. At the end of a session, it is normal practice to reduce both the master fader and the monitor level control to zero
before shutting down all phantom power from the console.
While on the subject of phantom power, never attempt, when phantom power is on, to remove or replace the screw-on capsule that many
capacitor microphones have. This has been known to burn out the FET
in the input circuitry of the impedance converter.
Capacitor microphones vary in their susceptibility to shifts in nominal
voltage in phantom powering. Generally, the variation is on the low side,
such as may be encountered in very long microphone cable runs. Symptoms
may be reduced signal output, increase in noise, as well as distortion.
When these conditions occur with normal cable runs, the round-trip cable
resistance should be measured, and the power supply itself checked for
possible problems.
Some capacitor microphones are designed to operate over multiple
ranges of phantom powering, for example, from 20 to 52 volts dc to
2:45 PM
Page 121
8: Electrical Considerations and Electronic Interface
Details of dc-to-dc
conversion for microphone
operation at P24 and P48
standards. (Data after
cover requirements of both P24 and P48 powering. What is required
here is a circuit that converts the applied voltage to the value required for
proper biasing of the capsule and operation of the impedance converting
preamp. The circuit for the Neumann TLM107 microphone is shown in
Figure 8–4. A major design challenge in such a circuit is the suppression
of noise that could result from the high switching rate of the input voltage
during the dc-to-dc conversion process.
This circuit provides capsule biasing voltages for its selectable patterns,
reducing them accordingly when the 10 dB pad is engaged. The 10 V dc
output is for powering the microphone’s electronics.
The present standard for phantom powering ensures that 48 V dc in a short
circuit loading condition through two parallel 6800 ohm resistors will produce a current of 14 mA dc, thus limiting the current availability for a given
microphone model. Some manufacturers have designed microphones that
can accommodate greater current for handling higher sound pressure levels
in the studio, and such microphones require lower resistance values in the
phantom supply in order to receive the higher current. Generally, this has
been carried out in proprietary stand-alone power supplies that a manufacturer may provide for a specific new microphone model.
Two-way compatibility is maintained. A new high-current microphone will work on a standard phantom supply, but it will not be able
2:45 PM
Page 122
to attain its highest degree of performance. A standard microphone will
work on the new supply, drawing only the current it needs.
Typical here is the “Super Phantom” powering that Josephson
Engineering has specified for certain microphone models, in which current
fed through a pair of 2200 ohm resistors in each leg of the power supply
is directed to the microphone. A short circuit loading condition here would
result in a current draw slightly in excess of 43 mA dc. International standardization activities are presently under way in this area.
The Audio-Technica model
AT3090 microphone;
photo of microphone (A);
circuit details (B). (Data
courtesy of Audio-Technica
2:45 PM
Page 123
8: Electrical Considerations and Electronic Interface
In an unusual design approach, Audio-Technica has introduced the
model AT3060 tube-type microphone which is powered at P48. A photo
of the microphone is shown in Figure 8–5A, and basic circuit details are
shown at B. Using a vacuum tube that will operate at a fairly low plate
potential of 37 Vdc, the tube filament requirement of 1.25 Vdc at a current draw of 10 mA is attained via an unusual integrated circuit (IC)
arrangement shown in the bottom-right portion of Figure 8–5B. The
voltage at the input to the first IC is 10 V at a current of 1.25 mA. The
three ICs in tandem progressively halve the voltage, while doubling
the current, attaining a value of 1.25 V at a current draw of 10 mA at the
output of the final IC. Various diodes are used for maintaining separation
between signal and dc paths.
The microphone element itself is an electret. The AT3060 has a
nominal sensitivity of 25.1 mV/Pa and can handle levels in the studio of
134 dB. The self-noise floor of the microphone is 17 dB(A).
Royer Labs has introduced integral electronics for phantom powering in two of their ribbon microphone models. Ribbons are particularly
susceptible to variations in downstream loading, and, with their relatively
low sensitivity, they may be subject to electrical interference in long runs.
Figure 8–6A shows a photo of the Royer Labs R-122, with performance
curves shown at B and C. The circuit diagram is shown in Figure 8–7.
The most unusual aspect of the circuit is the compound transformer,
which has four parallel sets of windings resulting in an optimum impedance match to the ribbon. The secondaries of these four sections are connected in series to attain a voltage gain of 16 dB, looking into the very
high input impedance of the buffer stage. The system has a maximum
level capability of 135 dB LP and a self noise no greater than 20 dB(A).
The sensitivity is 11 mV/Pa.
Classic tube-type capacitor microphones are as popular today in studio
recording as they have ever been. In a typical power supply (one for each
microphone), dc voltages are produced for heating the vacuum tube’s
filament, biasing the capsule, and providing plate voltage for the vacuum
tube amplifier. In many dual diaphragm designs, remote pattern switching is also included in the supply. Such a design is shown in Figure 8–8.
Note that the cable connecting the power supply to the microphone
contains seven conductors.
Most capacitor microphones have an integral output pad, as described in
detail in Chapter 3, under Details of preamplifier and polarizing circuitry.
2:45 PM
Page 124
The Royer Labs Model
R-122 powered ribbon
microphone: photo
(A); polar response
(B); frequency response at
1 meter(C). (Data courtesy
of Royer Labs.)
Circuit diagram for the
Royer Labs Model
R-122 microphone. (Figure
courtesy of Royer Labs.)
2:45 PM
Page 125
8: Electrical Considerations and Electronic Interface
Power supply schematic for a tube capacitor microphone with variable pattern switching at the power supply; pin 1A
provides the pattern control voltage. (Figure courtesy of Neumann/USA).
The effect of the pad is basically to shift the microphone’s entire operating
range from noise floor to overload point downward by some fixed
amount, normally 10 to 12 dB. The effect of this is shown in Figure 8–9.
Note that the total dynamic range of the microphone remains fixed, with
or without the pad.
Typical effect of 10 dB
microphone output pad on
maximum level capability
and self-noise of a
capacitor microphone.
A number of capacitor microphone models have an integral output
transformer built into the microphone body. Most early tube-type capacitor microphones had output transformers that were usually housed in
the power supply unit. The transformer often had split secondary windings and could be strapped for either parallel or series operation, as
shown in Figure 8–10. Typically, when the windings are strapped for
parallel operation (A), the output impedance is 50 ohms; for series strapping (B) the output impedance is 200 ohms. In either case the output
power capability remains the same:
Output power E2/R V 2/50 (2V)2/200.
In many cases it is necessary to reduce the output signal from the
microphone even lower than the parallel transformer strapping. H-pads
(shown in Figure 8–10 C) may be used both to maintain the impedance
relationship and produce the desired amount of attenuation. Today, most
transistorized capacitor microphones are transformerless and operate at
a fixed balanced output impedance.
2:45 PM
Page 126
Transformer strapping for
200 ohms (A) and 50 ohms
output impedance (B);
H-pad values for balanced
attenuation for microphone
impedances of 50 and 200
ohms (C).
In the very early days of broadcasting and recording, the dynamic microphones of the period had relatively low outputs and normally operated
into matching input impedances. Typically, a microphone with a 600ohm source impedance looked into a 600-ohm load, following the
matched impedance concept. Impedance matching was a holdover from
early telephone practice and found its way into broadcasting, and from
there into recording. Today, the bridging concept is well established. In a
bridging system, all output impedances are relatively low and all input
impedances are relatively high throughout the audio chain. Here, the
ratio of high to low impedance is normally in the range of 10-to-1 or
Figure 8–11 shows a simplified transformerless console input section
showing switchable line-microphone operation. At one time transformers
were felt to be indispensable in the design of microphone input circuitry.
Their chief advantages are a high degree of electrical balance and consequent high common-mode signal rejection. (A common-mode signal is
one that is identical at both inputs; typically, induced noise signals are
common mode.) In the era of vacuum tubes the input transformer was of
course essential. Today’s best solid state balanced input circuitry does not
mandate the use of transformers, and there are considerable economic
advantages to pass on to the user. Only under conditions of high electrical
interference might their use be required.
Today, most microphones have an output impedance in the range of
50 to 200 ohms. Most consoles have an input bridging impedance in the
range of 1500 to 3000 ohms.
A typical 250-ohm ribbon microphone may have an impedance modulus
as shown in Figure 8–12A. As long as the microphone looks into a high
2:45 PM
Page 127
8: Electrical Considerations and Electronic Interface
Simplified circuit for a
transformerless console
microphone/line input
Effect of loading on ribbon
microphone response.
impedance load, its response will be fairly flat. When used with a modern
console having 1500- or 3000-ohm input impedances, the frequency
response will be altered as shown at B. This is a response problem routinely
encountered today and frequently goes unchallenged.
2:45 PM
Page 128
Response of transformer
input circuit to varying
values of microphone
source impedance.
Another problem with improper loading is shown in Figure 8–13.
Here, a capacitor microphone is loaded by a console input transformer
that has a non-optimal termination on its secondary side. This can produce response variations similar to those shown in the figure as a function of the microphone’s output impedance. Note that as the input
impedance is reduced, the response develops a rise in the 10 kHz range.
This comes as a result of an undamped resonance involving the stray
capacitance between the primary and secondary windings of the transformer (Perkins, 1994).
Only in the lowest cost paging systems is one likely to come across an
unbalanced microphone input. For fairly short cable runs from microphone to amplifier, there may be no problems. For longer runs, where
there is greater likelihood for interference, the difference between balanced
and unbalanced operation is as shown in Figure 8–14. For balanced operation, shown at A, induced noise will be equal and in-phase in both signal
leads; it will be effectively canceled by the high common mode rejection
of the input circuitry. For unbalanced operation, shown at B, the induced
signal currents will be different between shield and conductor, and the
noise will be significant.
2:45 PM
Page 129
8: Electrical Considerations and Electronic Interface
For many field operations, stage microphones must be fed to both recording and sound reinforcement activities. Microphone splitters are used to
provide both an electrically direct feed to one operation, normally the
recording activity, and a one-to-one transformer feed to the other activity.
Circuit details for a typical passive splitter are shown in Figure 8–15. Here,
there are two secondary windings, providing an additional output for
broadcast activities. Ground lifts are often used to avoid hum due to
Balanced (A) versus
unbalanced (B) microphone
Details of a passive
microphone splitter.
2:45 PM
Page 130
ground loops. When using passive splitters, it is essential that all loads
fed by the splitter be bridging. Active splitters are often used in order to
avoid expensive transformers and to permit feeding the signal to low
input impedance loads.
Many leading recording engineers prefer to bypass console microphone
inputs altogether and use multiple individual stand-alone microphone
preamps instead. These microphone preamps are normally highly refined
versions of what may be found in the typical console, some offering significant improvements in areas of noise floor, common mode rejection,
increased output level capability, source impedance matching, calibrated
step-type gain trim, and rugged construction. Other features in some
models include equalization options and metering. These performance
attributes do not come cheaply, and a set of 16 individual preamps may
cost many times more than an excellent mass-produced 24-input console.
For many recording activities it may be difficult to justify the expense
of stand-alone preamps, considering that their outputs will be fed into a
host console for normal signal routing and bus assignment. However,
there are specific applications where their use is appropriate:
Direct-to-stereo via two microphones, where no console routing
functions are needed.
Pop multichannel recording, with each microphone assigned to one
recording channel. Here, the local host console is relegated to control
Front (A) and back
(B) views of a stand-alone
microphone preamplifier.
(Figure courtesy of FM
2:45 PM
Page 131
8: Electrical Considerations and Electronic Interface
room and studio monitoring only, with all studio microphones and
direct feeds going to their own preamps and from there to their individual recorder inputs. In this regard, the set of external stand-alone
preamps has taken the place of the “channel path” in a modern in-line
Many producers and engineers prefer to work in this manner when
called upon to go into an unknown studio to work with an artist who
may be on tour and wishes to lay down tracks, or perhaps do additional
work on an album in progress. One great benefit for the producer/
engineer is the comfort of working with a consistent set of tools that can
be taken from one field location to another, reserving for a later time
all post-production activities that will take place in a known, home
Figure 8–16 shows front and back panel views of a high-quality
stand-alone microphone preamplifier.
To a very great extent, the recording engineer working in a well maintained studio does not have to worry about microphone cables, except to
make sure that they are in good repair. Things may not be so simple for
the engineer working in a specific remote location for the first time and
finding that he may actually run short of cable! There are two concerns
with long cable runs: phantom power operation and HF losses due to
cable capacitance.
Typical electrical values for professional quality microphone cable are:
Gauge: #24 AWG stranded copper wire
Resistance/meter: 0.08 ohms
Capacitance/meter: 100 pF
P48 phantom powering has a fairly generous design margin for long
cable runs. For example, a 100 m run of cable will have a resistance, per
leg, of 8 ohms, and the total resistance in the cable will be 16 ohms.
Considering the relatively high resistance of the phantom feed network
of 6800 ohms per leg, this added value is negligible.
More likely, HF rolloff will be encountered in long cable runs, as
shown in Figure 8–17. Here, data is shown for cable runs of 10 and
60 meters, with source impedances of 200 and 600 ohms.
Another problem with long cable runs is their increased susceptibility
to RF (radio frequency) and other forms of electromagnetic interference.
Local interference may arise from lighting control systems, which can
generate sharp “spikes” in the power distribution system; these can be
radiated and induced into the microphone cables. Likewise, nearby radio
6:05 PM
Page 132
transmitters, including mobile units, can induce signals into microphone
Cables of the so-called “starquad” configuration (developed early in
telephony) can reduce interference by up to 10 dB relative to the normal
two-conductor configuration. The starquad configuration is shown in
Figure 8–18. Here, four conductors within the braided shield are twisted
throughout the run of the cable. Diagonally opposite pairs are coupled
and connected at each end of the cable to pins 2 and 3. The twisting of
the pairs ensures that induced noise components are equal in each leg of
the balanced signal pair, resulting in cancellation of noise components
at the receiving end.
No engineer should ever stint on cable quality. The best cables available
are of the starquad configuration, supple, and are easily coiled. In normal
use, cables may be stepped on, crimped by doors and wheels on roll-about
equipment, and otherwise subjected to daily abuse. In general, braided
shield is preferable to wound foil shield; however, in permanent installations this may not be important.
Effects of cable length and
microphone impedance on
HF response.
Details of starquad
microphone cable
2:45 PM
Page 133
8: Electrical Considerations and Electronic Interface
Interconnecting hardware should be chosen for good fit, with an
awareness that not all brands of connectors easily work together, even
though they nominally meet the standards of XLR male and female
Microphone “snakes” are made up of a number of individual microphone cables enclosed in one outer sheath, normally numbering 12 to 16
pairs. In an effort to keep the size down, foil inner shields are normally used
on each pair. At the sending end it is preferable to have the snake terminate
in a metal box with XLR female receptacles, rather than a fan-out of female
XLR receptacles. This recommendation is based on ease of reassignment of
microphones by cable number, should that be necessary. At the receiving
end a generous fan-out should be provided for easy access to the console’s
inputs, with each cable number clearly and permanently indicated.
Capacitive coupling between pairs is generally quite low, and for signals at microphone level we can usually ignore it. However, for some
applications in sound reinforcement, there may be microphone signals
sent in one direction along the snake with concurrent line level monitoring signals sent in the opposite direction. This is not recommended in
general recording practice.
A microphone emulator circuit is shown in Figure 8–19. The circuit
shown is single-ended (unbalanced) and should not necessarily be used
to assess details of interference. However, it is excellent for determining
HF cable losses over long runs. The 40-dB input pad will reduce an
applied signal of, say 0.2 volts to 20 mV, a typical capacitor microphone
sensitivity rating at 1 pascal. The 20 mV signal appears across a 200-ohm
resistor and thus simulates the output of an actual microphone.
The circuit shown in Figure 8–20A is useful for simple cable continuity and leakage testing. Measurement across like pin numbers should
result in a virtual short-circuit. Measurements across different pin numbers should be made with the ohmmeter set to a high resistance range in
order to identify any stray leakage between wire pairs. The more sophisticated cable tester shown at B is typical of items offered by a number of
companies. As an active device it can be used to test the following:
Cable continuity
Cable intermittent failure
Phantom power integrity
In-circuit continuity (via test tones)
Cable grounding integrity
Three types of cables can be tested using this system.
Polarity checking of a microphone/cable combination can be made
using a system such as is shown in Figure 8–21. Here, a LF positive-going
2:45 PM
Page 134
An unbalanced microphone
emulator circuit providing a
loss of 40 dB.
Details of a simple continuity and leakage detector for microphone cables (A); a more complex cable testing apparatus (B).
(Photo B courtesy of Behringer Audio Technology.)
Details of a system for
checking signal polarity of
acoustical pulse is fed to the microphone and analyzed at the other end
of the cable run using an analyzer that detects the presence of a positiveor negative-going pulse. If negative polarity is indicated, then there is
near certainty that a cable has been miswired or that an item of in-line
equipment is inverting the signal. Virtually all modern amplifiers and signal processing devices are non-inverting. Thus, most electrical polarity
problems encountered today are the result of local wiring mistakes.
2:45 PM
Page 135
8: Electrical Considerations and Electronic Interface
While normal phantom powering can easily handle cable runs up to the
100 meter range, there may be some environments where RF interference
is a chronic problem. One option is to use stage microphone-to-line
preamplifiers, which will raise the level of the transmitted signals by 40
or 50 dB with very low output impedance, thus providing considerable
immunity to induced noise and losses due to long cable runs. Stage
preamplifiers are available with remote controlled gain adjustment.
Another option is to provide on-stage digital conversion for each
microphone and transmit the individual digital signals via fiber optic
cables to a digital console in the control room. In this form, the signals
may be sent over extremely long runs with no loss.
One of the most common audio transmission problems is the ground
loop. Figure 8–22 shows how a ground loop is created. Electronic
devices are cascaded as shown. Note that there is a ground path, not only
in the cables that connect the devices but also in the metal rack that
houses them. Any ac power flowing in the rack will generate an external
magnetic field, and that magnetic flux which flows through the loop will
induce a small current through it.
It is necessary to break the continuity in the loop, and this is normally
done as shown in Figure 8–23. Microphone cables are grounded at the
Generation of a ground
Breaking the ground loop.
2:45 PM
Page 136
input to the mixer in order to isolate the microphone from electrostatic
(RF) interference and also to provide a continuity path for phantom
powering. From the mixer onward, it is customary to connect the wiring
shield at the output end only as shown. In this manner the continuity of
the ground loop conductive path is broken, thus minimizing the severity
of the ground loop.
Many audio transmission problems begin with the microphone and the
initial input gain setting looking into the console. Modern consoles are
well engineered, and essentially all the operator has to do to adjust the
input and output faders to their nominal zero calibration positions and
then set normal operating levels on each input channel using the input
trim control. This may sound simple but it requires a little practice.
If the operator sets the microphone trim too low, then there is the
risk that input noise in the console will become audible, perhaps even
swamping out the noise of the microphone; if the trim is set too high,
then there is the risk that the console’s input stage will be overloaded on
loud input signals. Since the console has a wider overall operating
dynamic range than a microphone, the requirement is only to adjust the
trim so that the program delivered from the studio via the microphone
will fit comfortably within the total dynamic range of the console.
Figure 8–24 shows a level diagram for a single input channel
through a typical console. Let us assume that we have a microphone
whose sensitivity is 21 mV for an acoustical signal of 94 dB LP and whose
self noise floor is 10 dB(A). Further assume that the microphone’s maximum output level (0.5% THD) is 135 adB LP.
The microphone’s output is equivalent to 31 dBu, where dBu is
defined as:
dBu 20 log(signal voltage/0.775)
A level of 94 dB LP is typical of many instruments in the pop/rock
studio when picked up at fairly close quarters and at normal playing
levels, and thus the engineer will set the input trim control so that this
electrical level will produce console output meter deflections in the normal operating range.
Note also that with this input setting the microphone’s noise floor of
10 dB(A) will correspond to an electrical level 84 dB lower than
31 dBu, or 115 dBu. This level is about 13 dB below the noise floor
of the console, which is in the range of 128 dBu. Thus, it is clear that
the audio channel’s noise floor will be essentially that of the microphone,
with little contribution from the console’s input circuitry. As the signal
progresses through the remainder of the console, the noise floor does not
change relative to normal operating level.
2:45 PM
Page 137
8: Electrical Considerations and Electronic Interface
Typical console gain
structure or level diagram.
The microphone input channel has the capability, for undistorted console output, of handling an input signal that is 23 dB higher than 94 dB, or
117 dB Lp. If the microphone’s input signal exceeds this amount on a
continuing basis, then the engineer must adjust the input trim to accommodate it. If the signal exceeds this amount only occasionally then the
engineer may make necessary adjustments at the input fader.
The microphone’s undistorted output level extends up to 135 dB LP,
which is 18 dB higher than the console’s maximum output capability.
As such, this microphone may be used in a wide variety of acoustical
environments, and the only adjustments that need to be made are the
microphone’s output pad (for very loud signal conditions) or resetting
the input trim as required.
During remix operations, a number of microphones are often fed to a
given output bus. For each new microphone input added to a given bus it
is apparent that the overall input levels must be adjusted so that the signal
fed downstream is uniform in level. Figure 8–25 shows the way that individual inputs should be adjusted so that their sum will be consistent,
whether one or more microphone inputs are used.
2:45 PM
Page 138
Combining multiple inputs.
Recording engineers engaged in direct-to-stereo recording, or in
making stereo mixes from multitrack tapes, must be aware of the process
discussed here.
The capacitor microphone discussed in Chapter 3 operates on the principle
of variable capacitance with fixed charge producing a variable signal output voltage. There is another method of deriving a signal from variable
capacitance that has been successfully used by only a few microphone
manufacturers, and that is the RF transmission principle. Today, only
the Sennheiser company of Germany manufacturers such microphones on
a large scale.
The RF capacitor microphone has exactly the same acoustical characteristics as a dc polarized microphone of the same physical geometry. The
only difference is the method of converting the capacitance variations into
a signal output.
Hibbing (1994) gives a complete account of the history and analysis
of RF microphones. He describes two general methods, and simplified
circuits for each are shown in Figure 8–26. The circuit shown at A works
on the phase modulation (PM) principle and as such resembles a small
FM transmitter-receiver combination in a single microphone package.
The variable capacitance alters the tuning of a resonant circuit; adjacent
to the tuned circuit are a discriminator section and a fixed oscillator operating in the range of 8 MHz. All three sections are mutually coupled
through RF transformers. The alternations of tuning resulting from audio
2:45 PM
Page 139
8: Electrical Considerations and Electronic Interface
pressure variations affect the balance of upper and lower sections of the
discriminator, and an audio signal output is present at the output.
Circuit details of the AM bridge design are shown in Figure 8–26B.
Here, a dual-backplate push-pull diaphragm is used to advantage as a
push-pull voltage divider, distributing the RF signal equally to both sides
of the bridge circuit when in its rest position. Under audio excitation, the
bridge output will be proportional to diaphragm excursion, providing a
high degree of linearity. It is this modulation principle that is used in the
Sennheiser MKH-20 series of studio microphones.
Early RF microphones were prone to internal instability and on
occasion interference from local radio transmissions. Over the years the
art has literally “fine-tuned” itself, and the modern Sennheiser MKHseries microphones are superb performers in all respects.
While not generally recommended, it is possible to operate two microphones at a single console input if phantom power is not required.
The price paid for this is a reduction in level of about 6 dB for each
microphone and of course the inability to adjust levels individually. The
circuit shown in Figure 8–27 shows how this can be done.
RF microphones;
simplified circuit for phase
modulation system (A);
simplified circuit for
balanced bridge operation
(B). (Data after Sennheiser.)
A circuit for paralleling the
output of two microphones.
(Data after Shure Inc.)
2:45 PM
Page 140
So-called digital microphones have entered the marketplace during
the last six or so years. Strictly speaking, these models are not actually
“digital” in the specific sense of directly generating a digital output code
from the diaphragm. Rather, they make use of traditional dc bias and
preamplification of the analog signal at the diaphragm. It is only after
this stage that analog-to-digital conversion takes place.
The advantage of these microphones is that certain problems in digital processing can be dealt with earlier, rather than later, in the audio
chain. For example, the useful signal-to-noise ratio of a well-designed
25 mm (1 in) condenser diaphragm can be in the range of about 125 to
135 dB. An ideal 20-bit system is capable of a signal-to-noise range
of 120 dB, and in a traditional recording system this will require truncation of the available dynamic range of the microphone by about 10 dB.
In and of itself, this may or may not be a problem, depending on other
electrical and acoustical considerations in the actual studio environment.
In the beyerdynamic model MCD100 series, the capsule looks into a
22-bit conversion system directly when the acoustical level is high (greater
than 124 dB LP). For normal studio levels (less than about 100 dB LP), –10
or –20 dB padding can be inserted ahead of the digital conversion stage in
order to optimize the bit depth. Sophisticated level control prevents the
system from going into digital clipping. The microphone and associated
signal flow diagram is shown in Figure 8–28A and B.
The Neumann Solution-D uses two 24-bit A-to-D converters operating in parallel and offset by 24 dB. These two digital signals are seamlessly recombined in the digital domain to produce a single digital output
signal with a net resolution of 28 bits (Monforte, 2001). Figure 8–29A
shows a view of the Solution-D microphone, and a signal flow diagram
is shown at B.
Details of beyerdynamic digital microphone system: view of microphone (A); signal flow diagram (B). (Data courtesy of
2:45 PM
Page 141
8: Electrical Considerations and Electronic Interface
Details of Neumann
Solution-D digital
microphone system: view of
microphone (A); signal flow
diagram (B). (Data courtesy
of Neumann/ USA.)
Both of these microphone systems have additional digital features,
including variable sampling rates, various interface formats, some degree
of built-in digital signal processing, and the ability to respond to certain
user commands via the digital bus. The Audio Engineering Society (AES)
is actively pursuing interface standards for this new class of products.
2:46 PM
Page 142
Outside of recording and broadcast studios, the wireless microphone is
virtually indispensable. The technology dates back to the 1960s, and
great strides in performance quality and overall reliability have been
made since that time. Television, staged music and drama performances,
religious services, and public meetings all make use of wireless microphones; and the ultimate freedom of movement offered by the technology
is seen as a boon by everyone involved in live performance.
The earliest wireless microphones employed the commercial FM
(frequency modulation) band at very low output power, and consumer FM
tuners were used for signal recovery. In time, the Federal Communications
Commission (FCC) allocated specific frequency bands in the VHF and
UHF television ranges for wireless microphones as well as for other shortrange communications needs. Today, virtually all major microphone
manufacturers offer wireless systems, and the user has much to choose
from. Most of these manufacturers have published detailed user guides
that cover both technical matters as well as usage recommendations.
In this chapter we discuss in detail the technology involved in wireless
microphone design and its application in a number of performance areas.
Wireless microphones can be licensed for operation in the following
radio frequency (RF) ranges:
1. VHF (very-high frequency) range:
Low band: 49–108 MHz
High band: 169–216 MHz
2:46 PM
Page 143
9: Overview of Wireless Microphone Technology
2. UHF (ultra-high frequency) range:
Low band: 450–806 MHz
High band: 900–952 MHz
The FCC gives priority to primary users, such as TV, radio, and commercial communications activities such as cellular telephones, pagers,
and two-way radio applications. Wireless microphone application is
considered a secondary activity, and as such is not allowed to interfere
with primary activities. On the other hand, users of wireless microphones are often subjected to interference from primary users and must
find their own solutions to these problems through the choice of more
appropriate, trouble-free operating frequencies.
In the US, a manufacturer or distributor of wireless microphone
systems must obtain a license from the FCC to sell the equipment, but
it is the responsibility of the final user/purchaser to observe and follow
all recommendations regarding proper use of the equipment. The allocation of specific frequency bands varies widely around the world and
within given countries, and the user must rely on good advice from both
manufacturer and dealer in purchasing the correct equipment for a given
application in a specified location.
Regarding overall quality of performance, the UHF range is preferred
in that it provides excellent line-of-sight reception within a given venue
with minimum radiation beyond the normal architectural confines of that
In the future, wireless microphone operations will have to coexist
with the demands of digital television and expansion of other classes of
communication. Both manufacturers, specifiers and users of wireless
products must keep abreast of all developments and restrictions in these
areas. Figure 9–1 shows some typical international frequency allotments
Some international wireless
frequency allotments. (Data
courtesy of Shure Inc.)
2:46 PM
Page 144
for wireless microphone activity (Vear, 2003). A more comprehensive
listing is given in Ballou (2002).
Wireless microphones work on the principle of FM transmission, as
shown in Figure 9–2A. A carrier at the assigned VHF or UHF frequency
is modulated by the audio program and sent to an RF amplifier contained
in the microphone case itself or in a “body pack” worn by the user. The
RF signal is propagated through space by a transmitter operating in the
power range of 10–50 mW. For UHF transmission the radiating antenna
can be housed in the lower portion of the microphone case. The antenna
for the bodypack is usually in the form of a short wire that hangs externally from the case.
Signal reception is shown in Figure 9–2B. Here, the RF signal is amplified and amplitude-limited to reduce static interference; the recovered
Principle of FM
transmission and reception:
simplified block diagram
of a transmitter (A);
simplified block diagram
of a receiver (B).
2:46 PM
Page 145
9: Overview of Wireless Microphone Technology
carrier is then demodulated to produce audio output at the receiving end
of the system.
The simple system shown in Figure 9–1 has two major problems:
the potential loss of signal due to multiple reflections from transmitter to
receiver as the performer moves about the stage, and the susceptibility
to noise due to the low radiated power and low modulation index of the
transmitter as mandated by FCC standards and the requirements for
minimal interference with adjacent channels.
Diversity reception is used to overcome the multiple reflection problem, and a combination of signal companding (complementary compression and expansion) with pre- and post-equalization is used to reduce
With a single receiving antenna there exists the possibility that primary and
reflected carrier signals will arrive at the receiving antenna out-of-phase,
causing a momentary dropout of the carrier, as shown in Figure 9–3A. This
can be substantially avoided through the use of diversity reception. In
diversity reception there are two receiving antennas which are spaced by a
distance between one-quarter the carrier wavelength and one carrier wavelength, and this reduces the likelihood of signal cancellation. The equation
relating radiated wavelength to frequency is:
wavelength () c/f
where (Greek letter lambda) is wavelength, c is the speed of radio
waves (300,000,000 meters per second), and f is the frequency in Hz.
In the frequency range of 900 MHz, carrier wavelength is in the
range of about 3 108 to 9 108 meters (about 13 in), so an antenna
spacing of about 8–35 cm (3–13 in) can be used to substantially reduce
carrier cancellation at the antennas. This is shown in Figure 9–3B, with
the two receiving antennas (labeled A and B in the figure). Most modern
diversity receivers are of the so-called “true diversity” type, in which the
signals picked up by both antennas are simultaneously demodulated by
two separate receivers so that the switching function from one to the
other, depending on signal strength, is virtually instantaneous and free of
any switching noises.
To reduce the basic noise floor of the relatively narrow deviation FM
transmission system used with wireless microphones, complementary
signal compression and expansion, known as compansion, are used.
Figure 9–4 shows, in effect, how this works. The input signal varies over
a range of 80 dB and is compressed in the transmitting circuitry to occupy
a range of about 40 dB. After demodulation the signal is expanded by
40 dB to restore the original dynamic range.
2:46 PM
Page 146
Diversity reception; single
receiving antenna (A); dual
(diversity) receiving antenna
reception (B).
Principle of operation of
signal companding
2:46 PM
Page 147
9: Overview of Wireless Microphone Technology
Typical HF pre-emphasis
(A) and de-emphasis (B)
curves used with wireless
Companding is normally coupled with complementary signal preemphasis and de-emphasis in order to minimize the audibility of noise
floor modulation by the complementary gain shifts. Typical fixed preand de-emphasis curves are shown in Figure 9–5.
Considering the combined effects of diversity reception, companding,
and complementary pre- and de-emphasis, the subjective dynamic range
of a wireless microphone channel is in the range of 100 dB, considering
an A-weighted noise floor as measured under no-signal conditions.
There are two basic transmitter units: the handheld microphone and the
body pack. These are shown in Figure 9–6 along with a receiving unit.
The microphone transmitter is convenient for most applications, and a
number of manufacturers provide a basic body on which a chosen microphone capsule assembly can be fitted.
For those applications where the microphone must be concealed
(drama and musical shows), or where the user demands the freedom of
both hands, the body pack is used. The body pack is worn by the user,
normally in a pocket or strapped to the user’s belt. A thin wire connects
a standard tie-tack microphone to the body pack. Most body packs also
have a line input option for use in picking up amplified instruments
which may be carried around on-stage.
The transmitter’s antenna wire must hang free in order to ensure
proper operation. Those systems operating in the UHF high band will
require only about an 8 cm (3 in) antenna, and these are often built
directly into the microphone transmitter.
A transmitter of either type will normally have the following controls
(the nomenclature may vary slightly from one manufacturer to another):
1. Frequency selection. This is fixed inside the transmitter and is not
readily accessible to the user.
2:46 PM
Page 148
Photograph showing a
typical wireless
microphone, bodypack
and receiver. (Photo
courtesy of Shure Inc.)
2. Sensitivity control. This is in effect a volume control that determines the level of the microphone signal reaching the transmitter
input. It can be adjusted for each user.
3. Power on-off. The power should of course be switched off when
the microphone is not in use in order to conserve battery life. In
actual operation however, the wearer should be instructed to leave
the microphone power on at all times in order to ensure program
4. Microphone mute. The mute switch should be engaged by the
wearer only when there is a lull in use, so that normal handling
noises will not be transmitted. The wearer must of course be aware
of the microphone’s status so that it can be turned on when needed.
5. Status lights:
Microphone on/off.
Low battery condition. Normally, a blinking of the on/off light
is used to indicate low battery power.
Signal overload. Normally, a red light will indicate the presence
of too-high signal peaks, warning the user to talk at a lower level
or to readjust the sensitivity control.
2:46 PM
Page 149
9: Overview of Wireless Microphone Technology
Today, receivers are fairly small, usually no more than one standard rack
unit high and one half-rack unit in width. While each unit can accommodate a pair of small diversity antennas, it is usually more convenient
to feed all receivers from a central diversity antenna pair with a splitter/booster unit, both of which may be located at the top of the rack
containing the receivers.
For each microphone or body pack in use there will of course be a
corresponding receiver. It is possible to assign multiple transmitters to
a single receiver, but it is obvious that only one transmitter can be used
at a given time.
Receivers vary in complexity and some will have more controls and
indicators than others. The following are considered essential:
1. Frequency selection. Must match that of the transmitter.
2. Squelch/mute threshold control. This control is used to set the RF
sensitivity of the receiver so that it will pick up the desired local
signal, but will mute when that signal is turned off.
3. Audio level. This control should be set for the desired output level
from the receiver. It is usually set for normal line output level.
4. Status lights:
Mute indicator. Indicates when the receiver is not picking up the
assigned transmitter.
RF level indicator. Indicates the level of incoming RF signal; a
persistently low level indicates the possibility of signal muting.
AF (audio frequency) level indicator. Indicates the level of audio
being fed downstream.
Diversity operation. Under normal operating conditions there
should be a continuous toggling action between the two indicator lights, indicating that diversity decisions are being made more
or less randomly – which is the intended operation of the system.
Since all wireless microphones and body packs are battery operated,
there must be a good supply of fresh batteries available at all times.
Rechargeable batteries may, over time, represent a saving over the cost
of non-rechargeable batteries. However, the non-rechargeable batteries do
offer the benefit longer operating time and slightly higher output voltage
than the equivalent rechargeable units. Figure 9–7 shows the expected
active life of several battery types. You can easily see that, in a very long
evening’s show, if an actor, singer, or master of ceremonies is on stage
continuously for up to three hours, it will be essential to outfit the transmitter with a standard alkaline non-rechargeable battery.
2:46 PM
Page 150
Operating life of various
battery types. (Data
courtesy of Shure Inc.)
For the average performance lasting about three-plus hours, any
rechargeable batteries that are in use will have to be replaced at some point
around the middle of the show, and this may not always be convenient to
do. For each rechargeable battery in use, it will be necessary to have two
others on hand – one freshly charged and the other being charged.
Furthermore, non-rechargeable batteries can be bought in bulk with considerable cost savings. The bottom line here is to weigh the pros and cons
of rechargeable versus non-rechargeable batteries very carefully.
It is essential that one person be placed firmly in charge of all wireless microphone operations in a given application involving more than a
few microphones. The major responsibility is to ensure that there will
always be a supply of freshly recharged batteries and that there be a rigorous schedule for all users to have batteries replaced when needed as the
show goes on.
On a more global basis, the wireless specialist must also be aware of
what channels are actually available for use in all geographical venues in
which a road show may be scheduled. Also, the wireless specialist must
be aware of any other shows or events that may be scheduled in adjacent
show or sports facilities. Not all problems can be encountered and corrected during an afternoon dress rehearsal. Patterns of RF interference
often change in the evening and must be anticipated.
Essentially, wireless microphones rely on line-of-sight between transmitter
and receiver. The RF signal will travel easily through curtains and thin wallboard materials provided that there is no metal mesh present. Ordinarily, a
wireless microphone should not be used within a distance of 3 m (10 ft)
of a receiver, due to the possibility of overloading the receiver’s first RF
amplification stage.
Normal usage may extend up to 300 m (1000 ft) if the venue is
relatively free of RF interference. In very large venues it may be advantageous to have several transmitter/receiver groups, just to make sure
that excessively long distances will not be encountered. Likewise, it is
important to ensure that all “roving” microphones will perform well
when tested over their entire anticipated range of movement. Do not hesitate to reassign frequencies if transmission problems are encountered.
Often, moving the receiver rack a small distance will cure a transmission
2:46 PM
Page 151
9: Overview of Wireless Microphone Technology
problem. While the actual number may vary, it is generally possible to
use up to 30 wireless microphones simultaneously in a given venue. In
fixed installations remember that wireless microphones should be used
only where there is a need for user mobility. Microphones used solely at
pulpits, podiums, and lecterns should all remain wired.
Do not treat the specification and application of wireless microphones as
an afterthought or as a trivial matter; there is too much that can go
wrong! Never make the mistake of believing that you can learn all you
need to know about these systems in one simple training session. We
strongly recommend that you acquire the wireless usage manuals available from many manufacturers. In particular we recommend the comprehensive guide written by Tim Vear (2003) for Shure Incorporated.
2:47 PM
Page 152
With the exception of handheld applications, microphones are rarely used
without accessories of one kind or another. Most accessories are used in
mounting the microphone in a stable configuration, ranging from desk
stands and flush mounts to large microphone stands or booms. Hanging
mounts are often used where stands are not appropriate. Under conditions of wind or mechanical instability, wind screens and shockmounts
are essential. Numerous electrical expedients, such as padding (electrical
attenuation), polarity inversion, and the like may be conveniently met
with in-line adapters. Likewise, stand-alone phantom power supplies for
capacitor microphones are often useful. In this chapter we will cover the
entire range of accessories.
The once familiar desk stand, shown in Figure 10–1A, has largely given
way to flush-mounted microphones in the boardroom and to the ubiquitous telephone handset in modern transportation terminals. Many desk
stands today are considered too obtrusive and have been replaced by
models with a thin gooseneck extension and small capsule. This type of
assembly (Figure 10–1B) can be either mounted in a heavy base or permanently mounted on the working surface of a lectern. In either case, the
microphone itself can be easily positioned for the talker.
Purely for the sake of nostalgia, some late-night TV talk show hosts
may have a vintage microphone sitting on their desk. Look carefully; you
will note that the performer is usually wearing a lapel microphone!
The “mike mouse,” shown in Figure 10–2, is made of open cell foam
and is a convenient way to keep a small microphone positioned close to
2:47 PM
Page 153
10: Microphone Accessories
Microphone desk mounts:
standard models (A);
dedicated models with
slender gooseneck and
microphone sections (B).
(Photo A courtesy of
Electro-Voice; Photo B
courtesy of AKG Acoustics.)
a surface in a stable position. These used to be familiar items in board
rooms and on the stage floor in performance venues. The microphone’s
position at a relatively large boundary results in minimal reinforcements
and cancellations due to reflections from the floor surface to the microphone. Today, we are more likely to see a very low profile boundary layer
(BL) microphone for such applications as these.
2:47 PM
Page 154
Foam floor-mount adapter,
top and bottom views.
(Photo courtesy of
Microphone stands come in all sizes and configurations and can span
adjustable heights from about 0.5 meter to 5 meters (1.5 ft to 15 ft), as
shown in Figure 10–3A. The high stands require a large footprint for
safety, and take great care when using them with large format microphone
models. These stands are useful where sound pickup in front of and above
an ensemble is sufficient. Where there is a need to position the microphone
within an ensemble, then a horizontal boom assembly attached to the
stand will allow positioning the microphone over the players. A comfortably large footprint is necessary, as is an appropriate counterweight to balance the microphone. Such booms are a mainstay in the popular studio.
Do not stint on quality. Good stands and booms are not cheap, and
well-made models will last indefinitely. A number of recording engineers
who are in the remote recording business have taken a page from those in
the motion picture trades. They have gone to grip manufacturers (suppliers
to the motion picture trade for stands, props, reflectors, scrims, and the like)
for purpose-built microphone stands at relatively reasonable prices.
For motion picture and video field sound pickup, telescoping handheld booms, as shown at B, are typical. Typical usage of a boom is shown
in Figure 10–3C. A good operator skillfully keeps the boom-microphone
assembly out of the picture, closely following the talkers and keeping the
microphone pointed at them.
A stereo mount is a piece of hardware that will accommodate two
microphones on a single stand, positioning them securely for stereo
pickup. The model shown in Figure 10–4A is articulated so that both
microphone angle and spacing can be adjusted. The assembly shown at
2:48 PM
Page 155
10: Microphone Accessories
Typical microphone stands
and boom assemblies (A);
telescoping handheld booms
as used in film and video
work (B); boom in
operation (C). (Photo A
courtesy of AKG Acoustics;
Photo B courtesy of M.
Klemme Technology.)
2:48 PM
Page 156
Stereo mounts; articulated
type (A); dedicated ORTF
assembly (B); closely
spaced, rotatable assembly
(C); long adjustable stereo
bar (D); and Decca tree (E).
(Photo B courtesy of
Schoeps GmbH; Photo C
courtesy of AKG Acoustics;
Photos D and E courtesy of
Audio Engineering
2:48 PM
Page 157
10: Microphone Accessories
B provides ORTF mounting for Schoeps cardioid capsules, and the
assembly shown at C provides XY or MS pickup using small format
capacitor microphones. The horizontal bar shown at D may be used for
a wide range of spacing between the stereo microphone pair. A so-called
Decca tree is shown at E and is used in front of large orchestral groups
in the manner of the British Decca record company. Small arrows indicate the positions of the three microphones.
Music schools and festival sites have a need for quasi-permanent microphone mounting so that ensemble and recording changes on-stage can
be made quickly. While there will always be a requirement for floor
stands as needed, overhead requirements can normally be met by a system of breast lines, pulleys, and cables attached to relatively permanent
microphone receptacles. Most of these systems have been improvised
and fine-tuned over time. Figure 10–5 shows details of a three-way
Details of a three-point
winch system for
positioning a microphone.
(Data after EMT.)
2:48 PM
Page 158
winch system designed some years ago by the German EMT company.
This figure will give the enterprising designer an idea of the complexity
involved. Such a system as shown here would be useful with a remotely
adjustable stereo microphone and is ideal for use with the Soundfield
For many performance applications, one or more single hanging microphones may be used with the arrangement shown in Figure 10–6A. This
mount uses a clamping action on the cable, providing a service loop of
cable that permits the microphone to be angled as needed. A horizontal
breast line can be tied to the cable loop for added positioning and minimization of cable twisting over time.
The method shown at B relies on a flexible wire to position a
small electret microphone at the desired angle. An array of such hanging
microphones would be effective for choral pickup in a house of worship.
The nature of wind noise is shown in Figure 10–7. A puff of wind
travels perhaps 1.5 m/s (5 ft/sec), while sound travels at a speed of
344 m/s (1130 ft/sec). Furthermore, the motion of wind occurs along a
narrow path straight ahead of the talker while speech radiates relatively
uniformly over the forward hemisphere.
Cable mounts; a cable
clamp for swiveling and
tilting of microphone (A);
flexible wire for aiming
small microphone (B).
(Photo B courtesy of Crown
2:48 PM
Page 159
10: Microphone Accessories
Action of wind and speech
on a microphone; the
nature of wind noise.
Relative effects of wind
on noise generation in
omnidirectional and
hypercardioid microphones
(A); noise generation in a
cardioid microphone with
and without a foam
screen (B). (Data after
Wuttke, 1992.)
Most users not familiar with microphones feel that they must talk
directly into the microphone; the important thing is that the microphone
be pointed at their mouth, and that is best done if the microphone is
placed to one side or slightly above the talker along the centerline and
aimed at the talker’s mouth.
Figure 10–8 shows some spectral details of wind noise. For a fixed
wind velocity, the data at A shows that the difference in the noise generated
2:48 PM
Page 160
by a hypercardioid microphone is about 20 dB higher than with an
omnidirectional microphone. The data shown at B shows the performance
of a cardioid microphone with and without a foam windscreen. Note that
the effectiveness of the windscreen diminishes at low frequencies. In order
to maintain the high midband reduction in wind noise, the size of the wind
screen has to be increased substantially.
Windscreens can be problematic. They nearly always affect frequency response, and if they are too dense they will affect directionality
as well. Wuttke (1992) points out that any windscreen used with a gradient microphone should have a clear inner open cavity in the region of
the microphone element so that the gradient effect can establish itself
directly at the microphone’s transducing element.
Wind effects do not only occur outdoors. The author has encountered several concert halls in which high volume velocity, low particle
velocity air handling systems produced wind noise in gradient microphones positioned 3–4 m (10–13 ft) above the floor.
A typical foam windscreen for handheld applications is shown in
Figure 10–9A. An oblong mobile screen is shown at B, and a so-called
Windscreens; a typical
foam screen for handheld
or studio use (A); handheld
screen for field use (B);
“furry shroud” for field use
(C). (Photo A courtesy of
AKG Acoustics; Photo B
and C courtesy of
2:48 PM
Page 161
10: Microphone Accessories
“furry shroud” is shown at C. These assemblies are intended for field use
in television and motion picture recording and are normally used with
rifle microphones mounted on booms or poles.
A nylon pop screen is shown in typical studio usage in Figure 10–10.
It is a virtually perfect antidote to the “popping” sounds of b and p
from vocalists and announcers in studio applications. It is semitransparent, so that the artist can easily see the microphone, and it has
virtually no attenuation at high frequencies as it effectively disperses
any puff of wind in the direction of the microphone. It is normally
attached to microphone hardware by a small clamp or by a flexible
wire, as shown.
FIGURE 10–10
A nylon pop screen in
normal operating
position. (Photo
courtesy of Shure Inc.)
2:48 PM
Page 162
Significant floor vibration transmission through the microphone stand
is not often a problem, but stands may be inadvertently hit by studio
performers. When this occur some form of shock mounting is necessary. Figure 10–11 shows the transmission curve for a typical vibration isolation system. Note that below the system resonance frequency,
fn, the vibration transmission coefficient is constant. Above the resonance frequency the transmission coefficient falls steadily to a low
value. In practice, we would prefer fn to be low enough so that negligible audible vibration will be transmitted to the microphone. In order
to do this we need a vibration isolator, commonly called a shock
A typical assembly is shown in Figure 10–12. A normal isolation
tuning frequency for a large format microphone might be in the 8 to
12 Hz range, well below the normal hearing range. Note that the microphone is entirely suspended by the elastic bands and is not in direct contact with the frame. Note also the microphone cable hanging from the
bottom of the microphone. If for any reason the microphone cable is
pulled taut and fastened to the stand, then the effect of the entire shock
mount may be negated. There should always be a loose service loop of
cable as it exits the microphone, and that loop should be taped to the
microphone stand or boom to avoid any downward gravity loading on
the microphone due to excessive cable length.
Microphones and their shock mounts are designed as a system. It
may be possible to mix and match models to some degree, but if the
resonance varies substantially the effectiveness of the shock mount may
be compromised.
Foam rubber isolating rings are routinely used in microphone
mounts that are sold with microphones. Because of their relatively low
mechanical compliance, these devices account for little in the way of
effective vibration isolation.
FIGURE 10–11
Universal vibration
transmissibility curve
through a shock mounting
2:48 PM
Page 163
10: Microphone Accessories
FIGURE 10–12
A typical large format
capacitor microphone and
its associated shock mount.
(Photo courtesy of
Figure 10–13 shows a number of in-line passive devices that can be
used in microphone circuits. The assemblies shown at A are known as
“turnarounds” and may be used when, for whatever reasons of
miswriting or design, XLR cables do not mate. The assemblies shown
at B are polarity or phase inverters and are used to correct polarity
mistakes in wiring. Some older line-level electronic items have inverted
outputs that may require correction. Polarity inverters may be used at
microphone level with phantom powering, but cannot be used with
An in-line balanced loss pad is shown in Figure 10–13C. Loss
values may be in the range of 20 dB or greater. The assembly may be
used with dynamic microphones, but not with any kind of remote
2:48 PM
Page 164
FIGURE 10–13
Various electrical
accessories; male-female
“turnarounds” (A); polarity
inverter (B); loss pad (C);
low-to-high impedance
transformer (D); and
low-pass filter (E).
powering for capacitor microphones. An in-line transformer, shown at
D, can be used to match a low impedance dynamic microphone with a
high impedance input in some nonprofessional applications. The in-line
LF-cut equalizer shown at E likewise may be used in nonprofessional
applications to correct for proximity effect in dynamic cardioid microphones. The microphone splitter was discussed in Chapter 8, under
Microphone Splitters.
Details of a phantom power supply are shown in Figure 10–14. This unit
enables a P-48 microphone to be used with a small PA-type mixer that
may not have built-in phantom powering.
2:48 PM
Page 165
10: Microphone Accessories
FIGURE 10–14
Photo (A) and circuit
diagram (B) for AKG
Acoustics N62E auxiliary
phantom power supply.
(Figure courtesy of AKG
2:50 PM
Page 166
Modern stereophonic recording, or stereo as we normally call it, makes
use of many diverse microphone arrays and techniques. At the basis of
them all are a set of fundamental two- or three-microphone arrays for
picking up a stereo sound stage for reproduction over a pair of loudspeakers. In stereo reproduction the listener is able to perceive images on
the stereo sound stage which may span the entire angular width of the
loudspeaker array. The sound sources that are perceived between the
loudspeakers are known as “phantom images,” because they appear at
positions where there are no physical, or real sources of sound.
It is essential that the reader has an accurate, if only intuitive, knowledge
of how phantom images are formed. In Figure 11–1 we show how a set
of real sound sources located progressively from the center to the rightfront of the listener produce time-related signals (phasors) at the ears.
At frequencies below about 700 Hz, each single sound source (S1
through S5) produces a signal at the right ear that leads the slightly
delayed signal at the left ear. The ears detect this as a phase difference at
low frequencies. At high frequencies (above about 2 kHz), the ears rely
primarily on the amplitude differences at the two ears, and a louder signal will appear at the right ear due to the shadowing effect of the head.
These two sets of cues reinforce each other, and the ear–brain combination interprets the differences at the ears as a set of apparent sources
spanning the front to front right.
For a sound source (S1) located directly in front of the listener, the low
frequency phasors will be equal at both ears, as will be the shadowing at
2:50 PM
Page 167
11: Basic Stereophonic Recording Techniques
Phasor analysis for real
sources with positions
ranging from center to
right-front of the listener.
high frequencies. These cues reinforce the judgment that the sound source
is directly ahead of the listener.
With only two loudspeakers not exceeding a spread of about
60 degrees, we can create low frequency phasors for sound sources that
span the entire listening angle between the loudspeakers. To create the
phasors for a sound source directly ahead of the listener, all we need to
do is feed the same signal to both loudspeakers, as shown in Figure 11–2.
Here, each ear receives two signals, one slightly delayed with respect to
the other. Since the program content of the signals is the same, we can
combine their phasors at each ear to create new phasors, shown here as
LT and RT. Since these two phasors are equal (that is, they have the same
amplitude and phase relationships), the ear–brain combination will
interpret them as producing a sound source directly in front of the
listener – as long as the listener is facing forward.
By varying the amounts of left and right components, phasors can be
created that will simulate all positions between the left and right loudspeakers. As we will soon see, a single pair of microphones can easily create this effect across the left-to-right span of the playback loudspeaker
2:50 PM
Page 168
Two loudspeakers
creating equal net phasors
at both ears.
In addition to the effect of low frequency phasors, small discrete signal
delays between the two microphones at high frequencies can also influence phantom source localization. In normal recording operations, both
amplitude and time delay cues are at work, and their combined effect on
phantom source localization can be roughly estimated from the data presented in Figure 11–3. This data was presented by Franssen (1963) and
gives reasonably accurate estimated stereo stage localization for wideband signals such as the spoken voice. An example showing use of the
graph is given in Figure 11–3.
In subsequent discussions in this chapter we will refer to approximate localization analyses as determined by Franssen’s method. These
will be shown as horizontal plots with specific indications of left, leftcenter, center, right-center, and right images as determined from the data
in Figure 11–3.
Coincident microphone arrays consist of a pair of directional microphones located virtually one atop the other and both individually
adjustable in their lateral pickup angles. As such, they respond only
to amplitude cues in the program pickup, since their proximity precludes
any time related cues. Coincident arrays are often referred to as X-Y
2:50 PM
Page 169
11: Basic Stereophonic Recording Techniques
Franssen’s data for
determining wide-band
stereo localization with
both amplitude and
time cues.
The most notable example of the coincident array is a pair of figure-8
microphones arrayed at a lateral angle of 90, as shown in Figure 11–4A.
This array, first described by Blumlein (1931, reprinted 1958), is unique
in that the sine and cosine angular relationships between the two pickup
2:50 PM
Page 170
pattern lobes retain constant signal power over the pickup angle due to
the relationship:
(sin )2 (cos )2 1
This identity ensures that any sound source located in the front
quadrant of the array will be picked up with uniform power, since
acoustical power is proportional to the square of its respective components. Specifically, when a sound source is directly in the middle of the
array, the values of both sin 45 and cos 45 will be 0.707, representing
a 3 dB reduction in level at both left and right loudspeakers. These two
half-power quantities will add acoustically in the listening space, yielding a power summation of unity. Localization of the Blumlein array is
shown as determined by the Franssen graph shown in Figure 11–4B.
Any discussion of the Blumlein array leads directly to a discussion
of the panpot (panoramic potentiometer), a dual fader arrangement that
has one signal input and two signal outputs. Circuit details are shown
in Figure 11–5A, and the two outputs are as shown at B. Panpots are
found in the microphone input sections of virtually all recording consoles and can be used to steer a given input signal to any position on the
two-channel stereo sound stage. The two fader sections give sine and
cosine values corresponding to the rotary fader setting, providing uniform acoustical power whatever the setting. In this regard the panpot is
analogous to the Blumlein crossed figure-8 microphone pair operating
in its front or back quadrant.
The crossed figure-8 array maps the frontal quadrant of its pickup
with virtual accuracy. Likewise, the back quadrant is equally well picked
up, but in opposite polarity to that of the front quadrant. The left and
right side quadrants are picked up with their two components in reverse
polarity (often described as out of phase or in antiphase), and this creates a spatial ambiguity. As long as the side quadrants and the back
quadrant are relegated to “back of the house” reverberation, there
should be no adverse problems due to the polarity reversal.
The crossed figure-8 array is also known by the term Stereosonic
(Clark et al., 1958) as it was modified and applied by the Electrical and
Musical Industries (EMI) during the early days of the stereo LP in
England. In their application of the Blumlein array for commercial recording, EMI recording engineers always introduced a slight amount of
in-phase crosstalk above about 700 Hz between the stereo channels. The
technique, known informally as “shuffling,” was used to match phantom
image localization at high frequencies with that normally produced by low
frequency phasor reconstruction at the listener’s ears (Clark et al., 1958).
In summary, Blumlein array performance:
1. Produces excellent stereo stage lateral imaging, due to the self
“panning” aspect of the frontal sine and cosine pickup lobes. Image
analysis by means of Franssen’s data indicates that the primary
sound sources will fill the entire stereo stage from left to right.
2:50 PM
Page 171
11: Basic Stereophonic Recording Techniques
Crossed figure-8s (A);
localization via
Franssen (B).
2. Conveys an excellent sense of acoustical space, due to the added
pickup of reflected and reverberant signals in the recording space.
3. Often presents difficulties in microphone placement; specifically, a
wide array of performers may require that the microphone pair to
be placed too far from the performers for the desired degree of
2:50 PM
Page 172
Details of the panpot:
circuit (A): signal
outputs (B).
Variable crossed cardioids
(A); localization via
Franssen’s data for 90
cardioids (B); variable
crossed supercardioids (C);
localization via
Franssen’s data for 120
hypercardioids (D).
Figure 11–6A shows a crossed pair of cardioids that can be splayed over
an angle of 90 to 135. When the splay angle is about 90 the sound
stage appears center-heavy (i.e. mono-dominant), and sources located at
far-left and right will appear panned well into the array. This is not a useful configuration, and most engineers will choose to widen the microphone splay angle up to 135. This has the effect of correcting mid-stage
balance with the sources at far left and right. But even then, the overall
stereo stage may appear a bit too narrow.
A better approach is shown in Figure 11–6C, where a pair of splayed
supercardioid microphones are used at an angle of 120. Here, the relatively narrow front lobe will spread left, center, and right stage events
naturally, while the smaller back lobes will pick up more room reflections and reverberation. In general, the crossed supercardioids, or hypercardioids, make a good alternative to crossed figure-8 patterns in live
recording spaces, since the rear lobes will pick up less room sound and
2:50 PM
Page 173
11: Basic Stereophonic Recording Techniques
its concomitant noise. Figure 11–6B and D shows the above two orientations as they appear in the Franssen localization analysis.
In summary:
1. Crossed cardioid performance produces a center-oriented stereo
stage and, as such, may be a useful adjunct in studio pickup.
2. It results in wider imaging when the splay angle is increased.
3. It has excellent monophonic compatibility (left-plus-right), due to
the absence of antiphase components in the two signals.
4. Splayed supercardioids or hypercardioids are probably the most
useful, offering a good compromise between wide-stage pickup and
direct-to-reverberant balance.
Figure 11–7A shows a typical M-S recording setup. A cardioid pattern
(the M component) is forward facing, and a figure-8 pattern (the S component) is side facing. Both M and S components are often recorded separately and are combined in a matrix (sum-difference) network as shown
to produce resultant left and right pickup patterns. An alternate matrix
transformer circuit is shown at B.
The system offers excellent monophonic compatibility and considerable postproduction flexibility. For example, the S component can be
reduced to produce a narrowing of the resultant array. Alternatively, the
S component can be increased to produce a wider resultant array.
Primarily as a result of this degree of postproduction flexibility, M-S
pickup has always enjoyed great favor with broadcast engineers, who
have to deal with concerns of FM and television mono and stereo compatibility in a wide marketplace.
Any XY pair can be converted into an equivalent MS pair, and vice versa.
Examples of this are shown in the data of Figure 11–8. The charts shown
at B and C can be used to define the MS form for any given XY pair.
While a cardioid pattern is normally used for the M component, it is
not essential; omnidirectional or hypercardioid patterns can be used as
well if the resultant left and right patterns and their angular orientation
fit the technical needs of the engineer. Those readers who wish to study
M-S recording in detail are referred to the work of Hibbing (1989) and
Streicher and Dooley (1982, 2002).
In summary, M-S performance:
1. Has excellent stereo-to-mono compatibility for broadcasting.
2. Has flexibility in postproduction remixing.
3. Is easy to implement in the field.
2:50 PM
Page 174
M-S system: a frontfacing cardioid and
side-facing figure-8, when
mixed as shown, are
equivalent to a pair of
splayed cardioid
microphones (A); alternate
method of achieving sum
and difference outputs
using transformers (B).
The beginnings of stereo recording with spaced microphones go back to
the experiments carried out by Bell Telephone Laboratories (Steinberg
and Snow, 1934) during the 1930s. Snow (1953) describes a wall of
microphones, each communicating with a matching loudspeaker in a different space, as shown in Figure 11–9A. As a practical matter this array
was reduced to the three-microphone/loudspeaker array shown at B.
Later, this three-channel array gave way to an arrangement in which the
center microphone was bridged (center-panned) between the left and
right channels.
There is a strong tradition of using spaced omnidirectional microphones for stereo recording, especially for smaller musical forms, such as
chamber music and solo piano. In Figure 11–3 we observed the stereo
stage as recreated by a single pair of omnidirectional microphones separated by about 0.6 m (2 ft). When such a pair is in proximity to a small
group of players, then both amplitude cues and time cues can be useful
2:50 PM
Page 175
11: Basic Stereophonic Recording Techniques
Several MS forms and their
equivalent XY forms (A);
conversion from XY to MS
(B and C); example: let an
XY pair consist of a two
supercardioids splayed at
an angle of 120 (2); from
B, locate the supercardioid
symbol and the right axis
and move to the left unto
the 60 curve is intersected;
then, read downward to
the horizontal axis and
note that the required M
directivity pattern will be
almost halfway between
supercardioid and figure-8;
then, going to the same
position on the horizontal
axis at (C), move upward
until the 60 curve is
intersected; finally, move to
the right axis and note that
the S level will have to be
raised about 2 dB relative
to the M level. (Data at B
and C courtesy of
in defining the reproduced stereo stage. The images so produced may
not have the localization precision or specificity of those produced by a
coincident pair, and many people, engineers as well as musicians, often
prefer the so-called “soft edged” quality of these recordings. The analogy of slightly blurred optical imaging is quite apt.
Actually, there is no reason why spaced cardioids should not be
used – specially in large reverberant spaces. Traditionally, however, most
recording engineers who prefer the imaging qualities of spaced microphones tend to favor omnidirectional models or subcardioids.
2:50 PM
Page 176
Spaced microphones; a
wall of microphones and
loudspeakers (A); three
microphones and three
loudspeakers (B). (Figure
courtesy of the Society of
Motion Picture and
Television Engineers.)
If two omnidirectional microphones are spaced apart more than
about 1 m (3.3 ft), there may be a “hole in the middle” of the reproduced
sound stage. This occurs because of reduced signal correlation between
the two microphone transmission channels, and hence very little that the
ears can lock onto to form center-oriented phantom images. A practical
example of a two-microphone setup is shown in Figure 11–10A. Here,
the performers in a string quartet are arrayed left to right, and the
Franssen analysis of stereo localization is shown at B with a 0.67 meter
microphone spacing. If the microphone spacing is increased to, say,
1 meter, the Franssen analysis of localization gives the results shown in
Figure 11–10C. Here, the hole in the middle becomes quite evident.
For large groups such as an orchestra, it is essential that there be a
center microphone panned equally between the two stereo channels.
When this is done, the hole in the middle is effectively filled in, and the
listener senses a continuum of sound from left to right, albeit with some
loss of image specificity. Another important characteristic of such wide
microphone spacing is the generation of added early time cues. As an
example, consider a left-center-right microphone array spanning a total
width of about 8 m (25 ft). A sound source on-stage fairly close to the
right microphone will be heard first in the right loudspeaker, followed
2:50 PM
Page 177
11: Basic Stereophonic Recording Techniques
FIGURE 11–10
Two spaced omnis with a
string quartet (A); Franssen
analysis for 0.67 m spacing
(B); Franssen analysis for
1 m spacing (C).
closely by a center phantom signal delayed approximately 12 ms and a
later signal at the left loudspeaker delayed a further 12 ms. These two
delayed signals will be at a reduced level, and they will effectively simulate early acoustical reflections in the recording space itself, as shown in
Figure 11–11A. Their effect on the subjective impression of music may
be seen in the data shown in Figure 11–11B (Barron, 1971), where they
may influence the perception of tone coloration.
A typical large ensemble may be recorded as shown in Figure 11–12.
The proportions are those the author has found to work in most
instances. In normal acoustical spaces, omnidirectional microphones will
work well, but in rooms that tend to be fairly live, subcardioids may
work better because of their back rejection of about 3 dB. The level setting of the center microphone is critical; it should be at least about 4 dB
lower in level with respect to the left and right microphones and should
be just sufficient to avoid the sense of a hole in the middle. Any greater
level of the center-panned signal is apt to make the stereo stage appear
too mono-heavy.
It is worth mentioning that most American symphonic recordings
made during the decades of the fifties and sixties were made basically in
this manner. These recordings have always been present in commercial
disc libraries, and they are as highly regarded today as they were in their
2:50 PM
Page 178
FIGURE 11–11
Widely spaced microphones
may simulate the effect of
early acoustical reflections
(A); Barron’s data on the
subjective role of reflections
in a performance
environment (B).
FIGURE 11–12
Three spaced omnis with
2:50 PM
Page 179
11: Basic Stereophonic Recording Techniques
In summary, in spaced omnidirectional recording using two microphones:
1. Microphones are normally spaced no farther apart than about 1 m.
2. The array can be placed relatively close to a small group of performers, creating a good sense of intimacy while retaining an
impression of room ambience.
3. The technique is useful primarily when precise imaging of instruments on the stereo sound stage is not essential.
In spaced omnidirectional recording using three microphones:
1. Wide spacing of microphones creates the effect of added early
acoustical reflections, thus enhancing a sense of ambience.
2. The center microphone “locks” the middle of the ensemble in
place, but often with relatively little precise localization.
3. Ample pickup of room reflections and reverberation retains a good
sense of acoustical space, along with significant direct sound from
the ensemble.
Near-coincident techniques employ a pair of directional microphones,
closely spaced and splayed outward, and as such combine some of the
attributes of both coincident and spaced microphone techniques. Some
options are shown in Figure 11–13. The near-coincident microphone
pairs combine both amplitude and delay effects in producing a stereo
sound stage. They are useful in a wide range of acoustical settings and
may be altered in both spacing and splay angle as required. Actually,
there are innumerable possible configurations, and the four shown here
are examples of some that have been described in the literature.
The ORTF (Office de Radio-Television Diffusion Française) technique was developed by the French Broadcasting organization, while the
NOS (Nederlandsch Omroep Stichting) technique was developed by the
Dutch broadcasting organization. The Faulkner and Olson (1979)
Stereo-180 arrays were developed by independent recording engineers.
Many engineers and musicians favor the ORTF array in particular and
use it in place of the more traditional coincident cardioid array (Ceoen,
1970). Figure 11–14 shows a Franssen analysis of localization properties
of the ORTF and NOS arrays.
Since the formalization of the ORTF near-coincident approach more
than 30 years ago, many engineers have experimented with their own
splay angles and microphone separation, perhaps in an effort to define
2:50 PM
Page 180
FIGURE 11–13
Near-coincident pairs.
ORTF (A); NOS (B);
Faulkner (C); Olson (D)
(cross-hatching indicates
antiphase pickup relative to
frontal lobe).
FIGURE 11–14
Franssen analysis of
ORTF and NOS pairs.
their own unique stereo signature. Williams has analyzed many microphone patterns in various splay angles and separation with the aim of
defining reasonable bounds on their performance. The data shown in
Figure 11–15 shows the useful range of cardioid microphone separation
and splay angle (Williams, 1987). The effective recording angles are indicated on the curves themselves for a variety of separation and splay angle
values. Those values in the crosshatched areas are not recommended
2:50 PM
Page 181
11: Basic Stereophonic Recording Techniques
FIGURE 11–15
Williams’ data showing
useful combinations of
separation and splay angle
for a cardioid microphone
because of problems in stereo imaging, ranging from insufficient stereo
separation to a tendency for separation to be polarized at the loudspeakers.
In summary, near-coincident microphone performance:
1. Combines the image specificity of coincident arrays with the
enhanced sense of image spatiality imparted by spaced microphones.
2. Allows considerable leeway in choice of splay angle and microphone spacing for modifying the recorded perspective.
Most engineers who make commercial recordings use mixed arrays in
which a central pair, either coincident or near coincident, is combined with
a flanking omnidirectional pair. The combination provides excellent flexibility and allows the engineer to alter perspectives in the orchestra without
necessarily making adjustments in the microphone positions themselves.
For example, a shift from a close-in perspective to a somewhat more
distant perspective can be made merely by altering the amount of the
spaced pair in the overall mix. Care must be taken not to exceed a level
range here of perhaps 1.5 to 2 dB, otherwise the changes may be too
apparent. Of course it goes without saying that such changes, if of a running nature throughout a large work, must be musically pertinent in the
first place.
Some of the more often used setups of this type are shown in
Figure 11–16. The ORTF-plus-flanking-omnis setup is shown at A. As a
2:50 PM
Page 182
matter of taste, some engineers may replace the ORTF pair with a pair
of widely splayed (120) hypercardioids, and the omnis may be replaced
with subcardioids in especially live rooms.
The variation shown at B makes use of four subcardioids. The center pair may be spaced up to 0.5 meter (20 in), and the splay angle
between them set in the range of 120. This mixed array provides good
flexibility, but overall it tends to pick up considerable room ambience
and reverberation.
The so-called “Decca tree” is shown at C. Developed by the English
Decca Record Company in the 1950s, the array makes use of three identical microphones on a “tree” in the center, with the middle microphone
gain set slightly lower than the left and right microphones. Flanking
FIGURE 11–16
Mixed arrays: ORTF plus
flanking omnidirectional
microphones (A); four
subcardioids (B); the Decca
tree (C).
2:50 PM
Page 183
11: Basic Stereophonic Recording Techniques
microphones of the same type are also used. The microphone of preference
here has been the Neumann M-50, which was discussed in Chapter 3. The
M-50 is essentially an omni microphone at low frequencies, exhibiting a
rising HF response above about 1 kHz and reaching a level 6 dB higher
than at low frequencies. Additionally, the M-50 tends to become more
directional as its on-axis response increases. The array produces a sound
rich in room ambience, along with an abundance of detail at high frequencies. The only difficulty associated with use of the Decca tree is suspending
the three central microphones in a stable manner. A little care is needed,
along with a fairly robust microphone stand.
The width of the stereo included listening angle is a subject of considerable discussion. Generally, stereo localization is at its most stable when
the loudspeaker–listener angle is moderate. At the same time, however,
the angle must be large enough to convey a realistic performance stage.
Most professional recording engineers would agree that an included
listening angle not exceeding about 45 or 50 is about optimum.
It goes without saying that the listening environment should be fairly
symmetrical about the centerline of the loudspeaker array and that the
listening space be free of noticeable flutter echoes and strong standing
waves. Regarding frequency response of the loudspeakers as they react
with the listening space, the response should be fairly uniform over the
range from about 160 Hz to about 4 kHz. Below 160 Hz a broad rise of
about 2 or 2.5 dB may enhance the presentation, especially at moderate
listening levels. Above about 4 kHz many listeners prefer a uniform
rolloff of the response slightly (no more than 2 or 3 dB). Above all, use
the best loudspeakers that your budget can maintain.
2:51 PM
Page 184
For many applications of coincident stereo or M-S recording, a stereo
microphone may be the best choice. A stereo microphone normally
embodies two closely spaced capsules arrayed one over the other, with
individual pattern adjustment for both capsules and provision for rotating one of the capsules with respect to the other. This degree of electrical
and mechanical flexibility is necessary in order to cover all aspects and
conditions of coincident recording.
Departing slightly from the notion of capsule coincidence, a number
of dual-microphone mounting techniques have been developed for stereo
recording and are embodied in acoustically appropriate structures. Most
of these are spherical, some purposely head-shaped, while others are
merely some kind of baffle with microphones on opposite sides. While
most of these embodiments work well enough for direct stereo pickup,
some may be more suited to the requirements of binaural recording.
In addition, there are a handful of compound microphones that have
been developed to pick up sound via four or more capsules for purposes
of surround sound pickup; these will be discussed in a later chapter. In
this chapter we will discuss the development of the conventional stereo
microphone and describe several commercial embodiments in detail.
Figure 12–1 shows how Blumlein (1931), using only a pair of omnidirectional microphones, synthesized a pair of cardioid patterns splayed at
90 below about 700 Hz, working in conjunction with a pair of leftright omnis shadowed by an absorptive baffle that separated them. The
shaded omnis were effective only at frequencies above about 1500 Hz.
This was in fact the first coincident stereo microphone array, and very
likely the first instance of generating a cardioid pickup pattern via the
electrical addition of figure-8 and omnidirectional components.
2:51 PM
Page 185
12: Stereo Microphones
Blumlein’s stereo recording
technique of 1931.
The first commercial stereo microphones appeared in the fifties and were
made by Georg Neumann and Schoeps in Germany and AKG Acoustics in
Austria. Two well-known current models are shown in Figure 12–2. The
cutaway detail of the Neumann SM-69 shows the vertical arrangement of
the Braunmühl-Weber dual diaphragm capsule assemblies, with the upper
one free to rotate as required. The AKG Acoustics model C-426 is
equipped with a set of lights to aid in azimuth adjustments in field setup.
2:51 PM
Page 186
Stereo microphones based
on the Braunmühl-Weber
dual diaphragm assembly;
the Neumann model
SM-69 (A); the AKG
Acoustics model 426 (B).
(Photos courtesy of
Neumann/USA and AKG
Figure 12–3 shows the Royer model SF-12 ribbon microphone. The
model consists of a pair of ribbon elements permanently mounted at 90
and dedicated to the Blumlein crossed figure-8 format. It has an XLR
five-pin output receptacle that fans out to a pair of XLR three-pin receptacles for normal console use.
The Josephson C700S is an unusual model that contains an omni
element and two figure-8 elements fixed at angles of zero and 90.
Because of this, it can perform as a stereo microphone or as a first-order
surround microphone. The three capsule outputs are combined in a flexible control unit to provide five simultaneous signal outputs. The choices
here are all first-order patterns individually ranging from omnidirectional to figure-8. The microphone is shown in Figure 12–4A, and the
signal flow diagram of the controller is shown at B.
Operating in a stereo mode, the system provides an X-Y pair of outputs whose patterns can be continuously varied from omni to figure-8
and whose included angle can vary from zero to 180, with all adjustments made entirely at the control unit. Operating in an M-S mode, a
2:51 PM
Page 187
12: Stereo Microphones
forward-oriented pattern of the user’s choice and a side oriented figureeight can be derived.
Operating in a surround mode, or as an adjunct to a larger studio
pickup plan, up to five outputs can be derived. An interesting application
here would be arraying the five outputs at angles of 72 and setting all
patterns roughly to hypercardioid. This could be used to achieve horizontal (azimuthal) plane pickup of ambient sound.
In this regard, the C700S behaves virtually like a Soundfield microphone operating in the horizontal plane. (See Chapter 15 under The
Soundfield Microphone.)
Stereo microphones that use dual diaphragm elements are normally operated with a remote control electronics unit that provides phantom powering and switchable pattern control. In most designs, the microphone
capsule assemblies themselves are rotated manually. Many remote control
units provide for Left/Right outputs or Mid/Side outputs, at the choice of
the user. Figure 12–5A shows front and rear photos of the Neumann CU
48i remote control unit, with a circuit diagram shown at B.
As a matter of convenient interface, there are two outputs from the
control unit terminating in conventional XLR-M connectors that accommodate P-48 phantom powering from the downstream console. The
microphone assembly itself is fed by a shielded 10-conductor cable which
is available in several interconnecting lengths to adapt to a variety of
installation requirements.
Royer model SF-12 ribbon
microphone. (Photo
courtesy of Royer
At one time, an accurately scaled artificial head containing two pressure
microphones at ear position was thought of only in the context of binaural recording. The basic binaural technique is shown in plan view in
Figure 12–6A, and a typical binaural head is shown at B. Here, the threedimensional sound field in the original recording space is accurately
mapped into the listener’s auditory domain via correct time cues and frequency response shaping around the head, along with equalization to
compensate for the secondary signal path between headphones and the
ear drums. Artificial head assemblies are manufactured by, among
others, Georg Neumann GmbH, Knowles Electronics, Brüel & Kjaer,
and Head Acoustics GmbH. In addition to binaural recording, artificial
head systems are widely used in psychological acoustics and evaluation
of performance spaces.
Binaural recordings are intended for headphone reproduction and
do not normally translate well in direct stereo playback, primarily
because of the limited channel separation at low frequencies when
auditioned over stereo loudspeakers. It is possible, however, to convert a
2:51 PM
Page 188
Details of the Josephson Engineering model C700S stereo/surround microphone system;
photo of microphone (A); signal flow diagram (B). (Data courtesy of Josephson Engineering.)
binaural recording into stereo by using a crosstalk canceling method,
such as that shown in Figure 12–6C. Because the head is small with
respect to the wavelengths of frequencies below about 700 Hz, sound
bends around the structure with little attenuation. There is however a
significant phase difference at the two microphones due to their spacing,
and this phase difference can be effectively converted to an amplitude
difference by applying equalization along with sum and difference networks. The technique has much in common with the basic Blumlein
setup shown in Figure 12–1.
Some manufacturers offer spherical baffles. These are similar to artificial heads, but are normally used directly for stereo recording with no
requisite LF crosstalk cancellation. Significant work in the development
of these recording methods has been carried out by Günther Theile
(1991). A further distinction is that their pressure microphones are
2:51 PM
12: Stereo Microphones
Page 189
Stereo microphone remote
control unit: photo of
Neumann CU 48i remote
pattern selector (A); signal
flow diagram of CU 48i
(B). (Data courtesy of
mounted directly on the spherical surface, resulting in relatively flat
response. A typical example of this is the Neumann KFM-100 system
shown in Figure 12–7A. Frequency response and polar response in the
horizontal plane are shown in Figure 12–7B and C.
Figure 12–8A shows a view of the Crown International Stereo
Ambient Sampling SystemTM (SASS). Small omnidirectional microphones
are positioned on each side at the apex of the foam-lined structure,
roughly at ear distance. The shading of the unusual structure results in
the polar patterns shown at B, which indicate a gradual transition above
500 Hz to a clear 45 pair of pickup patterns above about 3 or 4 kHz.
Various forms of flat baffles have been used over the years to
increase separation between near-coincident stereo microphones at high
frequencies. In the modern stereo era, Madsen (1957) appears to have
been the first to make use of such a baffle, as shown in Figure 12–9A. A
modern example here is the Jecklin disc (1981), shown at B. Here, an
acoustically damped disc with a diameter of 250 mm (10 in) is placed
between two omnidirectional microphones. For sounds originating at
2:51 PM
Page 190
90 to the array the delay between the two microphones will be about
0.7 ms, which is only slightly greater than the equivalent delay around
the human head. Shadowing effects for the far-side microphone will be
somewhat more complicated than for a head or a sphere, due to the
rather complex nature of diffraction effects around a plane as opposed
to a spherical or other smooth three-dimensional surface.
There are many competing stereo systems embodying spheres and
many types of baffles. The differences among them can be fairly subtle
and have to do basically with dimensions (most do not depart from
2:51 PM
12: Stereo Microphones
Page 191
Basic binaural recording
signal flow diagram (A);
photo of the Neumann KU
100 binaural head (B);
binaural to stereo
conversion (C). (Photo
courtesy of
normal head dimensions significantly), materials (absorptive or not), and
careful equalization in the microphone channels.
2:51 PM
Page 192
Neumann KFM 100 sphere
(A); frequency response (B);
polar response (C). (Data
courtesy of Neumann/USA.)
The Crown Internal SASS
system: photo of structure
(A); typical polar response
(B); circuit diagram of
system (C). (Data courtesy
of Crown International.)
2:51 PM
Page 193
Use of baffles in stereo microphone pickup: Madsen’s application with a pair of ribbon microphones (A); details of the
Jecklin/OSS disc (B). (Figure at A courtesy of Journal of the Audio Engineering Society.)
2:53 PM
Page 194
The topics covered in the previous two chapters form the basis for most
of our study in this chapter. We begin with a basic discussion of the
physical characteristics of musical instruments, moving on to a discussion of case studies with reference recordings made by the author. We
progress from solo instruments, through chamber ensembles, to large
orchestral resources, with emphasis on the practical decisions that every
recording project requires. A final section deals with the adjustment of
room acoustical conditions to fit the recording project at hand.
Our recommendations for microphone placement throughout this
chapter remain the same whether you are recording direct-to-stereo or to
multitrack. The difference of course is that multitrack offers unlimited
opportunities to fix your mistakes later – rather than fixing them with
retakes during the sessions.
One of the first things an aspiring recording engineer learns is that the best
seats in a concert hall are not necessarily the best places to put microphones. Those seats after all are where, by common agreement, orchestral
balance is judged to be ideal. Many of us have tried at least once to place
microphones at favored listening positions and found that the recording
sounded too reverberant, was dull at high frequencies, had insufficient
stereo stage width and, during quiet passages, was somewhat noisy.
Why should this be so? As we listen in a concert hall we are able to
distinguish direct sound cues (those that first reach our ears from the
stage), early reflections (those that help define the apparent stage width),
2:53 PM
Page 195
13: Classical Stereo Recording Techniques and Practice
and global reverberant cues (those that give an impression of the size of
the performance space). It is also an advantage that we can turn our heads
from side to side in order to further reinforce these cues and ultimately to
“zero in” on stage events and hear them in a familiar environment, even
in the presence of a reflected sound field that may be several times greater
in terms of acoustical power than the direct sound field itself.
A pair of stereo channels will have a difficult time duplicating this,
since so wide a range of desirable spatial cues cannot be reproduced as
such with only two loudspeakers. There are also performance constraints
in the consumer playback environment, be it one’s living room, automobile, or portable Walkman.
As a result, modern classical recording practice remains what it has
always been: a close-up event with the principal microphones placed at
Row A and a playback effect in the home that seems to give the impression of being seated at Row J. How this is done will become clear as this
chapter progresses. First, we need to understand some of the acoustical
characteristics of musical instruments, including their directional aspects
and dynamic ranges.
In general, instruments radiate forward and upward relative to the
player’s seating position; however, all musicians appreciate the reflectivity of an uncarpeted performing area. At low frequencies all instruments
are essentially nondirectional. As we progress upward in frequency,
radiation becomes more pronounced along the bell axes of brass instruments, while the directionality of woodwind instruments becomes fairly
complex. Figures 13–1 and 13–2 illustrate this.
String instruments radiate in a complex manner. At low frequencies
they are nearly omnidirectional, and at mid-frequencies there is a preference for radiation perpendicular to the top plate, or “belly”, of the
instrument. At the highest frequencies the radiation pattern becomes
broad again as radiation is largely from the very small bridge of the
instrument. These general trends are shown in Figure 13–3.
The various keyboard and percussion instruments generally have
complex radiation characteristics due to their shapes and sizes and often
interact significantly with their environments.
Figure 13–4 shows the dynamic ranges characteristic of the major
instrumental groups. Within a given frequency range, the dynamic capability may be no more than 35 or 40 dB at most; however, over a large
frequency range the total dynamic capability can be considerable. The
French horn is probably most notable here, covering a range of about
65 dB from the lowest to highest frequency ranges, but limited to about
35 dB within any given frequency range. Of all orchestral instruments
the clarinet has the widest overall dynamic range within a given frequency range, nearly 50 dB in the instrument’s middle range.
2:53 PM
Page 196
Polar response of brass
instruments along the bell
axis; values of f are given
for three brass instruments.
(Data after Benade, 1985.)
Directionality of
woodwind instruments:
basic characteristic (A);
off-axis measurement
graphs (B and C). (Data
after Benade, 1985.)
2:53 PM
Page 197
13: Classical Stereo Recording Techniques and Practice
The following discussions will largely detail the author’s experiences
in recording more than 280 compact discs. These are not the only
approaches, and the aspiring engineer is encouraged to experiment with
other approaches. References made to coincident pickup apply as well to
near-coincident pickup and to head-related techniques, as discussed in
preceding chapters.
One of the biggest problems in recording the piano may be the condition
of the instrument itself. Most good concert grand pianos are associated
with performance venues or schools of music and as such may be voiced
2:53 PM
Page 198
Directionality of string
instruments: violin in
transverse vertical plane
(A); violin in azimuthal
plane (B); cello in
horizontal plane (C).
Dynamic ranges versus
frequency; violin (A);
flute (B); oboe (C); clarinet
(D); trumpet (E); French
horn (F).
2:53 PM
Page 199
13: Classical Stereo Recording Techniques and Practice
for projection in recital or concert performance spaces. Often, this means
that the piano may sound a little aggressive close in. If this is the case,
then a technician should be asked to tone down the instrument to whatever degree may be necessary. Normally, this requires that only the upper
three octaves may need to be adjusted.
The piano can be recorded best in recital halls of moderate reverberation time, as opposed to a typical concert hall. Some microphone
approaches are shown in Figure 13–5; A and B show coincident pickup
while C and D show spaced omnidirectional microphones. With spaced
microphones, it is important not to overdo the spacing, and you should
take care that the microphones be approximately equidistant from the
sounding board to ensure reasonably coherent pickup of the lower vibration modes of the sounding board. Listen to various spacings before you
commit yourself.
In most cases it would be a mistake to move much farther from the
instrument than shown in the figure, the danger being pickup of too
much reverberation. When using a coincident approach, take the time to
experiment with various microphone splay angles and pickup patterns
with the aim of achieving the proper stereo stage width.
Some producers and engineers prefer to remove the piano’s cover
and place the microphones nearly overhead. This will result in increased
low frequency response and may or may not be appropriate for the
repertoire. Proceed with caution.
If possible, a stand-by technician-tuner should be on hand during the
sessions to do touch-up tuning or take care of any mechanical noises that
may develop in the instrument. Damper pedals are often noisy and
require attention. Often, a small piece of carpet under the pedal lyre will
“tame” the noise produced by an overactive peddler.
The harpsichord creates a shower of higher harmonics and relatively
little fundamental. As such, it can sound with great clarity in the fairly
reverberant environments which are normally chosen to enhance
baroque and classical literature for the instrument. The same general
pickup approach for the piano may be used here, with a slight preference
for spaced omni microphones. The harpsichord key action may generate
2:53 PM
Page 200
Recording the piano;
using coincident or
near-coincident microphones
(A and B); using spaced
microphones (C and D).
a good bit of noise, which can usually be alleviated with a 60 or 80 Hz
high pass filter.
Regarding the stereo soundstage presentation, a piano should appear
to be spread over a center portion about two-thirds the total stage width,
with reverberant cues coming from the entire stage width. (See Compact
Disc Reference 1.)
Again, the choice is between coincident and spaced microphones, and the
basic options are shown in Figure 13–6. With coincident pickup, watch
out for proximity effect, and be prepared to equalize it accordingly. If the
primary microphones do not pick up enough reverberation, you can use
a spaced pair of room microphones to enhance the sound. Do this only
if the room is quiet enough; otherwise, consider using a high quality digital reverberation generator with appropriate room parameters programmed into it. (See Compact Disc Reference 2.)
2:53 PM
Page 201
13: Classical Stereo Recording Techniques and Practice
Recording the guitar: using
coincident or near-coincident
microphones (A); using
spaced microphones (B).
Of all instruments, the organ is the only one that does not “travel.”
Unless it is a small portative instrument which is often used for baroque
music accompaniment, the organ is permanently built into a large space
and voiced for that space. No two organs are even remotely alike, and
the choice of venue is equally important as the choice of performer and
literature. The organ is usually heard at a large distance, and specific
directional effects from the instrument are the exception. The listener
thus hears the organ and its environment as an entity, rather than as an
instrument independent of its surroundings.
In many cases a single pair of omnidirectional microphones will suffice to pick up the instrument at a distance up to 6 or 10 m (20 to 33 ft).
If the instrument’s keyboard divisions are laid out horizontally, then a
coincident pair can be used to delineate lateral directionality. Usually the
instrument is arrayed vertically, and lateral imaging will be lost as such.
In some cases an instrument installed in the rear gallery of a church
consists of a main set of pipework along with a rückpositiv division suspended on the gallery railing. In this case the fore-aft imaging aspects
will be readily apparent in the recording. A suggested starting point is
shown in Figure 13–7.
2:53 PM
Page 202
Recording the organ:
elevation view (A); section
view (B).
In many European cathedrals the organ is located very high in the
rear gallery, as much as 15–20 m (50–65 ft) above the floor of the nave.
There are no microphone stands tall enough to put the microphones
where they need to be, and the normal approach is to hang microphones
from the ceiling or from lines suspended from side to side at the height
of the triforium gallery. Such preparations take time and can result in
microphone cable runs up to 100 or 150 m (330 to 450 ft).
If the engineer is required to work from the floor it may be useful to
experiment with both omnidirectional microphones on high stands,
along with a widely spaced pair of line microphones aimed at the instrument. These will provide added directionality at high frequencies and
may help to clarify the texture of the instrument.
A secondary pair of microphones in the middle of the venue may
flesh out the reverberant signature of the space, should that be needed.
These microphones may be rolled off at low frequencies to minimize LF
room rumble and to maintain clarity. In some spaces that are large but
not sufficiently reverberant, an artificial reverberation generator may be
useful. However, not many reverberators can duplicate accurately the
long reverberation time associated with large spaces. (See Compact Disc
Reference 3.)
Chamber music may be roughly defined as music written for groups
numbering from two to perhaps twelve performers, with one performer
on each part and normally performing without a conductor. For smaller
chamber groups there are basically two choices for the recording engineer: to record the group in a concert setting as they would normally be
seated on-stage; or to record them in a studio setting arrayed in an arc,
or even facing each other. There are advantages to both approaches.
The concert setup will be familiar to the players and as such may be
preferred at the outset. Conventional microphone placement would call
2:53 PM
Page 203
13: Classical Stereo Recording Techniques and Practice
Recording the piano trio:
concert setting (A); studio
setting (B).
for a coincident pair, perhaps with an additional flanking omnidirectional pair. This approach will make the ensemble sound “staged”; that
is, the listener will hear the group with a “Row G” perspective. If more
intimacy is desired the group can be placed in an arc or even a circle,
with microphones more closely placed. Figure 13–8 illustrates both
approaches. At A we see a piano trio (piano, violin, and cello) in a normal concert setup, with a frontal array of microphones. At B we show
an alternate approach of positioning the violin and cello players so that
all three members of the group now have direct eye contact. (See
Compact Disc Reference 4.)
A typical listener reaction to these two recording perspectives is that
the former places the listener in the concert hall (“you are there”), while
the latter places the performers in the listener’s living room (“they are
here”). The players, once they have adjusted to the alternate seating, usually become partial to it. The engineer and producer favor it because it
gives them more flexibility in establishing and altering balances as the
recording gets underway.
Recording a solo instrument or vocalist with piano is greatly helped
by departing from the concert setup, as shown in Figure 13–9. Here, the
vocalist has been placed a comfortable distance from the piano and is facing the pianist, thus ensuring good eye contact. Both piano and soloist are
individually picked up in stereo, and a mix is created in the control room.
If good judgments are made in balance and placement, everything will
sound perfectly natural. This approach allows the engineer and producer
to achieve the desired balance between soloist and piano at all times.
The question of when to use the piano with its cover on half-stick
often comes up quite often. In concert, half-stick may be necessary to
keep the instrument from overpowering soloists. However, in the studio
environment, it is preferred to keep the cover at full-stick and achieve the
necessary balances purely through placement of microphones and performers. (See Compact Disc Reference 5.)
The string quartet is one of the most enduring musical ensembles of
all time, and the medium has a rich and complex repertoire that continues
to grow. Normally the players are seated as shown in Figure 13–10A.
While the traditional frontal coincident array is often used, as shown at
B, many engineers favor an array of spaced omnidirectional microphones,
2:53 PM
Page 204
Recording vocalist with
piano in the studio:
microphone positions (A);
console assignments (B);
vocalist’s orientation to
microphone (C); distance x
is normally in the range of
0.5 m (20 inches) to 1 m
(40 inches) (C).
as shown at C. The center microphone would normally be mixed in at a
level about 4 to 6 dB relative to the two others and panned just
slightly right of center. Its purpose is to anchor the center of the stereo
stage (primarily the cello) without appreciably narrowing the overall
stereo stage width. Many engineers use a cardioid microphone for the
center pickup. The two flanking omnis should be positioned so that they
preserve the balance of the entire group. (See Compact Disc Reference 6.)
The chamber orchestra normally numbers from 25 to 40 players,
depending on the repertoire. The majority of the literature for the chamber orchestra is from the late 18th and early 19th centuries. The ensemble is best recorded with four microphones across the front, as discussed
in the previous chapter, with the addition of accent microphones as
needed. Accent microphones, also familiarly called “spot mikes”, are
used to add presence to an instrument or an orchestral section without
necessarily increasing loudness. On a small reflective stage they may not
be needed at all. But under most conditions we would use a secondary
2:53 PM
Page 205
13: Classical Stereo Recording Techniques and Practice
FIGURE 13–10
Recording the
string quartet; normal
concert seating (A); use
of coincident or
microphones (B); use of
spaced omnidirectional
microphones with an
accent microphone for the
cello (C).
coincident stereo pair for picking up the woodwind section, along with
single accent microphones for the first two stands of bass viols, harp and
celesta. The timpani are normally loud enough, but an accent microphone
may help bring out the articulation of the drums. A typical seating
arrangement is shown in Figure 13–11. (See Compact Disc Reference 7.)
Accent microphones are normally cardioids and are generally mixed
into the main array at levels according to the technique shown in
Figure 13–12. Many engineers prefer to delay individually all of the
accent microphones so that their signals will be essentially time-coherent
with the pickup of the main pair, as shown at A. That is, if an accent
microphone is located, say, 8 m from the main pair, it may be delayed by
8/344 seconds, or about 23 ms, to align the pickup with its acoustical
signal path to the main pair. While it is always correct to delay accent
microphones, it may not be necessary because of the masking effect of
the louder pickup from the main microphone array. Data shown at B
indicate approximately when delay is mandatory, while data shown at
C show the approximate level relationships between accent microphones
and main microphones.
2:53 PM
Page 206
FIGURE 13–11
Recording the chamber
orchestra; normal seating
FIGURE 13–12
Implementing accent
microphones: calculating
microphone delay (A);
determining when delay is
and is not essential (B); level
range for accent
microphones (C).
Remember that it is very important to pan the positions of the accent
microphones so that they correspond exactly to the physical position of
the instruments as heard on the stereo stage when no accent microphones are engaged. It is also common practice to attenuate slightly the
low frequency response of the accent microphones; this makes it easier
to add presence without adding loudness.
The full symphony orchestra may have a complement of strings of
14-12-10-8-8 (respectively, the number of first violins, second violins,
violas, cellos, and basses). Woodwinds may number from 8 to 16,
depending on the nature of the score. French horns may number anywhere from 4 to 12, depending on the score, and the heavy brass may
number 4 trumpets, 3 trombones, and a single tuba. The percussion
2:53 PM
Page 207
13: Classical Stereo Recording Techniques and Practice
resources, including timpani, may call for 4 players. Add to all of this the
requirements of perhaps two harpists and a keyboard player (celesta
and/or piano), and the head count can easily reach or exceed 90 players.
A symphony orchestra takes up much floor space, and a good rule is
to multiply the number of players by two to arrive at the approximate
number of square meters of total occupied floor space, taking into
account a normal complement of players in each section. Many modern
stages are not quite large enough to fit this paradigm, and many orchestral recordings routinely take place instead in large, fairly reverberant
ballrooms or other large meeting spaces. This is especially true in
England and on the European continent. Churches are often called into
duty for large-scale recording, but the excessive reverberation time of
many of these venues can be a problem.
Figure 13–13 shows plan and side views of the recording setup for a live
(with audience) recording of Gustave Holst’s “The Planets”, a typical
large orchestral work dating from 1916. Note that there are 14 microphones. Of these, the four across the front provide the basic pickup; all
others are accent microphones.
The main four microphones consist of an inner ORTF pair with a
flanking pair of omnidirectional microphones. The ORTF pair is aimed
downward at a point about two-thirds the depth of the stage, and the
intent here is to avoid picking up too much of the front of the orchestra
by relying on the off-axis attenuation of the cardioid patterns to balance
the front-back distance ratio.
The accent microphones are detailed below:
1. Harps: two instruments with a single cardioid placed between them.
2. Celesta: a single cardioid to delineate this soft instrument.
3. Horns: a single cardioid placed about 4 m (13 ft) above the section
and aimed downward toward the edge of the bells.
4. Woodwinds: an ORTF pair aimed at the second (back) row of
5. Timpani: two sets of drums were used, and a single cardioid microphone was placed midway between them about 2.5 m (8 ft) above
the stage.
6. Brass: a single cardioid placed about 4 m above the players, aimed
7. Basses: a single cardioid placed overhead about 1.5 m (5 ft) from
the first stand of players.
8. House pair: two widely spaced cardioids hanging from the attic and
positioned about 7 m (23 ft) above the audience and about 8 m
(26 ft) from the edge of the stage. These microphones were aimed
at the upper back corners of the hall.
2:53 PM
Page 208
FIGURE 13–13
Dallas Symphony
Orchestra, recording
Holst’s The Planets: Side
view (A); plan view (B).
All accent microphones were operated with a slight LF rolloff below
about 100 Hz. With the exception of the house pair, they were all fed in
varying amounts to a digital reverberation unit whose parameters were
set to match those of the occupied hall. The stereo output of the reverberation unit was returned to the two-channel stereo mix at an appropriate level – no more than to ensure that the effect of the accent
microphones blended with the natural pickup provided by the main four
2:53 PM
Page 209
13: Classical Stereo Recording Techniques and Practice
microphones across the front of the orchestra. An off-stage chorus was
located in the stage house behind the orchestra and the sound was introduced into the hall via the stage doors. No additional microphones were
used here.
The choice of instruments and sections to be given preference with
accent microphones was made in consultation with the recording producer, taking into account overall musical requirements of the score.
While the main microphones were virtually fixed in level, the individual
accent microphones were adjusted on a running basis by no more than
about 2.5 dB, according to musical requirements. (See Compact Disc
Reference 8.)
In many modern concert halls a large chorus may be placed in a choral
terrace behind the orchestra, calling for an additional stereo pair of
microphones. In many older halls, the orchestra is moved forward and
the chorus placed on risers positioned behind the back row of orchestra
players. If the hall is of the proscenium type, the chorus often finds itself
in the very dead up-stage acoustical setting, calling for its own microphones (two or three in stereo) along with the correct amount of artificial reverberation to match them with the normal orchestra pickup.
A recurring problem with recording works for chorus and orchestra
is leakage of the back of the orchestra (brass and percussion) into the
choral microphones. This can be alleviated by elevating the choral microphones as high as is practicable by overhead rigging.
If there is no audience present, an excellent alternative is to place the
chorus in the house behind the conductor and picking it up with its own
dedicated stereo microphone array. The effect is excellent in that the
chorus has immediate acoustical contact with the hall.
The main and woodwind ORTF microphone pairs are aimed downward
in order to adjust front–back balances in their respective pickup patterns.
This procedure is detailed in Figure 13–14.
The directional patterns are of course cardioid, and the idea here is
to trade off the balance of the nearer instruments with those toward the
rear by taking advantage of the off-axis attenuation of the microphones’
patterns. Considering the main ORTF pair, as shown in Figure 13–14, the
elevation angle has been set so that the primary axis of the microphones
is aimed at the woodwinds, brass and horns. When this is done, the offaxis pattern attenuation of the frontal strings will be about 2.5 dB.
However, the front of the orchestra is about one-third the distance from
the microphones relative to the more distant woodwinds, brass and
horns. The net pickup difference will then be 9.5 2.5, or about 7 dB,
favoring the front of the orchestra. We must remember however that both
woodwinds and brass instruments are essentially forward facing with
2:53 PM
Page 210
FIGURE 13–14
Downward elevation of
main ORTF microphones.
directivity indices that are in the range of 4 or 5 dB. Thus, with their
direct sound virtually aimed at the frontal microphones, the front–back
orchestra balance will generally seem quite natural. In some cases both
engineer and producer may wish to emphasize the strings even more by
lowering the elevation angle to favor the strings to a greater degree.
In a concert setting the addition of a piano for concerto use will require
that the center-front orchestral players move upstage as needed so that the
piano may be placed between the conductor and the house, as shown in
Figure 13–15A. If an audience is present, this repositioning often interferes with the placement of the main microphone pair. Ideally, it should
be just in front of the piano, but considerations of sightlines may force the
engineer to place the microphone stand just behind the instrument. With
hanging microphones there is no problem here at all; the main pair can
be placed in its ideal position. (See Compact Disc Reference 9.)
Vocal or string soloists are often placed to the conductor’s left and
will not interfere with the main microphone setup. Whenever possible,
these soloists should be picked up with a coincident or near-coincident
stereo pair, as opposed to a single microphone. The intent here is to preserve natural stereo perspectives in case the engineer has to increase the
level of the soloist’s microphones during very soft passages. Details are
shown in Figure 13–15B. (See Compact Disc Reference 10.)
A hall that is too live can be easily deadened by placing velour material
at strategic points in the performance space. Where possible, the velour
should be hung over balcony railings or freely hung away from the walls
as banners. A little will often go a long way, and it is highly recommended that an acoustical consultant be engaged to oversee this temporary modification.
It is a bit more complicated to liven a space, but the technique detailed
in Figure 13–16 produces surprising results. Before and after examples are
2:53 PM
Page 211
13: Classical Stereo Recording Techniques and Practice
FIGURE 13–15
Recording soloists with
orchestra. Piano (A);
vocalist (B).
given in Compact Disc Reference 11. The effect of the 0.004 in. (0.1 mm)
thick vinyl material is to reflect high frequencies arriving at a fairly low
grazing angle of incidence. It does not have to be stretched in any manner;
rather, simply laid over the seating area. (The vinyl material in question is
available at building supply stores under a number of trade names. No
material less than 0.004 in. thickness should be used.)
As we have mentioned before, artificial reverberation on accent
microphones is an accepted practice in current classical recording; however, there is no substitute for natural reverberation to enhance the primary pickup via the main microphones and house microphones.
Music of different periods will require varying amounts of reverberation,
both in level and in decay time. General recommendations for concert
venues, as a function of room volume and music type, are given in
2:53 PM
Page 212
FIGURE 13–16
Livening the recording
venue: section view (A);
measured reverberation
time with and without
plastic material (B).
Figure 13–17. In general, reverberation times exceeding about 2.5 seconds
will sound rather unnatural on orchestral musical, whatever the period.
Both classical and modern era music fare best with reverberation times on
the order of 1.5 s, while romantic compositions may do with reverberation
times up to 2 s. While ecclesiastical music is often heard in large spaces
with up to 4 or 5 s of reverberation time, in recording it is probably better
to err on the short side in order to keep the reverberation from interfering
with musical details.
Kuhl (1954) developed the data shown in Figure 13–18. Here, monophonic recordings were played for a listening panel whose individual preferences for reverberation time are plotted. The median points are taken
as target values. We need to add that these tests, if repeated today in
stereo, could result in slightly longer target reverberation time values, due
to the ability of stereo listening to delineate direct-to-reverberant details
more accurately than in mono listening. In any event, the tendencies
shown here are certainly relevant in today’s recording activities.
2:53 PM
Page 213
13: Classical Stereo Recording Techniques and Practice
FIGURE 13–17
Typical reverberation times
versus room volume for
various types of music (A);
normal variation of LF and
HF reverberation times
compared with the
midband (B).
Once the target reverberation time has been established, the amount
of reverberation introduced into the recording is critical. The sound
picked up by the house microphones and introduced into the stereo mix
will determine this ratio. This is an area where accurate loudspeaker
monitoring conditions and experience on the part of both engineer and
producer are of primary importance. In many cases, this important
judgement is not made until the postproduction stage.
It is up to the engineer and producer to determine the effective “distance”
of the pickup, relative to the stage, and make it convincing. Rather than
placing microphones at Row J, we make the recording fairly close-in, and
then we introduce reverberation to “zoom out” to where we want it to
be. A slight increase in the amount of signal from the house microphones
will result in a surprising increase in effective distance, so care should be
taken in making these balances. When musical reasons justify this, it may
be desirable to pan in the main flanking microphones very slightly toward
the center, just to give the direct stage pickup a slightly narrower image to
match the increase in apparent fore–aft distance. If this is not done, we
run the risk of making a recording with conflicting spatial cues; it may
sound both close-in and distant at the same time.
2:53 PM
Page 214
FIGURE 13–18
Kuhl’s data on preferred
recorded reverberation
times for three orchestral
As we have seen in previous chapters, today’s quality capacitor microphones may have dynamic ranges well in excess of 125 dB. However, in
classical recording, we are rarely able to take advantage of more than
about 90 or 95 of those precious decibels, as illustrated in Figure 13–19.
As we can see, the noise floor of a modern quiet venue is just about
at the same effective level as the self-noise floor of the primary microphones themselves. As the amplified signals from the microphones are
transmitted downstream to a digital recorder, the noise floor limitation
in the recording remains that of the microphones and the room.
Thus, the choices of both venue and microphones remain all important. Microphones with self-noise in the range of 7 to 13 dB(A) are
strongly recommended for all classical orchestral recording.
2:53 PM
Page 215
13: Classical Stereo Recording Techniques and Practice
FIGURE 13–19
Dynamic range of
recording microphones in a
concert venue.
1. Singing on the Water, Delos CD DE 3172, Carol Rosenberger,
piano. Recorded with two spaced omnidirectional microphones.
2. Bach: Violin Sonatas and Partitas, Arranged for Guitar, Delos CD
DE 3232, Paul Galbraith, guitar. Recorded with two spaced omnidirectional microphones.
3. Things Visible and Invisible, Delos CD DE 3147, Catharine
Crozier, organ. Recorded with two omnidirectional microphones
spaced about 4 m at a distance of 10 m.
4. Arensky/Tchaikowsky Piano Trios, Delos CD DE 3056, Golabek,
Solow, Cardenes Trio. Recorded as detailed in Figure 14–8B.
5. Love Songs, Delos CD DE 3029, Arleen Augér, soprano. Recording
made as shown in Figure 14–9. Some room reverberation added.
6. Mendelssohn/Grieg String Quartets, Delos CD DE 3153, Shanghai
Quartet. Recorded as shown in Figure 14–10C.
2:53 PM
Page 216
7. Haydn, Symphonies 51 and 100, Delos CD DE 3064, Gerard
Schwarz conducting the Scottish Chamber Orchestra. Recorded
as shown in Figure 14–12.
8. Holst, The Planets (coupled with Strauss: Also Sprach
Zarathustra), Delos CD DE 3225, Andrew Litton conducting the
Dallas Symphony Orchestra. Recorded as detailed in
Figure 14–13.
9. Shostakovich, Piano Concerto no. 2, Delos CD DE 3246, Andrew
Litton, pianist and conductor, Dallas Symphony Orchestra.
Recorded as shown in Figure 14–15A, but with main ORTF piano
between piano and audience.
10. Mahler, Symphony no. 2, Delos CD DE 3237, Andrew Litton,
Dallas Symphony Orchestra and Chorus. Recorded as shown in
Figure 14–15B, with two vocal soloists on either side of the conductor and with chorus located in choral terrace behind and at
sides of the orchestra.
11. Second Stage, Delos CD DE 3504, various orchestral movements
to demonstrate recording perspectives. Band 2 recorded in
Portland, OR, concert hall without livening; band 12 recorded in
same hall with livening technique as discussed in Figure 13–16.
2:54 PM
Page 217
Most recordings of commercial music originate not in traditional
performance venues but in professional recording studios. Most studios
used for pop and rock recording are large enough to accommodate no
more than 12 to 20 musicians comfortably, and only in large metropolitan centers will we find large sound stage environments that can accommodate a 50- to 75-piece orchestra. For the most part, microphone
placement in these studios is much closer to the instruments than is usual
in classical recording, and there is much more usage of accent microphones. In virtually all cases there will be extensive use of multitrack
recording, which allows for greater flexibility in postproduction.
In this chapter we will discuss a wide range of recording activities
and supporting functions that routinely take place in the studio. We will
also discuss aspects of studio acoustical treatment and arrangements for
instrumental isolation, both individually and in groups.
We will also develop in this chapter the notion of recording as an art
in its own right and not as a simple documentation of an acoustical
event. Many of our recommendations in balance and panning may seem
to be arbitrary, but their effectiveness in two-channel stereo playback
justifies them.
The modern studio should have a modern control room, fitted with
monitor loudspeakers that are symmetrically arrayed around the primary
workspace. They may be built in, or they may be placed on stands. It is
important that they exhibit essentially flat response over the frequency
range out to about 6 or 8 kHz, with no more than a slight rolloff above
that frequency. The response in both channels should be within about
2 dB of each other from about 100 Hz to about 8 kHz, and the systems
should be able, on a per-channel basis, to reproduce cleanly midband
signals of 105 dB at the engineer’s operating position.
2:54 PM
Page 218
The drum set is a vital element in virtually all popular and rock music
groups. While its actual makeup may vary from one player to another,
the layout shown in Figure 14–1 is typical. Its basic elements are:
1. Kick drum: played with a spring-loaded beating attachment operated by the right foot, striking the head of the drum (the head is the
stretched playing surface).
2. Snare drum: played by sticks or light metal brushes with both hands.
The snares (tight gut strings across the rear head) are normally
engaged and give the struck instrument its characteristic “snap”.
3. Hi-hat cymbals: consists of one fixed cymbal and a movable one
which is articulated by the player’s left foot. The pair of cymbals
are struck with a single stick in the right hand, and the left hand is
used to position them, varying the timbre.
4. Overhead cymbals: there are usually three of these, freely struck by
single sticks. The cymbals are often known by the following names:
“ride” cymbal, “crash” cymbal, and “sizzle” cymbal. The lastnamed one has multiple small pins located loosely in holes around
the periphery of the cymbal. These vibrate freely, metal-againstmetal when the cymbal is struck, producing a sound rich in high
5. Tom-toms: Commonly known as “toms”, these are small drums
struck with sticks.
Front view of the modern
drum set.
2:54 PM
Page 219
14: Studio Recording Techniques
Anyone who has ever witnessed a studio recording will recall the
drummer’s long setup time and the many adjustments made to the drum
set before playing gets under way. Spurious resonances and ringing must
be damped out, and the player will make absolutely sure that the various
moving mechanisms are silent in their operation.
Every instrument in the drum set will be precisely positioned, and
the microphones, no matter how closely you position them, must not
interfere with the player’s movements in any way.
Many small acoustical jazz groups require only basic pickup of the drum
set, consisting of an overhead stereo pair and a single kick drum microphone. The overhead pair will normally be cardioid capacitors with flat,
extended response. Place these microphones as shown in Figure 14–2,
with their axes aimed at the hi-hat cymbals and right tom, respectively.
Both microphones should be placed slightly higher than the player’s head
Simple pickup of the
drum set.
2:54 PM
Page 220
and slightly to the rear so that they will not distract the player. The general
aim here is to pick up the sound of the drum set very much as it is heard
by the player. Some engineers prefer to use a near-coincident spacing of
the overhead microphones; the wider spacing shown here will give a
slightly broader stereo presentation and is recommended for that reason.
The kick drum microphone is normally placed close to the instrument
on the far side from the player. Microphone choice varies; many engineers
use a dynamic microphone with known low frequency headroom reserve,
while others prefer a capacitor microphone. The differences here are
purely a matter of taste, but in either event make sure that the microphone can handle the loudest signals the player will produce.
The drum tracks are normally recorded dry (without any signal processing in the chain), with each microphone on a separate track. In stereo
presentation, the kick drum is normally panned into the center stereo
position, while the overheads are panned left and right, as determined by
musical requirements and spatial considerations.
Watch recorded levels carefully; the acoustical levels generated
close to the drum set can easily reach the 135 dB LP range during loud
playing, so make sure that your microphones can handle these levels
Where recorded track capability is not a problem and where the music
demands it, you can pick up individual elements within the drum set as
shown in Figure 14–3. The following rules should be observed:
1. Toms and the snare drum: You can use small electret capacitor clipon mikes, as shown in Figure 14–4, or you can place small diameter
capacitor microphones slightly above and behind the drum.
Whatever your choice, make sure that the microphone is away from
any position or path that the drumstick is likely to pass through
or near.
2. Cymbals and the hi-hat: These instruments will swing about their
central pivot points when struck, and microphones must be placed
carefully within the drum set away from the player and with
enough clearance so that they will not interfere with playing. The
normal position of the microphone should not be in line with the
edge of the cymbal but rather slightly above the cymbal. The reason for this recommendation is that the radiation pattern of the
cymbal changes rapidly along its edge with small angular changes.
Details here are shown in Figure 14–5.
When so many microphones are used on a single set of instruments it
is important to determine their real value in creating the final musical
product. Needless to say, all of the microphones should be of cardioid or
hypercardioid pattern in order to keep unnecessary leakage to a minimum.
2:54 PM
Page 221
14: Studio Recording Techniques
More detailed pickup of
the drum set. Front view
(A); top view (B).
If recorded channel capacity is an issue, there is always the possibility that
some degree of stereo sub-grouping of multiple tracks will save on the total
channel count. The engineer and producer should always think ahead and
have in mind, as early as the initial tracking sessions, what the downstream postproduction requirements might be.
Detail of a clip-on
microphone for drum
pickup. (Photo courtesy
of AKG Acoustics.)
Drums are very efficient instruments; the player does not have to expend
much physical power in order to produce very high sound levels in the
studio. Almost routinely, the drum kit and player are isolated from the
other musicians to reduce leakage from the drum set into other open microphones in the studio. Details of a drum cage are shown in Figure 14–6.
An effective cage can reduce the mid-band ambient level from the drum
set by up to 15 dB in the studio. It is important not to place the drummer in an isolation booth per se. Eye contact is essential, as is direct
speaking contact with other players. In spite of all the steps necessary to
minimize drum leakage, it is important that the drummer feel a part of
the playing group.
2:54 PM
Page 222
Microphone placement
for cymbals.
Details of a drum
cage: plan view (A); front
view (B).
Further use of baffles (also known as “goboes”) in the studio can
provide more acoustical separation among players as needed. The effectiveness of a single large baffle placed in a fairly well damped studio is
shown in Figure 14–7. Here, you can see that at mid and high frequencies the isolation is in the range of 15 to 22 dB.
During tracking sessions, it is customary to provide all players with
headphones so that “everyone can hear everyone else.” Generally, each
player can have a separate headphone mix, and in some large studios an
assistant engineer, using a monitor console located in the studio, provides
monitor mixes for all players.
2:54 PM
Page 223
14: Studio Recording Techniques
Effectiveness of baffles:
top view (A); side view
(B); isolation versus
frequency (C).
There are more percussion instruments than we can hope to detail here;
however, they all fall into several basic groups, according to the principle
of sound generation:
The non-pitched percussion instruments include:
Metallophones (metal resonators):
Triangle (struck with a small metallic beater)
Gongs and tam-tams (struck with soft beaters)
Finger cymbals (struck together)
Bell tree (graduated set of nested bells, usually struck in succession
with metal beater)
Cocolo (small metal chains loosely positioned around a metal
core that is rotated by the player)
Xylophones (non-metallic resonators):
Maracas (dried gourds filled with seeds and shaken)
Castanets (small wood-against-wood pieces shaken together)
Claves (hard wood sticks struck together)
Guiro (dried gourd with serrations; played with scraper)
Membranophones (instruments with stretched diaphragms):
Bongo drums (set of Latin drums roughly tuned high to low)
2:54 PM
Page 224
Tambourine (combination of metallophone and membranophone)
Other ethnic drums
The pitched percussion instruments include:
Orchestra bells (set of metal bars in keyboard layout)
Chimes (suspended tubular bells in keyboard layout)
Celesta (tuned metal bars played with keyboard)
Vibraphone (tuned metal bars played with mallets)
Xylophones (generic):
Xylophone (short tuned wooden bars in keyboard layout)
Marimba (tuned wooden bars with resonators in keyboard
The smaller percussion instruments are normally played by two or
three musicians, who move freely from instrument to instrument as
called for in the musical score. It is best to record them in stereo with
good left-to-right spread, assigning them to one or more stereo pairs of
recorded tracks. For example, a Latin work with might call for a significant marimba part to be recorded in stereo, while another stereo pair of
tracks will be assigned to a group of non-tuned percussion instruments.
Normally the microphone would be about 0.5–1 m (20–40 in) away
from the instruments. Recording the vibraphone in stereo is shown in
Figure 14–8. The same setup applies to the marimba and other tuned
mallet instruments.
A widely splayed pair of
cardioids for recording the
vibraphone in stereo.
2:54 PM
Page 225
14: Studio Recording Techniques
The piano is invariably picked up in stereo using two or three microphones.
Typical placement for pop/rock recording is shown in Figure 14–9.
In a jazz context the microphones would normally be placed slightly
outside the rim of the instrument, giving just a bit more distance to the
When picked up at such short distances the piano may sound quite
bright, and the hammers may have to be softened by a piano technician.
Any slight mechanical problem may become very apparent with close
microphone placement, and the technician should be prepared to fix such
problems. Keep in mind that any changes made to the hammers (hardening or softening) cannot always be undone easily or quickly. The technician and player should be in agreement on any changes made to the
Most engineers favor cardioid microphones for close placement,
since the proximity LF rise will usually enhance the sound of instrument.
Microphones placed outside the rim may be omnidirectional or cardioid,
at the choice of the engineer or producer. Some engineers prefer to use
boundary layer microphones, fastening them to the underside of the
piano cover using double-sided tape. This minimizes the effect of reflections inside the instrument.
As an aid to minimizing studio leakage into the piano microphones,
many engineers position the piano so that its cover, when open, faces
away from the other players, thus shielding the instrument from direct
sounds in the studio. If more isolation is required, the piano cover may
be positioned on “half-stick” and a blanket placed over the opening, as
shown in Figure 14–10.
Recording the piano: top
view (A); front view (B).
2:54 PM
Page 226
While the basic piano tracks should be recorded dry, reverberation
will always be added in postproduction. The presentation on the stereo
soundstage is normally in stereo; however, if the mix is a hard-driving,
complex one, the piano may sound clearer if mixed to mono and positioned where desired.
There are several methods for recording the bass:
FIGURE 14–10
Microphone on a floor stand
Nesting a microphone behind the tailpiece of the instrument
Microphone at bass amplifier/loudspeaker unit
Direct-output from instrument pickup into the console
Recording the piano with
added isolation.
Using a floor stand, as shown in Figure 14–11, a cardioid microphone
can be precisely placed to pick up the desired combination of fundamental sound along with the sounds of the player’s fingers on the strings. For
acoustical jazz it is important to achieve this, and we recommend this
approach as the best overall choice. It may be helpful to place a set of 1 m
(40 in) high baffles behind the microphone to reduce leakage from other
Some engineers prefer to place a small format omnidirectional
microphone in the space between the tailpiece and the body of the instrument. It is usually wedged in place using a piece of foam rubber, as
shown in Figure 14–12. The advantage as here is that any movements of
the instrument will not alter the microphone-to-instrument distance; a
disadvantage is that the pickup of finger articulation will be minimized.
FIGURE 14–11
Recording the bass
with a stand-mounted
2:55 PM
Page 227
14: Studio Recording Techniques
FIGURE 14–12
Recording the bass with a
microphone mounted on
the instrument.
FIGURE 14–13
Picking up the bass
acoustically from the bass
Under some adverse conditions the engineer will have no other
recourse than to place a microphone directly in front of the loudspeaker
in the bass amplifier/loudspeaker unit, as shown in Figure 14–13. Many
times, a quick stage setup or change-over will not provide the time necessary to work out the details of a better pickup method. Many bass
amplifiers are noisy and may be distorted at high output levels, both of
which can cause problems in postproduction. However, if the amp is
clean and noise-free, you can get a good recording this way.
A direct console feed from the instrument’s pickup is an excellent
way to achieve a good bass sound with no leakage whatever. Be sure that
you have a good active splitter box that will allow you to sample the signal directly out of the bass, while sending an uninterrupted signal on to
the bass amplifier. Details are shown in Figure 14–14A and B. Most
direct output feeds are unbalanced (see Chapter 8 under Microphone
Splitters) and should be fed into a console input set for line input operation. Phantom powering should be turned off.
Given the choice, most engineers will want to lay down at least two
bass tracks, one direct and the other via a microphone pickup. The two
will often be combined in postproduction. (Be on the lookout for antiphase cancellation when the signals are combined later on.)
2:55 PM
Page 228
FIGURE 14–14
Recording the bass: using
the direct output from the
bass (A); a pickup mounted
on the instrument’s bridge
(B). (Photo courtesy of
AKG Acoustics.)
Bass tracks are invariably compressed and equalized during postproduction. While much in the way of equalization can be done at that time,
it is always best to get the ideal sound on the basic tracks in the first
place. There are a number of excellent dynamic microphones for bass
pickup; these are often models that have been designed with a slight bass
rise and with adequate headroom at low frequencies. Many engineers
prefer to use large diameter Braunmühl-Weber capacitor types, since
these models often have distinctive high frequency signatures which both
engineer and performer may deem desirable.
In performance, most vocalists make use of handheld “vocal microphones”
whose response characteristics are internally equalized to produce a fairly
smooth response at close operating distances. In the studio, the vocalist is
normally recorded via a Braunmühl-Weber dual diaphragm large format
capacitor model. Years of acclimation have convinced most vocalists as
well as engineers that this is the only way to go.
There is much to be said in favor of this microphone type: the models are all slightly different in response, and the engineer can usually pick
one that truly enhances the performer’s voice at high frequencies. As a
group they are robust and can take high vocal levels in stride. Many
older tube models dating from the early 1960s are highly regarded and
are routinely used for these purposes. (See Chapter 21 for a discussion of
older classic microphone models.)
The normal studio setup is as shown in Figure 14–15. The microphone, set in cardioid position, is placed in front and above the vocalist,
with room for a music stand. The operating distance is approximately
0.6 m (24 in). A sheer pop screen is normally positioned in front of the
microphone as shown in order to control inadvertent pops of wind from
the vocalist. Be sure to angle the music stand so that there are no direct
reflections from the vocalist to the microphone. It is customary to provide
a stool for the vocalist, whether or not it is actually used. In a tracking
2:55 PM
Page 229
14: Studio Recording Techniques
FIGURE 14–15
Recording a vocalist: side
view (A); top view (B).
session, the vocalist will normally be surrounded by baffles on three sides,
and in extreme cases of studio leakage it may be necessary for the vocalist to perform in a vocal booth. This is a rather confining environment
and should be chosen as last resort.
Wherever the vocalist is located, it is important that headphone
monitoring be carefully tailored to the vocalist’s tastes. Typically, the
vocalist will want to monitor a stereo mix with a generous amount of the
vocal track itself, complete with stereo reverberation. In the unlikely
event that you are using a compressor on the vocal track going to tape,
do not put that signal in the vocalist’s monitor mix.
There are so many important details here that it is strongly recommended that you cover as many points in setup as you can before the
vocalist arrives for the session. You can always find someone on the studio
staff who will be glad to help you set up. The psychology of the moment
is critical; nothing is more comforting to a vocalist than hearing a truly
clean and clear sound the first moment the headphones are put on.
2:55 PM
Page 230
On many occasions the vocalist will want to record basic vocal tracks
after the initial instrumental tracking sessions have been finished. In this
case there will be few, if any, performers in the studio, and the isolation
measures we have discussed will probably not be necessary.
Be prepared to do vocal insert recording efficiently and smoothly, if
it is required. Make sure you have an experienced tape operator who can
keep accurate notes on the take sheets.
As vocal tracking gets underway the singer will want to come into
the control room to listen to the tracks that have just been recorded.
Make sure that you have a reasonable monitor mix up and running for
the singer and other artists as soon as they get to the control room.
While many of the points we are making here have nothing to do
with microphone technique as such, they are very important to the
smooth running of any pop recording session – and directly reflect on
your general competence and microphone choices.
You can use as many recording channels as you have to spare, but a
minimum of two may suffice, as shown in Figure 14–16. Since more than
one singer will be on each microphone, do not hesitate to reposition the
singers slightly in order to correct balances. Experienced backup singers
are always aware of this and will gladly follow your requests. If you find
it necessary to compress the stereo signals, make sure that you use two
compressors that are stereo-coupled in order to maintain overall balances.
A backup chorus, if needed, can best be recorded using the classical
techniques discussed in Chapter 13.
The acoustic guitar is equally at home in classical as well as popular
music of all kinds. While intimately associated with Spain, it has become
FIGURE 14–16
Recording a backup vocal
group in stereo, top view;
circles show approximate
positions of singers.
2:55 PM
Page 231
14: Studio Recording Techniques
FIGURE 14–17
Recording the acoustical
guitar with microphones
and direct input to the
a truly universal instrument. For most of its existence it survived only as
an acoustic instrument; however, during the last five decades it has
emerged in a dual role in both acoustical and amplified form. For rock
applications, the solid body guitar has become the chief instrumental
musical element. With its solid body structure, the only resonance present in the instrument lies in the strings, and the sound we hear is that of
the strings alone suitably amplified and processed.
In the studio, the acoustical guitar is normally recorded both with
microphones and via direct electrical output; in many cases a final mix will
make use of both inputs. In this section we will consider both options.
Figure 14–17 shows a conventional stereo method for recording the
guitar acoustically. Most engineers prefer spaced microphones as shown
here, but a coincident pair of cardioids is also an option.
When using spaced microphones, if the guitar is to appear in mono
in subsequent postproduction, it is best if only one of the recorded channels is used, suitably equalizing it for the desired spectral balance.
Alternatively, a coincident pair can be mixed directly to mono with no
phase cancellation problems.
There is much added flexibility in postproduction if a third track is
used to record the direct output from the guitar. The timbral difference
between the direct and microphone signals will be considerable, and this
can be used to create a fairly widely spaced stereo presentation.
Modern guitar amplifiers have stereo outputs, and when recording
the solid body instrument it would be a mistake not to record both outputs for the maximum postproduction flexibility.
Like the solid body guitar, the synthesizer (“synth”, as it is called) can
only be recorded directly from a stereo pair of outputs. In many cases
synth players will want to lay down additional tracks by overdubbing,
so manage your track capability accordingly.
2:55 PM
Page 232
As noted in Chapter 13, woodwind instruments have fairly complicated
radiation patterns, and this dictates that microphones should not be
placed too close to them if a balanced sound is desired. Often on television performances we often see small microphones clipped to the bells of
clarinets, primarily as a matter of expedience. If Figure 13-2 is any indicator, the resulting sound is likely to be very bass heavy, and considerable equalization will be needed to create a good timbre. Brass
instruments will fare somewhat better with such close pickup, since all
radiation is by way of the bell. Figures 14–18 and 14–19, respectively,
show how individual woodwind and brass instruments may be recorded
close-in for best balance. The dimensions shown in Figures 14–18 and
14–19 represent minimum values; you can always place the microphone
farther away, consistent with leakage problems in the studio. See Meyer
(1978), Dickreiter (1989) and Eargle (1995) for additional information
on the directional characteristics of musical instruments.
The French horn is a special case. It is normally heard via room
reflections, since the bell always fires to the rear of the player. If more
control is needed it is best to place a microphone oriented overhead at
90 to the bell axis. In that position the microphone will pick up some of
the “buzz” that characterizes the raw sound of the instrument, giving the
recorded sound additional presence. Whenever possible, a set of reflective baffles should be placed behind the instruments.
Data given in Figure 13–3 show the complex nature of radiation from
bowed string instruments, indicating how difficult it is to obtain a natural
FIGURE 14–18
Recording woodwind
instruments: flute
(A); oboe/clarinet (B);
saxophone (C).
2:55 PM
Page 233
14: Studio Recording Techniques
FIGURE 14–19
Recording brass
instruments: trumpet (A);
trombone (B); French
horn (C).
balance at close operating distances. Nevertheless, clip-on microphones
are often used on-stage when string instruments are used in combination
with much louder instruments. Since the bowed instruments are normally
used in multiples, we will discuss their pickup below under The Large
Studio Orchestra.
The harp can be recorded as shown in Figure 14–20. A spaced stereo
pair of omnidirectional microphones will produce the best timbre, but
in a crowded studio it may be mandatory to use cardioids for added
FIGURE 14–20
Recording the harp in
Our discussion here will be limited to three jazz ensembles ranging from
small to large and a single large studio orchestra, such as might be used
for film scoring.
Basic layout and panning
Figure 14–21A shows the studio layout for a basic jazz instrumental trio
consisting of piano, drums and bass. The simplicity of this ensemble
allows us to explore the basic advantages of stereo recording of both the
piano and drum set.
2:55 PM
Page 234
FIGURE 14–21
Recording a jazz trio:
studio layout (A); target
stereo soundstage (B).
Whenever possible, the engineer and producer should place the
musicians in the studio with the same left-to-right positioning as
intended in the final postproduction mix. The reason here is that any
studio leakage will be between adjacent instruments as they appear on
the stereo soundstage and thus not create any conflicting spatial cues.
Studio leakage is not necessarily a bad thing, and it can vitally enhance
a performance rich in acoustical cues when the instruments are so tightly
clustered together.
The piano has been placed at the left, since the open cover of the
instrument will reflect sound to both the bassist and drummer. The
bassist is placed in the center with both microphone and direct pickup.
The drum set is placed at the right and uses an overhead stereo pair with
additional microphones on both the snare and kick drums. The primary
purpose of the snare microphone is to pick up the delicate playing of
brushes on the snare drum.
Establishing the monitor mix
Both producer and engineer must be in mutual agreement on the stereo
layout of the recording. Here, the aim is to have the piano be reproduced
from the left half of the stereo stage, and this calls for panning the treble
piano microphone at the left and the bass piano microphone at the center. Because of the considerable leakage between the signals at both
2:55 PM
Page 235
14: Studio Recording Techniques
microphones the sound will appear from the left half of the stereo stage,
with high frequencies tending toward the left and middle and low frequencies tending toward the center.
The bass microphone and direct pickup will both be panned to the
center in whatever proportion best fits the music. The balance between
the two can in fact vary throughout the recording from tune to tune.
The kick drum microphone will be panned to the center, and the
drum overheads will be panned center and right. The microphone on the
snare will be subtly mixed in and panned between center and right.
In a typical studio environment, the piano and bass inputs will be fed
to the stereo inputs of an external reverberation generator whose stereo
returns will be panned to the left and right signal buses. The choice of
whether to use any artificial reverberation on the drum microphones is
open for consideration. Normally none is used.
The reverberation parameters for this recording would normally be
set for a LF (below 500 Hz) reverberation time of about 1 second, with
reverberation time above that frequency slightly longer.
The target stereo soundstage
The representation at Figure 14–21B gives a picture of the resulting
soundstage as intended by the producer and engineer. Note that the
stereo presentation is wide, yet preserves good center-stage imaging of
the bass as well as portions of the piano and drum set. The reverberation
return signals should be panned full left and right in order to simulate
the widest spatial separation. For a group as small as this, and with
instruments that are equally balanced acoustically, there may be virtually
no need for headphone monitoring.
Basic layout and panning
Figure 14–22A shows the basic studio setup and panning assignments for
a jazz group consisting of vocalist, two solo saxophones and a full
rhythm section. As before, the studio left-to-right setup is in keeping
with the anticipated sound stage.
The piano has been replaced by a Hammond electronic organ, a
mainstay in blues-type jazz vocals. The organ is recorded in stereo with
microphones next to its HF and LF loudspeakers. The Hammond B-3
model is never recorded direct-in because of the extensive pre-processing
of the signal and the fact that it normally uses the Leslie loudspeaker system, with its characteristic HF rotating loudspeaker array. The acoustical guitar is picked up with both microphone and direct output. Both
organ and guitar are arrayed in stereo, respectively on the left and right,
enhancing the musical dialog that often takes place between them.
The two saxophones are placed left and right. When they are playing in counterpoint or dialog, their left and right panned positions are
2:55 PM
Page 236
FIGURE 14–22
Recording jazz vocal with
small instrumental group:
studio setup (A); target
stereo soundstage (B).
maintained; when either instrument is playing a solo chorus it may be
panned to the center of the stereo stage. The reason here is purely tradition and expectation. Bass and drums are positioned as in the previous
Establishing the monitor mix
The first step is to establish a basic balance with bass, drums and organ;
listen for a smooth but detailed soundstage from left to right. Then the
guitar can be introduced into the mix. As a final step, balances among
the vocalist and saxophone soloists are made. Reverberation sends
should be taken from all instruments, except possibly the drums, and
returned to the monitor mix with full left and right panning.
The target stereo soundstage
The representation in Figure 14–22B shows the stereo soundstage as
intended by the producer and engineer. There are three basic spatial
layers in the recording: front-most are the three essential soloists. The
2:55 PM
Page 237
14: Studio Recording Techniques
secondary layer includes the rhythm elements, and the reverberation
return constitutes the back layer.
At times, the rhythm elements will take on primary importance, and
when the do they should be boosted in level according to their importance. A basic skill of any recording engineer is knowing when – and
how much – to alter the level of an instrument that moves from one layer
to another. An experienced producer is of great help here.
Always bear in mind that not all elements in the mix can have equal
importance at the same time. Both spatial and frequency dimensions must
be considered; whenever possible, important elements in the mix should
be perceived as coming from separate directions. Likewise, the mix, as
perceived overall, should exhibit a uniform spectrum from top to bottom.
Basic layout and panning
The standard large jazz band normally consists of four trombones, four
trumpets, and five saxophones in the brass and wind departments.
Rhythm elements consist of drums, bass, piano and guitar. Brass and
wind players may also double, respectively, on French horn, clarinet or
flute. There may be an additional percussion player as well. Figure 14–23A
shows a typical studio layout. The trumpets are usually placed on a riser
FIGURE 14–23
Recording a jazz big band:
studio layout (A); target
stereo soundstage (B).
2:55 PM
Page 238
behind the trombones, and it is not unusual for the various wind sections
of the band to stand when playing as an ensemble. Stereo pickup of both
saxophones and brass is typical, with accent microphones in place for
solos. The engineer needs accurate cueing from the producer for spotting
the solos. Usually, a single stereo reverberation system will adequately
take care of all ensemble demands for ambience. However, a vocal
soloist would certainly call for an additional reverberation system, most
likely with shorter reverberation time settings.
Target stereo soundstage
Both saxophones and brass are panned across the stereo soundstage to
form a continuous presentation. The spotting of solos, depending on
how extended they may be, are panned into the center. The rhythm elements form another layer, arrayed in stereo with piano at the left, drums
in the center, and guitar at the right. The overall soundstage is as shown
in Figure 14–23B.
In order to avoid a cluttered setup, most engineers would opt for
stereo microphones for the basic stereo pickup of the saxophones and
the brass. Pickup of the rhythm elements would be in accordance with
the suggestions made earlier in this chapter. The safest approach to
recording a group this complex would be to assign each microphone to
its own recording track, while at the same time making a two-track
stereo monitor mix. There are a number of engineers who are truly
adept at making “live to two-track” mixes of big bands. An automated
console, with adequate subgrouping of inputs, makes the job a bit
As discussed here, the large studio orchestra is typical of the ensemble
that would be used in motion picture sound track scoring. Superficially,
the group may resemble a small symphony orchestra. The main pickup
will in fact look a great deal like those discussed in Chapter 13.
The engineer must determine, in consultation with the composer/
arranger and the recording producer, the requirements of the score. Any
instrument that will be featured in a solo role, however momentary, will
probably need an accent microphone. The reason for this is that postproduction mixing of music tracks behind dialog and/or effects will ordinarily require that individual solo instruments be highlighted in order to
be heard effectively in the final film mix. This is a judgment that cannot
readily be made during scoring sessions.
However, the tracks laid down for opening and closing credits may
well be used exactly as balanced by the mixing engineer during the scoring sessions. The only new requirement introduced here is that the mix
will, in today’s creative environment, be done in five-channel format,
three in front plus two surround channels, for motion picture purposes.
2:55 PM
Page 239
14: Studio Recording Techniques
FIGURE 14–24
Studio layout for a large
studio orchestra.
The string ensemble, as in classical recording, always sounds best
when the microphones are no closer than absolutely necessary to delineate them with respect to the winds and brasses. Many a great score has
been ruined by too-close string microphone placement, with its characteristic “screech.” Figure 14–24 shows a typical studio layout for a large
scoring session. Note that the strings are picked up primarily via the
main four microphones across the front and the internal overhead pair
on the inner strings. The accent microphones are used only for highlighting solos, and they are often implemented only during postproduction sessions.
Half a century ago, recording studios tended to be fairly dry acoustical
environments in order to meet needs growing out of broadcasting and
film sound recording, with their limited dynamic ranges. Over the
decades we have seen more varied spaces as the recording arts have
evolved. Today’s recording studio is apt to have a number of environments within a larger space. For example, one end of the studio may be
more acoustically live than the other; or perhaps the entire studio will be
covered with movable wall sections that can be turned inward or out to
create local live or damped playing areas, as shown in Figure 14–25.
2:55 PM
Page 240
FIGURE 14–25
Live and absorptive areas in
a modern studio.
FIGURE 14–26
Reflections in the studio:
microphone positions (A);
response (B).
Modern studios will have several isolation areas. These are used primarily for vocals and any soft instrumental elements in the mix. Some
studios have a large space, virtually a second studio, adjacent to the main
area and large enough to contain a sizeable string group. Large sliding
glass doors provide the necessary sight lines into the main studio.
Discrete reflections in the studio can be problematic, as shown in
Figure 14–26A. For example, in recording cellos and basses, the microphone may be fairly close to the floor. If the microphone is about
2:55 PM
Page 241
14: Studio Recording Techniques
FIGURE 14–27
A boundary layer
microphone. (Photo
courtesy of Crown
FIGURE 14–28
The “three-to-one” rule in
omnidirectional microphone
placement: both sources
about equal in level
(A); source 1 louder
than source 2 (B).
equidistant from the sound source and the floor, as shown at position 1,
the reflected path can create an irregularity in response as shown at B.
Progressively moving the microphone toward the floor will minimize the
response irregularities.
In most cases it is best to place the microphone directly on the floor
and avoid interferences completely. Boundary layer microphones are
specifically designed for this application. A typical model is shown in
Figure 14–27. Normally, an omnidirectional pattern is chosen, but directional boundary layer microphones can be used if more separation is
Another problem arises in smaller studios where players are often seated
closer together than is desirable. Figure 14–28A shows what is called the
“three-to-one” rule. That rule states that, when using omni microphones, the distance from a microphone to its target sound source
2:55 PM
Page 242
should be no greater than one-third the distance from that microphone
to the nearest interfering source. The three-to-one ratio will reduce the
level from the interfering source by an average of 10 dB, which is quite
acceptable. When one instrument is significantly louder than the other,
then the adjustment shown at B must be made. Obviously, using cardioid
microphones will diminish interference problems all the more.
2:56 PM
Page 243
While two-channel stereo has been the mainstay of consumer sound
reproduction for nearly a half-century, surround sound as originally
developed for motion pictures has technical roots that go back even
earlier. For the consumer, surround sound as a central requirement in
music-only presentation in the home has had a spotty history.
Quadraphonic (four-channel) sound was introduced in the mid-1970s
and failed, chiefly because the technology proposed for it was not sufficiently developed. During the mid-1990s, surround sound was reintroduced to the consumer as an integral part of the home theater revolution,
with its five-channel loudspeaker array consisting of 3 loudspeakers in
front and 2 at the sides slightly to the rear of the listener. The basic plan
was patterned after the normal loudspeaker setup in motion picture
theaters of the day. A primary performance benefit of the “new” videobased surround sound was the use of a front center channel, which
anchored center-stage events accurately in that position – regardless of
where the listener was located. The added benefit of global ambience as
fleshed out by the back channels was equally beneficial.
The earliest five-channel carriers for music-only surround programming include the DVD Video format (with Dolby AC-3 encoding) and
the DTS (Digital Theater Sound) CD format, both presenting five fullrange channels and a single subwoofer effects channel operating below
100 Hz. (For obvious reasons, the format is known as 5.1 surround
sound.) As the new millennium got under way, two formats, DVD Audio
and Sony/Philips SACD (Super Audio Compact Disc), were both proposed for audio-only applications for surround sound. At the present
time, both these mediums have made a respectable bid for marketplace
acceptance, but neither has been the runaway success that many had
2:56 PM
Page 244
The first five years of the new century have seen so much development in surround technology, including microphone design, microphone
pickup techniques, and playback options, that it will be advantageous to
discuss the technology first in a chapter devoted solely to those topics.
The next chapter will then cover case studies of the various techniques.
For our purposes we will consider four categories of surround sound
recording and playback:
1. Stereo-derived: There are many formats here, and they are basically
derivations of current stereo techniques in that they make use of
both real and phantom images as well as ambient effects that arise
from decorrelated multichannel sources. Quadraphonic and current motion picture techniques fall under this category, as do special formats such as the TMH Corporation’s 10.2 configuration.
Virtually all of the surround remixes of both legacy and modern
multitrack source tapes fall in this category.
2. Single-point pickup: These techniques are used to sample and play
back a global three-dimensional sound field. The British Soundfield
microphone is an early example here using first-order directional
patterns. The more recent Eigenmike makes use of higher order
patterns. In either cases there is a direct correspondence between
the pickup and playback directions, and the general aim of these
systems is to duplicate natural spatial cues in the playback environment. The number of pickup elements is normally limited by the
order of the microphone patterns, but the playback setup can usually
handle more loudspeaker sources than microphones if their drive
signals are properly derived. In all of these systems it is essential
that the listener be positioned at the center, or “sweet spot”, of the
loudspeaker array.
3. Transaural, or head-related, pickup: This spatial sound transmission
technique uses a fairly small number of loudspeakers to duplicate at
the listeners’ ears the exact amplitudes and time relationships that
were physically present in the pickup environment. It is very critically dependent on loudspeaker–listener positioning, and its major
use is in controlled listening environments (such as those afforded
by computer workstations) where the listener’s head is confined to a
single position.
4. Systems with parallax: Like an acoustical hologram, these systems
allow the listener to move about in the listening space while virtual source locations remain stationary. Generally, a relatively
small number of recorded tracks may be involved, and the real
complexity comes in playback signal processing, where positional
information is superimposed on the various sources via impulse
measurements and signal convolution. The technique is experimental at present, but it holds great promise for use in special
2:56 PM
Page 245
15: Surround Sound Microphone Technology
For each category summarized above we will describe the general
technical requirements for both recording and optimum playback.
There are three acoustical elements necessary for convincing surround
sound reproduction:
1. Accurate pickup of direct sound originating on-stage; soundstage
imaging should be natural and unambiguous.
2. Pickup of sufficient early reflections from the stage and front portion of the recording space to convey a sense of room size and
dimension. These signals normally fall in the range from 25 to
60 ms after the receipt of direct sound and are generally conveyed
by all of the loudspeakers in the surround array.
3. Pickup of uncorrelated reverberation and its subsequent presentation over the entire loudspeaker array. The onset of the reverberant
field normally occurs at about 80 ms after the receipt of direct
Conventional two-channel stereo can deliver the frontal soundstage quite accurately for a listener seated directly on-axis, but cues
pertaining to early reflections and reverberation are limited by the
relatively narrow presentation angle in the playback environment. We
need to determine how many channels are actually necessary to do
justice to the demands of surround sound. Tohyama et al. (1995) developed data showing that a minimum of four channels is necessary to
create an accurate enveloping sound field in the consumer listening
environment. The basis for their study is shown in Figure 15–1. In this
figure, the inset shows an artificial head located in the completely
Interaural cross-correlation
(IACC) between the ears
of an artificial head
placed in a reverberant
2:56 PM
Page 246
IACC for stereo playback
of uncorrelated
reverberation in a typical
listening space.
diffuse sound field of a reverberation chamber, which is analogous to
the acoustical field component in a concert hall which conveys the normal sense of envelopment. The test signal consists of a slowly swept
band of noise ranging from 100 Hz to 10 kHz. The signals reaching the
microphones at the ear positions in the artificial head are compared,
and the mathematical cross-correlation between the ear positions is
measured and shown in the graph. At very low frequencies the correlation is unity, inasmuch as the microphone spacing is small relative to
the received wavelength.
As the signal increases in frequency the cross-correlation converges
to an average of zero, indicating that the received signals at the artificial
head are essentially uncorrelated, a basic condition for conveying a sense
of spatial envelopment at the ears of the listener.
Moving on to Figure 15–2, we now take an uncorrelated twochannel stereo recording made in the diffuse field of Figure 15–1 and
play it back in a typical living room environment. The artificial head is
again used to pick up the swept signal, and the interaural crosscorrelation is again measured and plotted, averaging three sets of stereo
loudspeaker playback angles. The graphed data show that the measured
signal correlation at the ears is not uniform, and in fact produces a
significant compromised sense of envelopment, especially at the critical
midrange frequencies between 800 Hz and 3150 Hz. As good as it
sounds, stereo is not capable of generating a convincing sense of spatiality
around the listener.
Figure 15–3 shows similar measurements, this time using four uncorrelated playback channels, averaging them at six different sets of bearing
angles around the measurement system. It is clear that the four-channel
array very nearly produces the same spatial cues as the reference condition shown in Figure 15–1, indicating that four (or more) channels of
ambience information can successfully produce an accurate enveloping
sound field in the home listening environment. A number of multichannel reverberation systems currently produce such decorrelated sound
2:56 PM
Page 247
15: Surround Sound Microphone Technology
IACC for quadraphonic
playback of uncorrelated
reverberation in a typical
listening space.
A typical playback setup for
quadraphonic sound
reproduction in the home
Figure 15–4 shows for reference purposes a typical quadraphonic playback arrangement. While the system could produce a good impression of
recorded ambience, the very wide spacing of the front pair of loudspeakers (90) was disturbing to many listeners, primarily because it did
not reproduce front-center phantom images convincingly. Because of
this, many consumers routinely reduced the frontal listening angle to
about 60, while leaving the back angle in the range of 90.
Surround sound for motion pictures has evolved over the years to the
arrangements shown in Figure 15–5. There are either two surround channels or three, depending on the vintage of the installation and its implementation of the Dolby EX-Plus center-rear surround channel. In the
motion picture theater, multiple surround loudspeakers are always used
in order to produce highly uncorrelated signals, the specific source of
which patrons cannot readily identify. This is a very desirable condition
and normally fits the use of the surround channels as carriers of special
environmental, often large-scale, effects in the typical motion picture.
2:56 PM
Page 248
Loudspeakers in the motion
picture theater: using two
surround channels (A);
using three surround
channels (B).
Typical home theater
surround sound
loudspeaker locations.
It is difficult to transport the complexity of the motion picture
approach into the home environment, and more often than not, only a
pair of loudspeakers will be used to convey a sense of surround envelopment for on-screen effects. A typical home theater loudspeaker array
is shown in Figure 15–6. Here, dipole loudspeakers are used for the
surround channels. The dipole loudspeaker exhibits deep nulls in its
response at angles of 90 relative to forward and rear axes, and the side
null axes are aimed at the primary listening area. The result is that the
signal from the surround channels reaches the listeners primarily by way
of room reflections and is thus less likely to call attention to itself as coming from distinct positions left and right of the listeners.
Dipole loudspeakers work well for surround music channels that are
primarily reverberant in their signal content. For more generalized music
presentation, the playback format shown in Figure 15–7 is recommended. This figure illustrates the standard loudspeaker configuration
recommended by the ITU (International Telecommunications Union,
1994) for the setup of listening environments for the mixing of surround
sound programs for professional activities. While these recommendations
2:56 PM
Page 249
15: Surround Sound Microphone Technology
ITU recommended
reference loudspeaker
locations for surround
sound monitoring; the
nominal 110 positions
may vary between 100
and 120.
are appropriate for surround sound presentation, it is felt by many
workers in the field that the left and right front loudspeakers, with their
target spacing of 60, may be too widely spaced for normal stereo listening
in the home.
In laying out a surround sound mix, the three frontal channels are
important in defining and positioning of primary musical elements. Our
hearing capability can lock on elements in the frontal soundstage and
identify their lateral position within a few degrees. This acuity is far
more stable with three frontal channels than it is using phantom images
in two-channel stereo, and it substantially enlarges the effective seating
area for all listeners.
While we are very aware of sounds originating from the sides and
rear, we have considerable difficulty in determining their actual positions. Sounds arriving from the sides are generally highly uncorrelated at
the ears and as such contribute to a sense of ambience in the reproduced
recording. Because of the ear’s difficulty in assigning specific directionality at the sides, a single pair of channels can handle the needs for conveying ambience very well, especially if the loudspeakers are arrayed in
large quantities, as in the motion picture theater, or as diffuse-radiating
dipole loudspeakers. Thus, the obvious use of the surround channels in
classical music is in conveying early reflections and reverberant sound in
the recording space. Note that the rear loudspeakers in the ITU configuration are positioned more to the sides than the back. In this position
they can convincingly deliver both early reflections and reverberant
Pop mixes are routinely made with secondary musical elements
panned into the surround channels; here we include such elements as
background vocals and rhythm fills. Live concerts usually fare very well
in surround sound through the combination of audience reactions and
secondary musical elements in the rear channels.
2:56 PM
Page 250
Layout of the TMH
Corporation proposed
10.2 array.
Extending the 5.1 concept to a greater degree, Holman (2000) has
suggested the 10.2 channel approach shown in Figure 15–8. The widefront channels are used primarily to enhance important early side reflections, which help to define the acoustical nature of a good recording
space; the center back channel ensures spatial continuity and permits
unambiguous rear imaging. The front overhead channels provide height
information which, like the front-wide channels, enhances spatiality and
adds realism to the presentation. The use of two low frequency channels
restores some of the spatial subtlety in the lower frequency range which
results from side-to-side pressure gradients at long wavelengths. Such
effects may be lacking when a single channel is used below 100 Hz.
Holman further points out that the exact positions of the added loudspeakers are subject to a moderate amount of leeway in actual positioning in the playback space. In practical application of the 10.2 format, the
notion of “virtual” microphones (see Chapter 19) is useful in the generation of secondary reflected and ambient signals.
While most surround sound recording is carried out using conventional
microphones and techniques, with assignment of the various signals into
the five cardinal points of a surround sound array, a number of microphone designs and microphone arrays are used for recordings intended
primarily for surround presentation. Some of them are discussed below.
2:56 PM
Page 251
15: Surround Sound Microphone Technology
Single-point quadraphonic
microphones: using
capacitor elements (A);
using ribbon elements (B).
(Figure courtesy of Journal
of the Audio Engineering
Going back to the quadraphonic era, a number of microphones were
developed that provided for one-point pickup of sound from four directions, a concept that is still useful in certain aspects of surround sound.
Typical here is the single-point four-cardioid array developed by
Yamamoto (1975), shown in Figure 15–9, either as a group of cardioid
capacitor elements (A) or ribbon elements (B) picking up sound in the
azimuthal (horizontal) plane. In either case, the mechanical and acoustical elements in the designs have been adjusted so that the target cardioid
directional characteristics of each element are maintained around the
circle with smooth transitions between adjacent pairs.
The Schoeps KFM 360 spherical array
Developed by Bruck (1997), the KFM 360 produces a dual set of MS
stereo patterns, one set facing the front and the other facing the rear.
Both MS sets are attached to a sphere 18 cm in diameter, as shown in
Figure 15–10A. The pattern choices are shown at B. The user can choose
the forward patterns independently of the rear patterns, thereby altering
front-back perspectives. The microphone array is fed to the DSP-4 control unit shown in Figure 15–11A, and a signal flow diagram for the unit
is shown at B.
The gradient microphones enable this fairly small array to produce
excellent fore-aft spatial delineation, and the effect can be further heightened through use of the rear channel delay function. Note that the control unit produces a front-center output via the Gerzon 2-to-3 matrix
circuit (see below under Systems with Parallax).
2:56 PM
Page 252
FIGURE 15–10
The Schoeps KFM 360
four-microphone array:
photo of left side of unit
showing imbedded pressure
microphone and external
gradient microphone (A);
front and rear pattern
selection for the KFM 360
(B). (Photo courtesy of
The Holophone Global Sound microphone system
This system has a set of five pressure microphone elements located on the
surface of an oval-shaped (dual radius) ellipsoid. The microphones can be
extended out a short distance from the ellipsoid for enhanced separation.
In its wireless form the system is excellent for field effects gathering.
2:56 PM
Page 253
15: Surround Sound Microphone Technology
FIGURE 15–11
DSP-4 control unit for
Schoeps KFM 360; photo
of unit (A); signal flow
diagram for control unit
(B). (Data courtesy of
The surround ambience microphone (SAM) array
An arrangement of four cardioids at 90 has been proposed by Theile of
the German IRT (Institut für Rundfunktechnik). As shown in
Figure 15–12, these microphones are located at the corners of a square
measuring 21 to 25 cm per side (8–10 in). The array is known as SAM
(surround ambience microphone). The spacing of the elements of the
array adds significant time cues to supplement the amplitude cues delineated by the microphone patterns. The array is generally intended to pick
up ambience in live spaces in conjunction with traditional accent microphones center channel pickup (Theile, 1996).
The sound performance lab (SPL) array
Figure 15–13A shows a photo of the German SPL array of five multipattern microphones. The array offers a high degree of flexibility, with
2:56 PM
Page 254
FIGURE 15–12
Details of the SAM
cardioid microphone array.
FIGURE 15–13
Details of the SPL
microphone array: photo
of array (A); plan view of
array showing normal
dimensions (B). (Photo
courtesy of SPL, USA.)
2:56 PM
Page 255
15: Surround Sound Microphone Technology
individual setting of distances between microphones (using telescoping
sections) and horizontal aiming directions for each microphone. The orientation shown at B represents normal operating distances among the
five microphones. An array such as this may be used to produce a basic
spatial “signature” of a recording environment. Accent microphones
may be used as needed with the basic array. The control unit provides a
number of functions, including microphone pattern switching, ganged
level control, sub-bass output control, and various panning functions.
Note that the SPL array bears a striking resemblance to the Decca tree,
as discussed in Chapter 13.
In this section we will describe methods of specifically picking up the
three frontal signals for optimum effect and separation. These arrays
may be used with a variety of pickup methods for general room ambience. One of the problems with three-microphone frontal arrays with
each microphone fed to its own frontal loudspeaker is the reproduction
of three phantom images produced by the microphones taken two at a
time: L-C, C-R and L-C. If there is considerable pattern overlap of the
three microphones, the phantom images between the outside pair will
conflict with the real image produced by the center loudspeaker. This can
be minimized by using hypercardioid or supercardioid patterns for the
left and right microphones, which will minimize this tendency
Klepko (1997) describes the microphone array shown in
Figure 15–14. Here, the left and right microphones are hypercardioids,
and the center microphone is a standard cardioid. The use of hypercardioids for left and right pickup will minimize the strength of any phantom image between the left and right loudspeakers in the playback
Schoeps proposes the frontal array shown in Figure 15–15. Known
as OCT (Optimum Cardioid Triangle), the array uses increased spacing
and supercardioids facing full left and full right to minimize phantom
images between outside loudspeakers. The center cardioid element is
positioned slightly forward of the two supercardioids to give the center
channel signal a slight anticipatory time advantage for frontal pickup,
relative to the left and right microphones. The added omni microphones
FIGURE 15–14
Three-microphone frontal
array proposed by Klepko.
2:56 PM
Page 256
FIGURE 15–15
Three-point frontal array
proposed by Schoeps.
FIGURE 15–16
Details of a coincident
array of three second-order
at left and right are primarily for restoring the normal LF rolloff caused
by the supercardioid microphones.
A second-order cardioid coincident frontal array can be conceived as
shown in Figure 15–16 (Cohen and Eargle, 1995). While a number of
engineering obstacles still remain in realizing this configuration, the
array has pickup characteristics that virtually eliminate phantom images
between left and right loudspeakers, while providing uniform pickup
over a 150 frontal angle. Adjacent crosstalk along the common left and
right microphone axis is about 15 dB, clearly indicating that signal
commonality between L and R will be minimum. The equation of the
second-order pattern is:
(0.5 0.5cos )(cos ).
2:56 PM
Page 257
15: Surround Sound Microphone Technology
The foregoing discussion has dealt with minimizing crosstalk, or
leakage, from the center channel into the front flanking channels, and the
microphone arrays we have shown will accomplish this. However, in
many pop mixes there is a preference for a strong frontal presentation of
a soloist, and many remix engineers and producers will purposely feed
some of the center channel into the left and right flanking channels to
accomplish this. The level of this added feed is about 3 dB lower than
that present in the center channel, and it produces something resembling
a “wall of sound” from the front. This a judgement call on the part of
the engineer and producer and is based on musical requirements rather
than technical requirements. Proceed with caution when doing this.
The reader will have noticed that several of the microphone arrays discussed so far have no specific center channel microphone. In these cases,
as well as in remixing older stereo program material for surround sound,
it is often necessary to synthesize a center channel from a stereo pair.
Gerzon (1992) describes a matrix network that accomplishes this,
albeit with some loss of overall left-right separation in the frontal threeloudspeaker array. The circuit, shown in Figure 15–17, may look fairly
complex, but it can easily be set up on any console that has switchable
polarity inversion in its line input sections as well as a flexible patch bay.
It can also be set up on the virtual console pages of many computer based
editing programs.
Here is an example of its application: Assume we have a stereo
recording with three main signal components: left, phantom center and
right. The stereo channels may then be represented as:
LT L 0.7C
RT R 0.7C
where C represents the panned center signal, such as a vocalist, and L
and R represent discrete left and right program elements. LT and RT comprise the left-total and right-total stereo program. The matrix separation
FIGURE 15–17
Gerzon 2-3 matrix for
deriving a center channel
from a stereo program pair.
2:56 PM
Page 258
angle is normally set in the range of 45, giving values of sin and cos of
0.7. For this value, the three output signals of the matrix will be:
Left 0.85L 0.5C – 0.15R
Center 0.5L 0.7C 0.5R
Right –0.15L 0.5C 0.85R
You can see that L, C and R components (shown in boldface) are dominant, respectively, in their corresponding output channels. It is also clear
that the crosstalk has increased noticeably. A higher value of will
increase left–right separation, but at the expense of center channel level.
Gerzon recommends that a value of 55 be used at frequencies above
about 4 kHz, while a value of 35 be used at lower frequencies. Such an
approach actually calls for two matrices and may be needlessly complicated. If you wish to implement the Gerzon matrix, we recommend that
you proceed carefully in establishing an acceptable balance.
Single-point surround pickup arrays are generally intended to map the
global sound field in a performance space into a loudspeaker array in the
playback space. A coincident stereo microphone pair is, in a manner of
speaking, a rudimentary version of this, but generally we think in terms
of four or more transmission channels with microphones in a threedimensional array. The Soundfield microphone, as introduced by Gerzon
(1975), uses four microphone to create an array of first-order directional
patterns, each of which can be assigned to a loudspeaker position in
space matching that of the microphone. Johnston (2000) has suggested
an array of seven rifle microphones with a corresponding playback setup.
Meyer (2002) proposes a third-order microphone array, known as the
Eigenmike, which supports a playback array of 16 loudspeakers.
Questions of practicality are bound to arise when the number of
playback channels increases beyond five or six – at least as far as the
home environment is concerned. On the other hand, there are many
special entertainment venues and occasions where multichannel techniques can be taken to the limit.
As we discussed in Chapter 5, any first-order cardioid pattern may be
produced by combining an omnidirectional pattern with a figure-8 pattern. By selectively combining a single omnidirectional capsule with three
figure-8 elements, individually oriented in left-right, up-down, and foreaft directions, it is possible to produce a first-order cardioid pattern
pointed at any direction in space. Gerzon (1975) developed an array of
four subcardioid patterns, each oriented parallel to the sides of a regular
tetrahedron and contained within the structure shown in Figure 15–18A.
2:56 PM
Page 259
15: Surround Sound Microphone Technology
FIGURE 15–18
Details of the Soundfield
microphone. Photo of
microphone (A); diagram
showing signal flow in the
control unit (B); details of
the B-format (C); photo of
control unit (D); detail
functions of the control
panel (E). (Photos courtesy
of Transamerica Audio
A rear view of the four capsule elements comprising the A-format is
shown at B. The array elements occupy a relatively small space, providing balanced resolution in all directions over a wide frequency range.
The orientation of the subcardioid elements, as seen from the rear of the
assembly, is as follows:
Forward elements:
Back elements:
Left-up (LU)
Right-down (RD)
Right-up (RU)
Left-down (LD)
5:01 PM
Page 260
FIGURE 15–18
These four A-format elements are combined to produce the four
B-format components, as shown:
W Pressure component (LU RD RU LD)
X Fore-aft velocity component (LU RD) (RU LD)
Y Left-right velocity component (LU LD) (RD RU)
Z Up-down velocity component (LU RU) (RD LD)
Graphic details of the B-format are shown at Figure 15–18C, and
these four elements can be combined to produce the entire range of the
first-order cardioid family oriented in any direction. A photo of the front
panel of the control unit is shown at D, and further details of the individual controls are shown at E. Normally, four separately resolved output are used for surround sound applications, along with mono and
stereo outputs.
In its stereo (two-channel) mode of operation, the following manipulations can be carried out remotely at the control unit:
1. Rotation of the patterns, with no physical contact or manipulation
of the microphone assembly itself.
2. Electrical adjustment of the stereo pickup plane for any desired
downward tilt angle, with no physical movement involved.
3. A front-back dominance control to “focus” the stereo pickup
toward the front (for greater soloist dominance) or toward the back
(for greater reverberant pickup).
2:57 PM
Page 261
15: Surround Sound Microphone Technology
As such, the Soundfield microphone has great application as a permanent fixture in auditoriums and concert venues in schools of music
and at music festivals, where many instrumental groups, each requiring
its own stereo microphone pickup setting, may be used during a single
program with virtually no interruption in the flow of things. To facilitate
stereo applications, the Soundfield microphone may be used with a simplified stereo-only control unit.
When used in connection with surround sound presentation, the
Soundfield microphone’s four primary outputs can be positioned anywhere in the listening space and B-format settings adjusted so that the
output from each of the four (or more) loudspeakers corresponds exactly
to that of a first-order microphone oriented in that direction.
Figure 15–19 shows a perspective view of the Johnston-Lam (2000)
microphone array. Five hypercardioid microphones are oriented at 72
intervals in the horizontal plane. The vertical microphones are short line
models with broad response nulls at 90 to minimize the amount of lateral signal pickup in those channels. All seven microphones are located
on a sphere with a diameter of approximately 290 mm (11.5 in). In
normal practice the array is placed at a typical listening position in a
performance space, elevated about 3 m (10 ft) above the floor.
According to the authors, microphone spacing, orientation, and
selection of response patterns are chosen to preserve the interaural time
differences (ITD) and interaural level differences (ILD) picked up by the
microphones and which are subsequently heard by a listener located at
or near the center of the playback loudspeaker array. The ITD and ILD
FIGURE 15–19
A view of the Johnston-Lam
microphone array.
2:57 PM
Page 262
are the primary cues the ears need for localization, and the authors suggest that recreating them in this manner is a more realistic goal than
attempting to recreate the minute details of the actual sound field itself.
The authors state that the approach is essentially compatible with playback over a standard ISO surround loudspeaker array if the height information is appropriately mixed into the playback array. Ideally, the
loudspeakers should be positioned at 72 intervals about the listener.
Meyer and Agnello (2003) describe a relatively small spherical array
about 75 mm (3 in) in diameter that contains 24 miniature omnidirectional electret microphones equally spaced on its surface. These microphone outputs can be selectively combined and adjusted in relative delay
to create first, second, or third-order pickup patterns arrayed equally in
three dimensions.
The required number of microphones for optimum spherical coverage
is given by:
Number of microphones (N 1)2
where N is microphone order. For example, in the Soundfield microphone first-order array the equation gives a value of 4, and as we have
seen there four elements in the B-format from which all remaining patterns are derived. In a second-order system there are 9 elements, and in
a third-order there are 16 elements. Figure 15–20 shows a view of the
FIGURE 15–20
Photograph of the
Eigenmike. (Photo
courtesy of mh-acoustics,
Summit, NJ.)
2:57 PM
Page 263
15: Surround Sound Microphone Technology
FIGURE 15–21
Formation of Eigenbeams:
third-order pattern along one
axis (A); third-order pattern at
and each of 20 (B); a
two-dimensional polar
plot of a third-order
hypercardioid pattern at
3 kHz (C). (Data courtesy of
Meyer and Agnello, 2003.)
In a manner analogous to the A and B formats that used in the
Soundfield microphone, the outputs of the multiple elements are combined in two stages. The resulting flexibility here allows Eigenbeams, or
directional microphone elements, to be formed continuously over the
sphere independently of the number of actual sensors. Figure 15–21A
shows a synthesized third-order hypercardioid pattern in three dimensions and positioned along one axis. At B is shown the same pattern positioned at and angles, each of 20. A two-dimensional view of the
pattern for 3 kHz is shown at C.
In a typical application involving an array intended for sound
field reconstruction, third-order patterns would maintained above
about 1.5 kHz. At progressively lower frequencies the directivity
would be reduced to second-order and finally, below 700 Hz, to firstorder. The compromise here is between spatial resolution and LF noise,
as discussed in Chapter 6. Alternatively, the entire array size can be
increased to offer a better set of tradeoffs between pattern control and
noise at LF.
In addition to audio applications the array has great promise in
acoustical architectural measurement applications, where multichannel
recording can be used to store multiple impulse data for later directional
analysis off-line.
2:57 PM
Page 264
Cooper and Bauck (1989) use the term transaural to describe sound
recording and reproduction systems in which the directional response is
determined by conditions existing independently at the listener’s ears. By
comparison, normal stereo reproduction is defined principally in terms
of what comes out of the loudspeakers, and of course the sound from
each loudspeaker is heard by both ears.
Transaural recording and playback is a direct outgrowth of binaural
sound, which we discussed in Chapter 12. The basic plan is shown in
Figure 15–22, where the two channels of sound picked up by an artificial head are fed through a crosstalk canceller. In this stage, delayed signals from each loudspeaker are cross-fed and arrive at opposite ears with
the exact delay time required to cancel the normal stereo leakage, or
crosstalk, signals.
The details of the Schroeder-Atal crosstalk cancelling circuit are
shown in Figure 15–23, where the binaural signal pair is transformed by
the two signals labeled C, where C A/S. S and A are, respectively, the
transfer functions from the loudspeakers to listener for the near-side
signal (S) and the far-side signal (A) at the ears. The performance of the
crosstalk-cancelling circuitry is normally set for a given listening angle
), and the system’s cancellation performance will be optimum for
FIGURE 15–22
Transaural technology:
a binaural signal can
be played back directly via
headphones or over a pair
of loudspeakers using
crosstalk cancellation.
2:57 PM
Page 265
15: Surround Sound Microphone Technology
FIGURE 15–23
Details of the
Schroeder-Atal crosstalk
cancellation network.
loudspeakers positioned at that angle. Off-axis listening, or listening onaxis at other angles, will result in non-optimum performance.
Transaural processing works most convincingly when the A and S
data are actually the same as those measured for a given listener – that is,
when the listener’s head is actually substituted for the artificial head in
making the measurements. Generally, however, measurements made on a
typical artificial head will give results that are a good approximation to
the shape of an average adult head. In some laboratory playback systems, individual head-related transfer functions (HTRFs) can actually be
accessed for improved performance.
For studio use Cooper and Bauck have developed the panning system shown in Figure 15–24. Using impulse response methods, the
HTRFs for a given head are measured at each desired azimuthal position
about the head (A). These measurements are stored and can be introduced into a panning system (B) that will translate a given input signal
to a specific position over the frontal 180 listening arc as perceived by a
listener. Positioning can be synthesized for each 5 angular increment.
Using a system such as this, a two-channel transaural presentation
can be generated from a five-channel surround program in which the
listener will hear each of the five channels originating from its proper
2:57 PM
Page 266
FIGURE 15–24
Details of a transaural panning scheme: making HTRF measurements (A); overall detail of the panner (B). (Data after
Cooper and Bauck, 1989.)
position in virtual space. The effect we have described is as shown in
Figure 15–25.
A note on performance: transaural virtual images are only as stable
as the listener’s head is stationary. A pair of loudspeakers is required for
each listener, and if listeners move their heads from side to side, the virtual images will move with them. With the aid of a technique called head
tracking, compensating signals can be fed back to the synthesizer, partially correcting the tendency for the images to move when the listener’s
head rotates from side to side.
The transural technique has great promise when used in conjunction
with computer monitors and television sets. Today, many TV programs
are broadcast in surround sound that has been encoded for transaural
playback. For TV loudspeakers located about 0.5 m (20 in) apart, a listening distance of approximately 2 m (6.7 ft) will usually suffice to give
the listener a good impression of what transaural techniques can provide.
An on-axis location is critical, and normally only a single listener can
2:57 PM
Page 267
15: Surround Sound Microphone Technology
FIGURE 15–25
Overall view of a
transaural presentation
of a five-channel surround
program using a stereo pair
of channels.
appreciate the effect to the fullest extent. The same technology is often
applied to video games and other computer programs. Here, the viewing
distance is much closer than is the norm for TV, and appropriate adjustments will have to be made.
Transaural systems have generally required extensive internal signal
processing, but the costs of such operations continues to drop, making
them more accessible to modern application in listening environments,
such as computer workstations, where listening positions are stable and
can be accurately estimated.
In a landmark paper on motion picture sound technology, Snow (1953)
described the “perfect” record/playback system shown in Figure 15–26.
A horizontal array of microphones communicate, each independently, to
a horizontal array of loudspeakers in the theater environment. A diverging wavefront at the microphones creates a sequentially firing set of individual wavefronts from the loudspeakers which coalesce, via Huygen’s
law on wavefront reconstruction, into a near copy of the original wavefront impinging on the microphones. We may think of this as an acoustical hologram in which listeners in the theater are free to move around
and hear individual acoustical sources in their proper left-right, fore-aft
relationships – in other words, as fixed in space as real sources.
The hardware requirements are staggering to consider, and such a
system has long been little more than a dream. However, modern technology has come to the rescue with multiple small loudspeakers, small
2:57 PM
Page 268
FIGURE 15–26
Snow’s multichannel
solution to motion picture
stereo. (Data courtesy of
Journal of the Society of
Motion Picture and
Television Engineer.)
FIGURE 15–27
Playback environment for
holographic sound
presentation. (Data after
Horbach, 2000.)
amplifiers, and relatively low-cost digital signal processing capability.
Horbach (2000) describes one approach to this.
A playback environment is visualized as shown in Figure 15–27.
Here, a “wraparound” loudspeaker array is placed in front and smaller
linear arrays are placed along the back wall. Ideally, we would like the
loudspeakers to be small and very closely packed; in practice, one loudspeaker every 15 cm (6 in) is about as dense as we can reasonably get.
The data that is actually recorded consists only of the relative dry,
unreverberated instrumental tracks and other basic musical elements
produced acoustically in the studio. Details of the overall acoustics of the
recording space will have been gathered using impulse measurements,
and this added data, along with the actual music tracks, will be used to
reconstruct the recording during the playback process.
2:57 PM
Page 269
15: Surround Sound Microphone Technology
FIGURE 15–28
Recording environment for
holographic sound pickup.
(Data after Horbach,
A typical recording setup is shown in Figure 15–28. Here, a large
number of omnidirectional microphones are arranged in crossed line
arrays as shown. The small squares indicate the target locations of individual instruments in the recording space.
The first step is to build up a large library of impulse responses. At
this stage, impulse signals are produced at each instrument location
(indicated by the numbered squares in the figure). For each studio location, impulse measurements will be made at all of the microphone locations. If there are 10 studio locations and 31 microphone locations, there
will be 10 31, or 310 impulse response files. Each impulse file is relatively small and of course needs to be made only once. Other impulse
files detailing the overall studio reverberation characteristics may be
made as well.
When the final program mix is constructed, the engineer and producer bring together all of the individual recorded source tracks and all
of the impulse data files. A stage plan is then laid out with each instrument or musical element assigned a specific position, corresponding to
one of the numbered squares in Figure 15–28. The impulse files that
define the sound stage boundaries for that signal position are then
accessed and assigned to a specific playback channel. When all of the
impulse files are convolved with the audio track and monitored over the
multi-loudspeaker playback array, the producer and engineer will hear
the audio track positioned accordingly behind the loudspeaker array.
This assignment process is repeated until all of the audio tracks have
been appropriately positioned on the virtual stage.
2:57 PM
Page 270
FIGURE 15–29
Playback of a virtual
source and the directions
perceived at two listening
positions. (Data after
Horbach, 2000.)
FIGURE 15–30
Overview of signal flow in
a system with parallax.
Figure 15–29 shows how a typical virtual source is made to appear
fixed in a given location behind the loudspeaker array. Within certain
limitations, it is also possible to position virtual sources in front of the
loudspeaker array.
The methods we have just described are more akin to modern pop studio music production techniques than to verbatim recreation of natural
acoustical sound fields. While the basic data transmission rate from studio to playback is not inordinately high (and carries with it many opportunities for beneficial data reduction), the actual on-site playback data
processing rate can be quite high. As shown in Figure 15–30, only the
raw audio information is continuous; impulse data files and reconstruction information are one-time only events and can be downloaded via a
preamble to the program file. Needless to say, the playback hardware
must be efficiently designed to handle all of this on a timely basis.
2:58 PM
Page 271
In this section we will examine six surround recordings that demonstrate
a number of the techniques discussed in the previous chapter. These
examples range from fairly modest studio resources up to large concerted
works that include chorus, organ, piano, and percussion ensembles in
performance with normal orchestral resources. In each case, recording
channel limitations restricted us to no more than eight tracks. This called
for careful planning, since the eight “stems” had to be used not only for
creating surround mixes but also for making any adjustments in the stereo
mix during later postproduction. In this connection we will introduce a
concept of subtractive mixing, as it applies to demixing and subsequent
remixing, of tracks to attain an ideal balance.
All of the projects studied here were targeted for both stereo and
standard 5.1 channel playback. In each case we will present the basic
microphone choice and layout, channel assignments, mixing assignments
in both stereo and surround, and an overall commentary. All of the projects discussed here have been commercially released, and a discography
is included at the end of this chapter.
The items to be discussed are:
1. Berlioz: March to the Scaffold from Symphonie Fantastique
2. Tchaikowsky: 1812 Overture (orchestra with chorus)
3. Gershwin: Rhapsody in Blue (orchestra and piano)
4. Berlioz: Te Deum (orchestra with chorus, vocal soloist, and organ)
5. Bizet-Shchedrin: Carmen Ballet (strings and large percussion
6. Schnittke: Piano Concerto (strings and piano)
2:58 PM
Page 272
What is the purpose of a surround sound musical presentation? One
answer, especially for classical music, is to present that music in the context of the original performance venue with its acoustics intact. For pop
and rock music another goal may be to place the listener in the center of
an imaginary stage, with sound elements widely positioned all around.
Both are respectable goals, and there should be no prejudice favoring one
over the other.
The recordings described here all fit neatly into the former, or so-called
direct-ambient, approach in which there is a stage at the front and an
enveloping ambience, based on the peformance space, that reaches the
listener from all directions. This does not happen casually, and there are
some general procedures that should be followed. They are:
1. Controlled monitoring environment: An ITU standard loudspeaker
monitoring array was used with reference levels set to within 0.25 dB.
Even though the mix was made with engineer and producer sitting
at the “sweet spot,” the program was auditioned at various points
within the periphery of the loudspeakers to ensure that it was
generally and broadly effective.
2. Decorrelated ambience: The basic ambient signature of the performance space is presented over all loudspeakers except the front center,
and these four ambience signals are all decorrelated, either through
signal processing or through the original spacing of microphones.
3. Control of early reflections: In any performance venue, early reflections arising primarily from side walls help to define the boundaries
of the space and convey a sense of room size. If the reflections are
too early and too high in level, they may add muddiness and
confusion. The optimum delay range for them is between 25 and
40 milliseconds, presented at a level between 10 and 15 dB
relative to the primary signal. If we’re lucky, these signal components are produced directly in the room and are picked up naturally; at other times, the engineer has to create them via signal
processing using modern delay and reverberation generators. These
signals are normally introduced into the mix only in the left and
right-front channels as well as the two rear channels. In some cases
it is effective to delay the rear signals slightly, relative to the front
left and right signals. You will see application of these techniques
as we progress through this chapter.
4. Layers: When multiple microphones have been used in a recording –
and this is the norm today – it is often helpful to think in terms of
structural layers in the recording. For example, the main frontal
microphones (and soloist microphones if appropriate) comprise a
single layer. The ensemble of accent, or spot, microphones in the
2:58 PM
Page 273
16: Surround Recording Case Studies
orchestra will comprise another layer, and the house microphones,
which convey only reverberation and ambience, will comprise yet
another layer. Normally, if level changes are made in postproduction, the entire layer will be altered as a unit in order to avoid any
skewing of the sound stage. During the recording process however
individual microphones, or microphone pairs, may be altered on a
running basis as dictated by musical requirements.
5. Track assignments and allocation: When using fewer recording
tracks than there are microphones, informed decisions must be
made regarding which microphones to combine when laying down
the initial tracks. In general the decision is made to subdivide the
recorded elements into basic pre-mixed tracks, or stems, which can
be more easily manipulated during postproduction. Since all of the
works recorded here were to be released first in stereo, a decision
was made to aim for a finished live-to-stereo mix on tracks 1 and 2,
while ensuring that the remaining stems yielded enough flexibility
for creating surround mixes as well as for making minor adjustments in the stereo mix itself. While all eight tracks were edited at
the same time, the stereo tracks were used as as the basic guide for
all edit decisions.
The recording venue was the New Jersey Performing Arts Center, a 2500seat hall with a midrange reverberation time of about 2.3 s. Zdenek Macal
conducted the New Jersey Symphony Orchestra. Diagrams of the stage
layout are given in Figure 16–1, showing positions of the microphones.
The recording was made over three evenings with audience present and
was to be used for both stereo CD production and release in surround
sound. The direct-to-stereo mix was monitored at the sessions and fed
directly to channels 1 and 2 of the six-channel recorder (20-bit/48 k
sampling). Details of microphone deployment are shown in Table 16–1.
Note the use of only two omnidirectional microphones in the setup.
This is in keeping with the principle stated in the chapter on classical
recording in which the main ORTF pair and its associated flanking omnis
provide the basis of the pickup. All other microphones are secondary to
the main four and are intended for subtle heightening of instrumental and
sectional presence and balance. As such, these microphones function best
with the side and back rejection that the cardioid pattern offers.
The accent microphones were all fed, in varying amounts, to a digital
reverberation side chain whose parameters were adjusted to match those
of the hall itself. The stereo returns from the reverberation generator
were panned left and right into the overall stereo mix. The house microphones were aimed at the upper back corners of the hall to minimize
pickup of direct sound from the stage.
2:58 PM
Page 274
Stage and microphone
layout for “March to the
Scaffold:” center section
view (A); plan view (B).
The major components of the stereo soundstage are shown in
Figure 16–2A. In creating the surround soundstage, it was necessary to
“demix” the main ORTF pair from the stereo array and re-pan them into
the frontal sound loudspeaker array as shown in Figure 16–2B. This was
done using a technique known as subtractive mixing, which can be
2:58 PM
Page 275
16: Surround Recording Case Studies
explained as follows:
Let the main array consist of the following four components:
where LF and RF are the outputs of the left and right flanking microphones, and LM and RM are the outputs of the left and right main ORTF
TABLE 16–1
Microphone Deployment for “March to the Scaffold”
Major components
Stereo mix left
Stereo mix right
L flank
R flank
L house
R house
Mic height
Track 1
Track 2
3.5 m (12 ft)
3.5 m (12 ft)
3.5 m (12 ft)
3.5 m (12 ft)
4 m (13.3 ft)
4 m (13.3 ft)
Accent microphones (these appear only in the stereo mix)
L woodwinds
3.5 m (12 ft)
R woodwinds
half-right 3.5 m (12 ft)
1 m (40 in)
2 m (80 in)
half-right 3 m (10 ft)
Basses (1st stand)
2 m (80 in)
Target recorded
soundstages for “March to
the Scaffold:” stereo (A);
surround sound (B).
Track 3
Track 4
Track 5
Track 6
2:58 PM
Page 276
microphones. The symbol g represents a gain factor for the two ORTF
signals, and it is carried over into the signals recorded on tracks 3 and 4
of the digital recorder.
The next step is to subtract the ORTF pair from LT and RT. This can
be done because the value of g is the same in tracks 1 and 2 as it is in
tracks 3 and 4; the result is:
Having “stripped out” the ORTF pair from the left and right frontal
channels of the surround array, we are free to reposition those signals
anywhere in the array we wish. The usual approach is to reintroduce
them by panning LM slightly left of the center channel and RM slightly
right of the center channel, as shown in Figure 16–2B. As a matter of
practicality, it is possible to carry out subtractive mixing on any console
that has phase reversal stages in each input strip and the facility for
patching and multing various signals. With practice, the process can be
easily carried out by ear, listening for the “null zone” as you alternately
raise and lower the level of the antiphase components.
A functional view of the subtractive mixing process is shown in a
basic form in Figure 16–3. Here, four tracks have been recorded: a
stereo mix (LT and RT) on tracks 1 and 2, along with the main ORTF
Functional view of the
subtractive mixing process.
2:58 PM
Page 277
16: Surround Recording Case Studies
pair (LM and RM). Through subtractive mixing we want to separate the
flanking microphone signals (LF and RF) from the stereo mix. This is
accomplished by feeding the ORTF pair into faders three and four in
antiphase. When these are then fed into Groups 1 and 2, the LM and RM
signals will be canceled, leaving only LF and RF at the outputs of Groups
1 and 2. For the cancellations to be complete the levels must be matched.
This may be done by adding test signals at the head of all recorded tracks
before the session begins, ensuring that all gain settings downstream
from the faders are fixed. With a little practice, you can approximate this
condition by raising and lowering the phase-inverted faders through the
null region and then isolating the actual null position by ear.
Having adjusted the frontal array as discussed above, we then move
on to the requirements of the rear channels. In this recording, the distance
of the house microphones from the main array resulted in a relative delay
of about 44 ms, which is too great by about 20 ms. This value was compensated for on the virtual console of the digital editing system by delaying all signals other than the house microphones, producing a net time
onset of reverberation of 24 ms that was within normal bounds for good
surround performance. The final step was to re-introduce signals from
the main ORTF pair into the back channels to simulate added early
reflections from the sides. To do this, the outputs of LM and RM were
high-passed at about 200 Hz, delayed about 20 ms, reduced in level and
fed respectively to the left-rear and right-rear channels.
The recording venue was McDermott Hall in the Meyerson Symphony
Center, Dallas, TX. Andrew Litton conducted the Dallas Symphony
Orchestra and Chorus. As before, the recording was intended for both
stereo and surround sound release. The recorded track assignment was
composed of: stereo mix (tracks 1 and 2), main pair (tracks 3 and 4),
flanking pair (tracks 5 and 6) and house pair (tracks 7 and 8). Figure 16–4
shows stage and microphone layouts. The composite stereo monitor mix
was assigned to channels 1 and 2 of the digital recorder. Details of microphone deployment are shown in Table 16–2.
The basic pickup is very much like that of the previous example. The
use of omnidirectional microphones for chorus pickup is a judgemental
choice; the broad omni pattern ensures that more members of the ensemble will be covered than if cardioids were used. A drawback here is the
increased leakage of brass and percussion (back of the orchestra) into the
chorus microphones, a problem that required carefully planned gain
adjustments during the sessions.
In the a capella choral opening of the work, the flanking microphones were panned to the rear loudspeakers, giving the impression of a
chorus placed at the back of the hall. As the introduction progressed,
2:58 PM
Page 278
Stage and microphone
layout for 1812 Overture:
center section view (A);
plan view (B).
these microphones were progressively panned to front left and right.
Stage details are shown in Figure 16–4. The stereo and surround soundstages are shown in Figure 16–5. In setting up the surround mix, the
main microphone pair was demixed and re-panned center left and right
in order to produce a center channel. All accent microphones appear
2:58 PM
Page 279
16: Surround Recording Case Studies
TABLE 16–2
Microphone Deployment for 1812 Overture
Major components
Stereo mix left
Stereo mix right
L flank
R flank
L chorus
R chorus
L house
R house
Stage height
3.5 m (12 ft)
3.5 m (12 ft)
3.5 m (12 ft)
3.5 m (12 ft)
4.5 m (14.5 ft)
4.5 m (14.5 ft)
4 m (13.3 ft)
4 m (13.3 ft)
Accent microphones (these appear only in the stereo mix)
L woodwinds
3.5 m (12 ft)
R woodwinds
half-right 3.5 m (12 ft)
1 m (40 in)
2 m (80 in)
Basses (1st stand)
2 m (40 in)
Recorded soundstages for
1812 Overture: stereo
soundstage (A); surround
soundstage (B).
Track 1
Track 2
Track 7
Track 3
Track 4
Track 8
Track 5
Track 6
2:58 PM
Page 280
only in the front left and right channels. The house microphones were
located 8 m (27 ft) from the main pair and were panned into the rear
channels. They were aimed at the upper back corners of the house.
The opening of this work as auditioned in surround sound is an
excellent example of the condition shown in Figure 15–3, in which all
channels contain mutually uncorrelated reproduction of the chorus.
Positioning the chorus at the back of the surround soundstage during the
open measures of the work was purely experimental – but is musically
appropriate and does not sound contrived.
The work was performed by Andrew Litton, pianist and conductor, and
the Dallas Symphony Orchestra in McDermott Hall at the Meyerson
Symphony Center. The normal orchestral forces were reduced to 25 players,
in keeping with the original arrangement of the work for the Paul
Whiteman Orchestra in 1924. In this regard the recording is more like
that of a studio orchestra than a full symphonic ensemble, and microphone distances were reduced slightly for more presence. The cover of
Stage and microphone
layout for Rhapsody in
2:58 PM
Page 281
16: Surround Recording Case Studies
the piano was removed and the instrument oriented as shown so that the
conductor–pianist had direct eye contact with the players. Close pickup
of the piano, within 0.5 m (20 in), was mandated by the high ambient
orchestral levels surrounding the instrument. Figure 16–6 shows the
stage and microphone layout. Microphone deployment details are shown
in Table 16–3. The stereo monitor mix was assigned to tracks 1 and 2 of
the digital recorder.
Details of the stereo and surround soundstages are shown in
Figure 16–7.
The solo piano is a dominant feature in this recording and as such
needs to be positioned slightly in front of the orchestra. After some
experimenting in postproduction it was felt that the direct sound of
the instrument needed to be present in all three frontal loudspeakers, as
discussed in Chapter 15 under Frontal Arrays. By using only partial subtractive mixing, an approximate balance was reached, as shown in
Table 16–4.
The relationships given here produce a total relative level of 0 dB,
and the 3 dB advantage of the center channel ensures that the listener
will identify the source primarily at the center loudspeaker. The rear
channels were derived primarily from the two house microphones, supplemented by a reduced signal, high-passed at 200 Hz, from the main left
and right ORTF pair.
TABLE 16–3
Microphone Deployment for Rhapsody in Blue
Major components
Stereo mix left
Stereo mix right
L flank
R flank
L solo piano
R solo piano
L house
R house
Stage height
Track 1
Track 2
3 m (10 ft)
3 m (10 ft)
3 m (10 ft)
3 m (10 ft)
0.5 m (20 in)
0.5 m (20 in)
10 m (33 ft)
10 m (33 ft)
Accent microphones (these appear only in the stereo mix)
L woodwinds
3 m (10 ft)
R woodwinds
half-right 3 m (10 ft)
Orchestral piano
1 m (40 in)
2 m (80 in)
half-right 2 m (80 in)
Bass and tuba
2 m (80 in)
L drum set
1 m (40 in)
R drum set
half-right 1 m (40 in)
Track 3
Track 4
Track 7
Track 8
Track 5
Track 6
2:58 PM
Page 282
Recorded soundstages for
Rhapsody in Blue: stereo
soundstage (A); surround
soundstage (B).
TABLE 16–4
Piano level
Solo Piano Balance in Rhapsody in Blue
Front Left
Front Center
Front Right
–6 dB (1/4 power)
–3 dB (1/2 power)
–6 dB (1/4 power)
Berlioz’s Te Deum is a large work for orchestra, chorus, tenor solo, and
organ. It was performed with audience at the Cathedral of Saint John the
Divine in New York in 1996 during a convention of the American Guild
of Organists, Dennis Keene, conductor. The Cathedral is the largest gothic
structure in the world, with a nave extending 183 m (601 ft) front to back,
and the reverberation time of the occupied space is about 5 s. The orchestra was positioned in the large crossing, and the chorus was located on
multiple risers in the sanctuary. The vast dimensions of the space are such
that there are virtually no early reflections. The first reflections are from
walls that are about 25 m (83 ft) distant, so that the reflected sound blends
in comletely with the onset of diffuse reverberation.
2:59 PM
Page 283
16: Surround Recording Case Studies
As the postproduction of the recording got underway it was apparent
that the chorus sounded “very present and very distant” at the same
time. The obvious solution was to add ambience to the recording using
a program in the Lexicon model 300 reverberation generator. This program adds only a set of simulated early reflections and as such gave an
effect of immediacy and naturalness not present in the basic recording.
Another problem we faced was where to put the house microphones.
The audience in the nave numbered 4000, and there was no way to position stereo microphones 40 m (130 ft) above them. The solution here
was to position the house microphones 25 m (80 ft) above the chorus
adjacent to the organ pipes. As a result of this, you will hear the organ
Stage and microphone
layout for Te Deum:
elevation view (A);
plan view (B).
2:59 PM
Page 284
TABLE 16–5
Microphone Deployment for Te Deum
Major components
Stereo mix left
Stereo mix right
L flank
R flank
L chorus
C chorus
R chorus
L house and organ
R house and organ
Vocal solo
Stage height
Track 1
Track 2
3 m (10 ft)
3 m (10 ft)
3 m (10 ft)
3 m (10 ft)
3.5 m (11.5 ft)
3.5 m (11.5 ft)
3.5 m (11.5 ft)
25 m (83 ft)
25 m (83 ft)
2 m (80 in)
Track 3
Tracks 3/4
Track 4
Track 5
Track 6
Tracks 7/8
Accent microphones (these appear only in the stereo mix)
L woodwinds
3 m (10 ft)
R woodwinds
half-right 3 m (10 ft)
2 m (80 in)
1.5 m (5 ft)
and reverberant signature of the Cathedral both from the rear channels,
and this is very much the way the work was originally heard in Paris in
1855, where the organ was in the rear gallery.
As with the previous examples, the stereo monitor mix was recorded
on tracks 1 and 2 of the digital recorder. Figure 16–8 shows views of the
recording and microphone setup, and Table 16–5 shows the deployment
of microphones. Figure 16–9 shows the resulting stereo and surround
sound stages.
The Carmen Ballet is an arrangement of music from the opera Carmen
for strings and large percussion resources. The recording was made by
the Monte Carlo Philharmonic Orchestra, conducted by James DePreist,
24–27 June 1996, in the Salle Garnier in Monte Carlo, Monaco.
Figure 16–10 shows views of stage and microphone layout, and
Table 16–6 shows microphone and track deployment.
The engineer and producer made the decision early on to assign each
of the percussion microphones to a separate track. This provided utmost
flexibility in rebalancing any of the percussion resources as they might
crop up later in postproduction. In a sense, the percussion elements were
2:59 PM
Page 285
16: Surround Recording Case Studies
Recorded soundstages
for Te Deum: stereo (A);
surround sound (B).
FIGURE 16–10
Stage and microphone
layout in plan view.
2:59 PM
Page 286
TABLE 16–6
Microphone and track deployment for Carmen Ballet
Major components
Stereo mix left
Stereo mix right
Perc. left flank
Perc. left main
Perc. right main
Perc. right flank
House left
House right
Microphones appearing only in the stereo mix
String left flank
String left main
String right main
String right flank
Mic height
3.5 m (11.5 ft)
3.5 m (11.5 ft)
3.5 m (11.5 ft)
3.5 m (11.5 ft)
4 m (13 ft)
4 m (13 ft)
Track 1
Track 2
Track 3
Track 4
Track 5
Track 6
Track 7
Track 8
3.5 m (11.5 ft)
3.5 m (11.5 ft)
3.5 m (11.5 ft)
3.5 m (11.5 ft)
3 m (10 ft)
an unknown quantity in terms of balance and relative levels, whereas the
string ensemble was, by comparison, very stable and predictable.
Another important decision involving engineer, producer, and conductor was made regarding the relative perspectives of the string ensemble and the percussion ensemble. Physical constraints meant that the
large array of percussion instruments would have to be placed behind
the strings, but musical balance required that the two ensembles be
essentially equal in all respects. In other words, they had to occupy the
same apparent location on the recorded sound stage as the strings rather
than be heard from their natural position behind the strings.
The solution was simple; it was to place an identical set of main
microphones in front of each ensemble, treating them equally in terms of
presence and level. Figure 16–11 shows the resulting sound stages for
both stereo and surround presentation.
The recording was made on the large scoring stage at Lucasfilm’s
Skywalker Ranch in San Rafael, CA. Constantine Orbelian was both
pianist and conductor of the Moscow Chamber Orchestra. The recording space has dimensions roughly of 27 m (90 ft) by 18.3 m (60 ft) by
9 m (30 ft), and its boundaries can be configured for a variety of acoustical requirements. For this recording the room was set for “moderately
live” acoustics. Figure 16–12 shows details of studio setup and microphone placement. Note that the piano has been placed, with its cover
2:59 PM
Page 287
16: Surround Recording Case Studies
FIGURE 16–11
Recorded sound stages for
Carmen Ballet: stereo (A);
surround (B).
FIGURE 16–12
Studio and microphone
layout in plan view.
2:59 PM
Page 288
removed, so that it is well into the middle of the string group. Removing
the cover was a necessity for ensuring eye contact among all players.
Table 16–7 shows details of microphone and track deployment. Both
stereo and surround sound staging are shown in Figure 16–13.
TABLE 16–7
Microphone and track deployment for Schnittke Piano Concerto
Major components
Stereo mix left
Stereo mix right
Right ORTF
Piano left
Piano right
House left
House right
Mic height
3 m (10 ft)
3 m (10 ft)
2 m (80 in)
2 m (80 in)
3.5 m (11.5 ft)
3.5 m (11.5 ft)
Track 1
Track 2
Track 3
Track 4
Track 5
Track 6
Track 7
Track 8
These microphones appear only in the stereo mix
Main left flank
3 m (10 ft)
Main right flank
3 m (10 ft)
1.5 m (60 in)
FIGURE 16–13
Recorded sound stages for
Schnittke Concerto: stereo
(A); surround (B).
2:59 PM
Page 289
16: Surround Recording Case Studies
Because of fore–aft spacing limitations in the studio we introduced
an additional 20 milliseconds of delay to the house microphones in order
to “position” them as desired.
Hector Berlioz, “March to the Scaffold,” Symphonie Fantastique, DVD Music
Breakthrough, Delos International DV 7002, band 15.
Pyotr Tchaikowsky, 1812 Overture, DVD Spectacular, Delos International
DV 7001.
George Gershwin, Rhapsody in Blue, DVD Spectactular, Delos International DV
7002, band 12.
Hector Berlioz, Te Deum, DVD Spectacular, Delos International DV 7002, band 6.
Bizet-Schedrin, Carmen Ballet, DVD Music Breakthrough, Delos International
DV 7002, bands 3–5.
Alfred Schnittke, Piano Concerto, Delos SACD 3259 (multichannel hybrid disc).
3:00 PM
Page 290
The applications dealt with in this chapter include broadcast, news
gathering, paging in public spaces, conference management, and safety
alert systems. A number of microphone types used in these applications
have already been discussed in Chapter 10, which covers microphone
accessories. We refer to figures from that chapter where necessary as we
underscore the specific applications in this chapter.
The desk stand microphone mount is one of the oldest fixtures in
communications. It long ago disappeared from broadcast in favor of the
much more flexible pantograph assembly that allows the microphone to
be supported on a heavy, stable base with the microphone itself positioned conveniently in front of the user. The desk stand, as shown in
Figure 10–1, may still be found in a few boardrooms or paging systems,
but in the vast majority of transportation terminals the telephone handset
now takes the place of the desk stand.
The telephone handset itself has undergone some important modifications.
While the older carbon button transmitter (microphone) still has some
advantages in normal telephony, it is often replaced by an electret element
3:00 PM
Page 291
17: A Survey of Microphones in Broadcast and Communications
when the handset is to be interfaced with a paging or an announcing
system. Care should be taken to use a handset/interface which will allow
the system to mute before the handset has been placed back in its cradle,
thus avoiding the disagreeable “thunk” of hanging up.
In very noisy environments, including airplane cockpits, ship decks, and
heavy duty machinery rooms, a noise-canceling microphone is used, often
in conjunction with a hands-free headset-microphone combination. The
noise-canceling microphone is a gradient model designed for flat response
at very short working distances. Stated differently, the normal proximity
LF rise has been equalized for flat response relative to mid and HF, with
the result that distant pickup is rolled off at LF. A typical model is shown
in Figure 17–1A, and frequency response is shown at B. Close sound
sources enter primarily by way of the front opening, while distant sound
sources enter equally at front and back openings and are reduced 6 dB per
halving of frequency below about 1 kHz. In the figure, the crosshatching
shown at B indicates the effective range of noise cancellation. In order to
gain this degree of effectiveness it is essential that the microphone be positioned virtually next to the talker’s mouth. Such microphones are carefully designed with sufficient screening and mesh to attenuate close-in
breath sounds.
Positioning microphones very close to wall or floor boundaries has long
been a general practice among knowledgeable engineers, but it was
Crown International who, during the 1970s, introduced a line of microphones optimized for the purpose. These were known by the term PZM
(pressure zone microphone), indicating that these microphones responded
only to signal pressure components that were present at the boundary.
Boundary layer microphones are available from many manufacturers.
While the early models were primarily omnidirectional, modern versions
may have cardioid and hypercardioid patterns. When a directional pattern is used, its axis is parallel to the boundary layer and the microphone
senses pressure gradient components that are parallel to the boundary.
The BL microphone is often placed on a large room boundary or in the
center of a large baffle, and when mounted in this manner the BL microphone picks up sound with a minimum of reflections. These microphones
are ideal for placement on boardroom tables and altars, and have found
a permanent home on-stage in legitimate theaters and concert halls. A
typical omnidirectional model is shown in Figure 17–2A.
When placed on a supporting baffle, the larger the baffle the better the
LF response, as shown in Figure 17–2B. When the baffle diameter is large
with respect to wavelength, the microphone takes advantage of pressure
doubling at the boundary surface. At progressively lower frequencies
3:00 PM
Page 292
Photo of a noise-canceling
microphone (A); response
of the microphone to near
and far sound sources (B).
(Data courtesy of Crown
the boundary “unloads” and the microphone’s sensitivity drops by 6 dB
as it approaches a free-space operating condition.
For many applications an automatic microphone mixing and gating
system is used to limit the number of open microphones only to those
actually in use. (We discuss the electronic requirements for this in a later
section.) While conventional microphones are used in most of these
systems, Shure Incorporated has devised a mixing system in which the
microphones themselves can make a distinction in sound direction.
Figure 17–3 shows details of the Shure AMS26 microphone, which has
3:00 PM
Page 293
17: A Survey of Microphones in Broadcast and Communications
The boundary layer
microphone (A); response
depends on size of mounting
baffle (B). (Photo courtesy of
Crown International.)
Details of the Shure
AMS26 microphone. (Data
courtesy of Shure Inc.)
two elements, one forward-facing and the other rearward-facing. Only
the front element is used in direct speech pickup; the other element is used
only in a sensing circuit that allows the combination to reject random,
distant sounds, while gating on frontal speech sound. It is obvious that
the AMS type microphones must be used with AMS input circuitry for
proper system performance.
On convention floors and in other active areas covered by radio and television broadcast, the roving reporter always wears a headset-microphone
3:00 PM
Page 294
The headset-microphone
combination. (Photo
courtesy of AKG
combination and carries a handheld vocal microphone for interviews.
Three communications channels are in use here: two for the reporter and
one for the interviewee. A typical unit with an electret microphone is
shown in Figure 17–4.
In teleconferencing, two or more remote locations are connected via telephone lines so that communication, both audio and low data-rate video,
is enabled. Each of the remote spaces will need a center microphone that
is broadly omnidirectional in the horizontal plane but limited in vertical
coverage. Olson (1939) describes a microphone with a toroidal pattern,
which is uniformly receptive to sound in the horizontal plane, but which
has substantial rejection in the vertical direction. The microphone is of
the interference and has polar response that varies with frequency.
Cooper and Shiga (1972), as part of their research in quadraphonic
sound, describe a relatively simple microphone with a toroidal pattern,
as shown in Figure 17–5. The system consists of two figure-8 microphones oriented at 90 and placed one over the other. The two microphones are processed through all-pass networks with a 90 phase
relationship between them, as shown.
The pistol-shaped microphone shown in Figure 17–6 is used for field newsgathering activities. It can be quickly re-aimed as needed. The M-section
may be used alone for simple news gathering applications, while the
3:00 PM
Page 295
17: A Survey of Microphones in Broadcast and Communications
Conferencing microphone
with a toroidal directional
A handheld MS
microphone for news and
effects gathering. (Photo
courtesy of Neumann/
MS combination is very useful for sound effects recording, where stereo
remix in postproduction is often required.
The present-day tie-tack microphone grew out of the relatively heavy
lavalier microphone which hung around the user’s neck. Those were also
the days before wireless microphones, so the whole assembly was a
clumsy one, with the user tethered by a cable. The heavy lavalier was
also susceptible to noise pickup through contact with the wearer’s clothing. The modern tie-tack microphone is very small electret model and can
be clipped directly to the wearer’s lapel or tie, as shown in Figure 17–7A.
Details of the microphone itself are shown at B. The tie-tack microphone
5:56 PM
Page 296
The modern tie-tack
microphone: typical
application (A); photo of a
typical model (B); normal
equalization of a tie-tack
microphone (C). (Photo
courtesy of Crown
is ubiquitous, barely visible in just about every talk or news show on
television. It is also the microphone of choice for lecturing in those situations where the talker must have freedom of movement. The response
pattern is normally omni, inasmuch as the operating distance is very
short. For feedback-prone environments a cardioid version can be used.
The microphone’s response is normally internally equalized as shown at
C; the rolloff at MF compensates for the chest cavity resonances of the
talker, and the HF boost compensates for the microphone’s position far
off-axis of the talker. Obviously, if a tie-tack microphone with a directional pattern is used, the main axis of the pattern must point toward the
talker’s mouth.
Sports events on TV make extensive use of telephoto shots of on-field
action, and it is desirable to have a microphone that can produce a matching sonic effect. Some of the larger rifle and parabolic microphones discussed in Chapter 6 are of some limited use on the playing field, as are
certain arrays discussed in Chapter 19. The “crack” of a baseball bat is a
great effect if it can be made loud and sharp enough. In many cases it has
been picked up with a small wireless microphone located at home plate.
3:00 PM
Page 297
17: A Survey of Microphones in Broadcast and Communications
The announcer’s booth is constructed as shown in Figure 17–8A. The
space should be large enough for two persons so that interviews can be
done comfortably. The acoustical treatment should consist of absorptive
panels or other elements deep enough sufficiently to damp the lower
frequency range of a male announcer (the range down to about 125 Hz),
and the space should be free of any small room sonic coloration. Double
Radio announcer’s booth:
construction details (A);
microphone usage (B).
3:00 PM
Page 298
glazing should be used in the windows of the booth that look into the
control room and studio. The table in the center of the room should have
a well-damped surface, such as felt, to minimize scuffing noises, shuffling
of papers, and the like. We recommend that an acoustical consultant be
engaged for advice regarding construction, room treatment, and the type
of air ductwork necessary to minimize noise.
Details of microphone placement are shown at B. The engineer
should be aware that modern broadcast practice relies a great deal on
voice processing. Careful limiting, compression, and downward expansion are often used to maintain consistent levels and minimize noise.
Additionally, some degree of equalization may be applied to the voice.
Some announcers with “big, radio-type voices” may exhibit pronounced
vocal waveform asymmetry, which may cause a high signal crest factor
(see Chapter 2). The effect of this may be to restrict the maximum modulation level available for that announcer. Modern voice processors have
circuitry that can be engaged to minimize high signal crest factors without audibly changing the timbre of the voice, thus providing for better
overall on-air transmission.
Large conferences more often than not are presented in midsize meeting
rooms using a classroom setup. The basic requirement grows out of the
fact that all attendees are participants and may be called upon to address
the group as a whole. Often, several levels of priority are required. A
chairperson, of course, is in charge of the meeting’s proceedings.
Additionally, sub-chairpersons may be responsible for smaller delegations. Here is a general description of such a system:
Chairpersons use a special station that allows them priority of all
delegate stations of lower priority. Under normal conditions, when talking, the chairperson’s voice is heard at all stations. When comments from
the delegates are desired, the delegate engages his or her station; and this
mutes the local station’s loudspeaker so that feedback does not take
place. Comments are then heard over the rest of the stations. The maximum number of units that can be engaged at a given time can be set by
the chairperson, and the overall system gain is automatically adjusted
downward in order to inhibit feedback. All the microphones are cardioid
in pattern and are positioned, via the gooseneck, for fairly close placement to each delegate. In the system described here, up to 25 delegate
stations can be connected in series-chain fashion, and a single base control unit can accommodate up to four such chains. An extension unit
allows 75 additional delegate stations to be used. Figure 17–9A shows
a photo of a typical user station that contains microphone, local loudspeaker, and control functions. A typical block diagram for system layout
is shown in Figure 17–9B.
3:00 PM
Page 299
17: A Survey of Microphones in Broadcast and Communications
Conference systems: a
typical user station (A);
block diagram of typical
system (B). (Data courtesy
of AKG Acoustics.)
The modern airline terminal paging system is very complex. The telephone
handset (with electret transmitter) is universally used for live announcements. Local stations exist at each gate, and the range of that system is
limited, to the extent possible, to loudspeakers that specifically cover the
local gate area. Most messages are obviously made by gate ticketing personnel and relate to the flight presently scheduled. Local gate systems
operate best when they consist of a large number of small loudspeakers,
each operated at a fairly low level and positioned uniformly throughout
that gate area. This ensures minimum “spill” into adjacent gate areas.
Global paging is used for announcements that may affect a number
of substations at one time, and messages are normally recorded in a
“queue” system for replay a few seconds later. The necessity for the
queue is to allow for a number of such announcements to “stack up,” if
need be, to be replayed in the order in which they were placed. This is a
great convenience for the person making the page, since it ensures that
the attendant can make the page immediately, get it into the queue, and
then go on about their business.
At the highest level of priority are emergency pages or announcements, and these can override all local paging activities. Many such
announcements are prerecorded and may be actuated by fire or other
3:00 PM
Page 300
alarm systems. It is imperative that recorded announcements be made by
professional announcers and appropriately processed (peak limited and
spectrally shaped) for maximum intelligibility. In certain parts of the
world it is also essential that such announcements be made in a number
of languages.
The signal level of many paging systems is adjusted automatically and
continuously over a range of several decibels, depending on the ambient
noise level in various parts of the terminal as measured by sampling
microphones placed throughout the terminal.
Paging in factories or on the deck of a ship is difficult. Noise levels are often
so high that the excess levels over the speech range (250 Hz to 3 kHz)
required to override the noise may be uncomfortable to all concerned.
Intelligibility is actually reduced at such high levels. Today, key personnel
on duty in noisy areas are often provided with pagers and cellular telephones for use in emergency situations; feeds from these channels may be
routed directly, if needed, into audio communications channels.
3:01 PM
Page 301
It is rare to find any public venue today in which speech is not routinely
reinforced. Today’s auditoriums, worship spaces, classrooms, and arenas
tend to be larger than in the past, and patrons have come to expect the
comfort and ease of listening with minimum effort. In this chapter we will
present an overview of sound reinforcement system design, with emphasis
on those factors that determine the intelligibility of reinforced speech.
Music reinforcement spans the gamut from small-scale club systems
to high-level reinforcement of rock and pop artists in major venues, both
indoors and out. For some years, the emphasis in professional music
reinforcement has been on basic sound quality, providing an impetus to
both loudspeaker and microphone product development. We will begin
our discussion with speech reinforcement.
The basic requirements for a speech reinforcement system are that it
Adequate and uniform speech signal levels for all listeners
Adequate speech intelligibility for all listeners
Natural speech quality for all listeners
Stable performance under all operating conditions
Before we discuss these requirements, let us look at the development
of mass speech communication over the centuries:
In Figure 18–1A, a talker is addressing a group of listeners outdoors
and must rely solely on the natural speech volume level to ensure that all
listeners can hear. If the audience is a large one, those listeners in the
3:01 PM
Page 302
The progress of assisted
speech reinforcement over
the centuries.
front will receive higher levels than necessary, while those at the back
will have to strain to hear and understand the talker. Outdoor noises
cannot be avoided.
In Figure 18–1B, the talker is elevated on a hillside. From this position the sound coverage of the audience will be more uniform, and more
acoustical power from the talker will reach the audience.
In Figure 18–1C, the audience has been placed on a terraced area of
an amphitheater, further isolating it from outdoor disturbances.
In Figure 18–1D, both audience and talker have moved indoors. The
freedom from outside interference, along with the early reflections from the
side walls and ceiling, will increase both sound levels and intelligibility.
The final step, shown in Figure 18–1E, is to improve the directivity
of the talker with respect to the audience by adding a reflector behind the
talker. This increases the level of the talker at mid-frequencies, further
aiding intelligibility at a distance.
At each step, there has been an improvement in at least one of the
following aspects: speech level, uniformity of coverage and minimizing
outside disturbances. Until recent years, the development shown in
Figure 18–1E was deemed sufficient to take care of virtually all speech
requirements in small- to moderate-size lecture rooms.
The analysis given here is based on the work of Boner and Boner (1969).
Figure 18–2 shows the primary elements in a modern speech
reinforcement system. There are four basic elements: talker, microphone,
3:01 PM
Page 303
18: Fundamentals of Speech and Music Reinforcement
Basic elements of an
outdoor speech
reinforcement system.
loudspeaker, and listener; these are separated by the distances DS, D0, D1
and D2, as shown.
Assume that the unaided talker produces an average speech level of
65 dB LP at the microphone, 1 m (40 in) away. By inverse square law, the
level at the listener will be 26 dB lower, or 39 dB LP (see Chapter 2 under
Acoustical Power). Both the loudspeaker and microphone are assumed
to be omnidirectional.
When the system is turned on, the electrical gain of the microphone/loudspeaker combination can be increased until the loudspeaker
provides a level from the microphone equal to that of the talker, or
65 dB. This condition produces unity gain through the system, and this
results in acoustical feedback through the system as shown in
Figure 18–3. Feedback causes the familiar “howling” effect that we have
all heard with improperly controlled sound reinforcement systems.
To have a workable system it will be necessary to reduce the system
gain by about 6 dB, which should result in an acceptable stability margin.
When this adjustment has been made, the loudspeaker will now produce
a signal level at 10 m (33 ft) (distance from the loudspeaker to the microphone) of 59 dB.
Now, we determine what level the loudspeaker will produce at the
listener at a distance of 15 m (50 ft). Again, by inverse square law, we
calculate that level to be 55.5 dB.
The acoustical gain of the system is defined as the difference in level
at the listener with the system off compared to the level at the listener
with the system on:
Acoustical gain 55.5 dB 39db 16.5 dB.
A general equation for the potential acoustic gain (PAG) in dB that
an outdoor system can produce is:
PAG 20 log D1 20 log D0 20 log Ds 20 log D2 6
Later, we will see how the use of more directional microphones and
loudspeakers can improve the maximum acoustical gain of a system.
Another graphical way of looking at the acoustical gain of the system is shown in Figure 18–4. We may think of the system as effectively
3:01 PM
Page 304
The origin of acoustical
The concept of EAD.
moving the listener closer to the talker. When the system is off, the
distance is of course D0, or 20 m (66 ft). When the system is turned on,
the talker has in a sense “moved forward” toward the listener. This “new
listening distance” is known as the equivalent acoustic distance (EAD).
We can calculate EAD by considering the following two facts: the talker
produces a level of 65 dB at a distance of 1 m (40 in), and with the system
turned on the listener hears the talker at a level of 55.5 dB.
We now ask the question: at what distance from the talker will the level
have dropped to 55.5 dB? The level difference is 65 55.5, or 9.5 dB.
Using the nomograph shown in Figure 2–7 we can see directly that level
attenuation of 9.5 dB corresponds to a distance of about 3 m (10 ft),
relative to a reference distance of 1 m (40 in). Therefore, with the system
turned on, the listener will hear the talker as if that talker were located
at a distance of 3 m (10 ft) from the listener.
3:01 PM
Page 305
18: Fundamentals of Speech and Music Reinforcement
When we move the speech reinforcement system indoors its analysis
becomes a bit more complicated. The inverse square attenuation with
distance that we observed outdoors has now been replaced with a combination of inverse square loss and indoor reverberant level produced by
room reflections.
An indoor system is shown in Figure 18–5A. We can see the inverse
square paths to the listener, along with the reflected contributions from
the room. When we look at the total contribution of both direct and
reflected sound, the picture is as shown at B. Close to the source, the
attenuation with distance follows the inverse square relationship; however, at some given distance, the reverberant field begins to dominate.
The reverberant field is fairly uniform throughout the room, and at
distances far from the loudspeaker the reverberant field is dominant. As
we discussed in Chapter 2 (under The Reverberant Field) the distance
from the loudspeaker at which both direct and reverberant fields are
An indoor speech
reinforcement system:
system layout (A);
attenuation of sound with
distance in a reverberant
environment (B).
3:01 PM
Page 306
equal is known as critical distance (DC). At DC, the level is 3 dB higher
than either of its components. In this example we are assuming that
both talker and microphone are in the reverberant field produced by the
A complete analysis of the PAG of an indoor system is fairly complex, but it basically takes the D terms in equation (18.1) and converts
them to limiting values of critical distance. When this is done, two of the
log terms in the equation cancel, and the net result is:
PAG 20 log DCT 20 log Ds 6 dB
where DCT is the critical distance of unaided talker in the direction of the
listener, and DS is the talker-to-microphone distance.
The critical distance may be calculated by the equation:
DC 0.14QR (meters or feet)
where Q is the directivity factor of the sound source and R is the room
constant in the enclosed space. The room constant is:
(square units)
where S is the surface area in the space and is the average absorption
coefficient in the space. (Note: the value of R is expressed in either square
meters or square feet, depending on the system of units being used in the
When we examine equation (18.2), we can see that the first term is
dependent on the room characteristics and cannot easily be changed.
Thus, the second term is the only one that we can easily change. This is
intuitively clear; the best (and easiest) way to increase the potential gain
of an indoor system is simply to move the microphone closer to the
talker. In addition, the use of a directional microphone and employment
of loudspeakers that aim the bulk of their output directly at the fairly
absorptive audience will also improve system gain.
In an indoor environment there are three factors that largely determine
how intelligible transmitted speech will be:
1. Signal-to-noise ratio (dB). This is the ratio of average speech levels
to the A-weighted local noise level. For best intelligibility, a signalto-noise ratio of 25 dB or greater is recommended.
2. Room reverberation time (s). When the reverberation time exceeds
about 1.5 s, the overhang of successive syllables will have a detrimental effect on intelligibility. Strong room echoes in particular will
have a deleterious effect on intelligibility.
6:00 PM
Page 307
18: Fundamentals of Speech and Music Reinforcement
3. Direct-to-reverberant (D/R) speech level (dB). When this ratio is
less than about 10 dB, the reverberant level tends to mask speech in
much the same way that random noise does.
In a later section we will discuss some of the methods by which the
intelligibility of a speech reinforcement system may be estimated while
system layout is still at the design stage.
Figures 18–6 and 18–7 show some of the microphone types that are
used in speech reinforcement. The hand-held vocal microphone is used
as shown at Figure 18–6A. Vocal microphones are normally designed so
that they produce fairly flat response when positioned about 5 to 10 cm
(2 to 4 in) from the performer’s mouth. (Typical response of a vocal
microphone is shown in Figure 7–2.) At such small operating distances,
the vocal microphone is fairly immune to feedback – a classic example
of reducing DS to a very low value. The best microphones for vocal use
are those that have integral multiple screening surrounding the capsule
to minimize the effects of inadvertent puffs of wind from the talker.
Many of these microphones have a pronounced “presence peak” in the
3–5 kHz range for added brightness and improvement of articulation.
Many performers feel very much at home with a vocal microphone in
hand – and they often feel at a loss without it. Proper microphone etiquette
must be learned; never blow on the microphone to see if it is on; always
Microphones for speech
reinforcement; the
handheld vocal microphone
(A); head-worn microphone
(B). (Photo courtesy of
AKG Acoustics.)
3:01 PM
Page 308
Microphones for speech
reinforcement: podium
microphone (A); podium
usage (B); boundary layer
usage on an altar (C).
(Photo courtesy of Crown
hold it slightly to the side, outside the breath stream, and maintain
a consistent operating distance.
The head-worn microphone (shown at B) has long been a staple in
communications activities, but it was not until its adoption by singer
Garth Brooks that it became a staple for on-stage performance. The microphone element is an electret, normally with a cardioid or hypercardioid
pattern, and equalized for close use. When properly worn it is stable in its
positioning and offers excellent performance. It is invariably used with
wireless bodypacks, and for lecturers or panelists it provides complete
freedom of movement. Caution: it must be properly fitted and not be
allowed to slip into position in the breath stream of the talker or singer.
3:01 PM
Page 309
18: Fundamentals of Speech and Music Reinforcement
For permanent podium or lectern mounting there are numerous
miniature electret cardioid or hypercardioid models mounted on flexible
gooseneck extensions, as shown at Figure 18–7A and B. These can be
unobtrusively located to one side and positioned at a distance of
30–50 cm (12–20 in) from the mouth of the talker. A small windscreen
is recommended. It is important that the gooseneck portion be out of the
range of movement of papers or notes used by the talker and that the
talker’s normal motions not be impeded.
For use on flat surfaces, such as tables used in panel discussions, or
altars in houses of worship, a boundary layer microphone (shown at C)
is essential. An omni pattern often works best, but cardioid models may
be necessary to minimize local noises. The cardioid, however, will be
3:01 PM
Page 310
more subject to impact noises than the omni. The operating distance is
normally in the range of 45–60 cm (18–24 in).
The tie-tack microphone, which was introduced in the previous
chapter, has the advantages of a small electret and is very popular
primarily because it is inconspicuous. It is important that it be fastened
to the user’s lapel or tie with enough slack in the cable to avoid pulling
or tugging on the cable as the performer moves around. For sedentary
applications the microphone may be wired directly, but for most
purposes it is used with a wireless bodypack. Position the microphone as
high as possible on the tie or lapel; however, be aware that in a high position normal up and down head movements may cause audible shifts in
level. Make the right compromise. The tie-tack microphone’s response is
normally shaped as shown to minimize radiation from the chest cavity
and to maintain HF response.
When microphones are positioned close to reflecting surfaces, the
delayed reflection may combine with the direct sound, producing some
degree of comb filtering in response. Figure 18–8A shows an omni
Delayed reflections; an
omni microphone in the
path of a reflection (A); a
hypercardioid microphone
with the reflection reduced
by the off-axis attenuation
of the pickup pattern (B).
3:01 PM
Page 311
18: Fundamentals of Speech and Music Reinforcement
Cancellations due to
multiple microphones;
improper implementation
(A); proper
implementation (B).
microphone mounted on a podium in such a way that it will pick up
a distinct reflection from the reading desk, producing uneven response.
Moving the microphone to one side will alleviate this problem to some
degree. A better solution is to use a hypercardioid microphone, whose
off-axis response will attenuate the reflection, as shown at B.
Another common problem is the use of two microphones where one
is sufficient. Improper usage is shown in Figure 18–9A. In broadcasting
of important events, doubling of microphones is often done for transmission redundancy in case of failure of one channel, but more often
than not both microphones end up operating in parallel. For a talker
directly between the two there may be no problem. But talkers do move
around, and the combined signals from both microphones will cause
peaks and dips in response as shown. The solution is shown at B, where
both microphones are mounted in coincident fashion and splayed
slightly to increase the effective pickup angle. In this case the position of
the talker will not be a problem since the talker’s distance to the two
microphones remains the same.
The basic loudspeaker approach for a venue depends on many things,
but the reverberation time is paramount. Consider the space shown in
Figure 18–10.
If the reverberation time is less than about 1.5 s, a single central
loudspeaker array is usually the best solution. It has the advantage of
single point radiation and will sound quite natural to all patrons. It must
of course be designed so that it covers the audience seating area smoothly
3:01 PM
Page 312
FIGURE 18–10
Speech reinforcement
in large spaces: low
reverberation time
(A); moderate
reverberation time (B);
high reverberation
time (C).
and minimizes sound projected onto the walls. The design approach is
shown at A.
If the same space has a reverberation time between about 1.5 and
3 s, then it may be more effective to supplement the central array with
additional side wall arrays at about halfway the length of the space. The
primary array is then adjusted so that it covers the front of the space, and
the secondary arrays are delayed so that the progressive wavefronts from
the main and secondary arrays will be effectively “in-step” at the back
of the space. The primary aim is to increase the D/R ratio in the back half
of the space. The design approach is shown at B.
Finally, if the same space has a reverberation time exceeding about
3 s, a conventional distributed system may be required. Here, there is no
central array as such, but rather a series of smaller loudspeakers located
3:01 PM
Page 313
18: Fundamentals of Speech and Music Reinforcement
on the side walls, sequentially delayed in order to produce a natural
effect. In this approach, shown at C, each listener is relatively close to a
loudspeaker and will thus benefit from the increase in D/R ratio. In many
venues, these side mounted loudspeakers will be simple vertical column
models, which have broad horizontal response (to cover the audience
effectively) and narrow vertical response (to avoid excess signal “spill”
onto the walls). Some installations include an undelayed target loudspeaker located at the front of the room which helps to define the actual
sound source at that natural position.
As you can see, the design choices must be carefully made, and at
each step in the design process a new analysis of the system’s intelligibility estimate must be made. Very large venues such as sports arenas and
stadiums are obviously more difficult to design and analyze.
Signal flow diagrams for the three systems of Figure 18–10 are
shown in Figure 18–11. As the number of delay channels increases, the
target coverage areas of loudspeakers in each delay zone are restricted to
a specific portion of the seating area. The aim here is to provide coverage
FIGURE 18–11
Signal flow diagrams for
the systems shown in
Figure 18–10.
3:01 PM
Page 314
where it is needed for maximizing the D/R ratio. Ideally, the “spill”
outside each area should be minimal, but in practice it is difficult to
The usual cause of acoustical feedback, even in properly designed
systems, is the careless raising of microphone input levels to “reach” for
a weak talker. Another reason could be that too many microphones are
open at the same time. When a skilled operator is at the controls, things
rarely go wrong. But many systems are operated by amateurs or volunteers who do not understand the basic nature of feedback. In order to
make systems failsafe in the hands of such operators, manufacturers have
come up with a number of electronic devices that inhibit feedback to a
greater or lesser degree. We discuss some of them below.
The frequency shifter was developed during the 1960s as a means of
minimizing feedback by shifting the amplified sound up or down in
frequency by about 4–6 Hz. As such, the effect was not easily noticed on
speech. Music was however another matter; even slight frequency shifts
are audible as a slow beating effect, especially in sustained passages.
When the frequency is shifted, it is difficult for the conditions necessary
for feedback to become established; however, excessive gain will result in
a time-varying “chirping” effect as the system tries unsuccessfully to
reach steady feedback. A signal flow diagram of an early frequency
shifter is shown in Figure 18–12. Today, the technique is more sophisticated, consisting of a slow, random frequency shifting action rather than
a fixed relationship.
In the hands of properly trained acousticians the insertion of narrowband notch filters into the audio chain can minimize feedback, resulting
in improvements of up to 4 to 6 dB in overall gain capability (Boner and
Boner, 1966). The technique involves driving the system slowly into feedback, determining the frequency of feedback, and inserting a narrowband filter at the feedback frequency. This procedure is done sequentially
for the first three or four feedback modes of the system; beyond that
FIGURE 18–12
Details of a frequency
3:01 PM
Page 315
18: Fundamentals of Speech and Music Reinforcement
point there will be diminishing returns. The method is now normally carried out using parametric equalizer sections, as shown in Figure 18–13.
Based on digital signal processing, sophisticated methods of signal analysis and treatment can produce systems that can detect the presence of
sustained feedback, determine the feedback frequency, and automatically
engage the necessary filtering to control the feedback. A simplified signal
flow diagram is shown in Figure 18–14.
Automatic mixers are widely used in many systems employing a number
of microphones. For example, a house of worship may have microphones present on the lectern, pulpit, baptistry, and altar. It is clear that
only one of these microphones will be in use at a given time. Gating the
microphones on and off does not necessarily require an operator; using
a properly adjusted automatic microphone mixer, the control of the
microphones can be smooth and foolproof. Figure 18–15 shows a basic
signal flow diagram for an automatic mixer.
In many applications it will be necessary for more than one
microphone to be open simultaneously, and under this condition there is
a function in the mixer that will attenuate system gain according to the
Gain reduction 10 log NOM
FIGURE 18–13
Narrow-band equalization.
FIGURE 18–14
A feedback eliminator.
3:01 PM
Page 316
FIGURE 18–15
Basic signal flow
diagram for an automatic
microphone mixer.
where NOM is the number of open microphones. The desired gain
reduction amounts to 3 dB for each doubling of open microphones and
will result in a uniform reinforced level in the space.
A well-designed automatic mixer will also provide some immunity
against false triggering of gating functions due to variations in the ambient noise level of the space.
Where a speech reinforcement system has been installed in a venue, the
effective intelligibility of that system can be determined directly by syllabic testing using standardized techniques. Essentially, a talker stands at
the microphone and reads a set of random syllables, all embedded in a
“carrier” sentence. An example of this is: “Please write down the word
cat;” now I want you to write down the word man.” And so it goes. The
purpose of the carrier sentence is to present the test syllables within the
acoustical masking context of continuous speech. The results for various
listeners in various parts of the auditorium are then analyzed and the
accuracy of their responses expressed as a percentage. If a listener gets
a score of 85% on random syllabic testing, then that listener will likely
understand about 97% of normal speech in that space.
If a speech reinforcement system is still on the design drawing board,
its effectiveness may be broadly estimated, based on certain acoustical
parameters that the supervising acoustician can arrive at. For example,
the acoustician can estimate the normal direct speech sound levels at
a typical listener that will be produced by the targeted loudspeakers. The
acoustician can also determine, based on reverberation analysis, what
the reverberant level at the listener will be, as well as the reverberation
time itself. The acoustician can also make a reasonable estimate of the
3:01 PM
Page 317
18: Fundamentals of Speech and Music Reinforcement
noise level likely to be encountered in the space, based on the intended
degree of noise isolation and elimination.
If all of these estimates can be made for the octave frequency band
centered at 2 kHz, and if it can be safely assumed that there are no
unusual reflections in the room’s reverberant decay pattern, then an
estimate of the system’s articulation loss of consonants (%Alcons) can be
made using the following set of equations:
%Alcons 100 (102(A BCABC) 0.015)
A 0.32 log
10 EE E N
B 0.32 log
T12 C 0.5 log
ER 10LR / 10
ED 10LD / 10
EN 10LN / 10
As an example, assume that our preliminary room and system simulations
give the following values: T60 4 seconds, LR 70 dB, LD 65 dB and
LN 25 dB. Calculating the values of A, B and C:
A 0.036
B 1.76
C 0.24
Entering these values into equation (18.5a), gives %Alcons 14%.
Figure 18–16 indicates a subjective assessment of this system’s
anticipated performance as on the borderline between adequate and poor.
FIGURE 18–16
Subjective descriptions of
%Alcons values.
3:02 PM
Page 318
Based on this observation, the acoustician and architect may find common ground between them for improving the estimate. Most notably, a
reduction of reverberation time (difficult, and often costly, to accomplish)
or an improvement in loudspeaker coverage for higher direct sound levels
(easier to accomplish) might be considered.
Traditionally, legitimate theaters and playhouses have been fairly small
spaces, and professional actors have long been known for their ability to
fill a house without electroacoustical assistance. This is still the model in
most established venues. However, modern venues tend to be larger than
earlier ones, and the proliferation of multi-purpose halls can result in
staging of drama in houses that are actually too large for the purpose. In
these cases a speech reinforcement strategy as shown in Figure 18–17 is
very effective. Here, three boundary layer microphones are placed on the
apron of the stage as shown. Ideally, these microphones should be in the
direct field of the actors, but in some large houses the microphones may
be in the transition region between direct and reverberant fields.
The signals are routed to carefully equalized and delayed loudspeaker channels located in the proscenium. In many cases, the acoustical gain of the system at mid-frequencies may be virtually none; the
important effect of the system is the amplification of high frequencies,
which enhances intelligibility. The purpose of the delay is to ensure that
the first-arrival sound at the listeners is from the stage.
If omni microphones can provide the requisite gain at high frequencies, they are the first choice. If cardioid microphones are needed for
FIGURE 18–17
Stereophonic reinforcement
of speech in the legitimate
theater: stage layout (A);
relationship (B); typical
electrical channel (C).
3:02 PM
Page 319
18: Fundamentals of Speech and Music Reinforcement
feedback minimization, be aware that footfalls on the stage may become
a problem, which will require high-pass filtering.
Special effects and off-stage events are normally handled by a sound
designer and may be fed into the house from a variety of directions.
Since the rock musicals of the 1960s, sound reinforcement has become a
requirement in the musical theater, both for pit musicians and singer–
actors on stage. Today, hardly a touring company goes on the road without a sound specialist and all the necessary equipment to do the job right.
Here are some of the microphone requirements:
1. On-stage: Singers wear wireless tie-tack microphones, usually
positioned at the middle of their hairlines. This affords complete
freedom of movement and expression. See the notes in Chapter 9
regarding the use of large numbers of wireless microphones.
2. Overhead pickup: Rifle microphones may be flown to pick up some
degree of stage ambience or group action.
3. Orchestra pit: The pit orchestra has always been limited in personnel,
and individual clip-on microphones may be used on each instrument.
These signal outputs must be pre-mixed to give the front-of-house
(FOH) mixer the necessary flexibility in handling the overall production. Assistant mixers may be needed in complex situations.
Loudspeakers are normally deployed in vertical columns at the sides
of the stage, with other loudspeakers positioned in the house as necessary for special effects.
The mega-event concert performance of a major pop/rock artist is usually
presented in a large indoor arena, outdoor stadium, or even an expansive
field. Patrons pay dearly for front seats, and even those who are seated at
great distances will expect sound pressure levels in the 110–115 dB range.
Most of the instruments have direct feeds to the console, and only
vocals, drums and an occasional wind instrument will actually require
microphones. All microphones are positioned very close to the sound
sources to minimize feedback since on-stage monitoring levels can be
very high. Vocalists–instrumentalists in particular have gravitated to
head-worn microphones, since these give freedom of movement as well
as a very short and consistent operating distance for vocal pickup. A solo
vocalist may still prefer a handheld microphone since this has long been
a standard “prop” onstage.
A potential source of feedback is from stage floor monitor loudspeakers operating at high levels into the performer’s microphones, even
3:02 PM
Page 320
when those microphones are operated at very short distances. As a hedge
to this, a new technique known as in-the-ear monitoring has become
popular. It uses small receivers that are actually placed in the wearer’s ear
canal, with operating levels carefully monitored.
The field of active acoustics has been around for at least three decades and
has made steady but gradual progress. One of the earliest systems was
assisted resonance (Parkin, 1975). The first notable installation was in
London’s Royal Festival Hall. The hall lacked sufficient volume for its seating capacity and target reverberation time, and the system was installed to
increase reverberation time in the range 60–700 Hz. The system consists of
172 channels. Each channel contains a microphone placed in a Helmholtz
resonator, amplifier, and a loudspeaker. The microphones are positioned
well in the reverberant field. Because of their narrow frequency bands the
individual channels are quite stable, and the increase in apparent reverberation time results from the high-Q nature of each resonator. Figure 18–18A
shows a view of a single channel, while the increase in reverberation time,
with the system on and off, is shown at B.
LARES stands for Lexicon Acoustical Reverberance Enhancement
System. The system consists of a set of Lexicon reverberation channels
randomly feeding an ensemble of loudspeakers. Microphones are located
above and in front of the performance area. The increase in reverberation time is the result of the reverberation generators themselves, and the
gain of the systems is stabilized by randomly varying delay sections in
FIGURE 18–18
Parkin’s assisted resonance.
3:02 PM
Page 321
18: Fundamentals of Speech and Music Reinforcement
FIGURE 18–19
The LARES system.
FIGURE 18–20
Delta stereophony.
each reverberation channel. Reverberation time is independently
adjustable and is not dependent on reverberant level. A signal flow diagram for the system is shown in Figure 18–19.
Delta stereophony, shown in Figure 18–20, is used to increase loudness of stage events without compromising the natural directional cues
from the stage. The system includes loudspeakers positioned at stage
level as well as overhead. The number of delay channels normally does
not exceed about six or eight, and the system is easily reconfigured for
various kinds of stage events.
3:04 PM
Page 322
A microphone array may be considered to be a set of microphone
elements arrayed in space whose outputs are individually processed and
summed to produce a given output. In a basic sense, an early cardioid
microphone created by summing separate pressure and gradient elements
may be thought of as an array, but this is not what we have in mind. The
arrays and techniques discussed here normally use multiple elements to
achieve a target directional response that may be specific for a given
application. In other cases, signal processing may be applied to single
microphones, or microphone channels, to enhance performance in some
unique respect.
Adaptive systems are those whose signal processing coefficients
change over time in order to accommodate or maintain a given signal control parameter. Basic adaptive techniques will be discussed as they apply to
echo cancellation and as they are used to simulate, via data reduction
techniques, the output of microphones in recording applications. Other
applications of adaptive techniques deal with the manipulation of array
directivity as a function of source tracking or other requirement. The
discussion begins with basic line array theory.
The simplest line array consists of a group of equally spaced microphones whose outputs are summed directly. Let us assume a set of
four omnidirectional microphones arrayed vertically, as shown in
Figure 19–1. Let the spacing, d, between adjacent line elements be 0.1 m
3:04 PM
Page 323
19: Overview of Microphone arrays and Adaptive Systems
(4 in). The far-field directivity function R() is given by the following
R() sin (1/2 Nkd sin )
N sin (1/2 kd sin )
where N 4 and d 0.1. The measurement angle is given in radians,
and k 2f/c, where c is the speed of sound in m/s.
Figure 19–2 shows directivity plots in the vertical plane for the
four-element array at frequencies of 400 Hz, 700 Hz, 1 kHz and 2 kHz.
A plot of the directivity factor of the array is shown in Figure 19–3 by
A four-element in-line array
of omnidirectional
Far-field directional
response for a four-element
array of omnidirectional
microphones with interelement spacing of 0.1 m:
400 Hz (A); 700 Hz (B);
1 kHz (C); 2 kHz (D).
3:04 PM
Page 324
Directivity factor for
in-line arrays of 4, 6, 8 and
10 elements.
Extending the frequency
range of uniform coverage.
the curve marked N 4. The value of d/ 1 along the baseline of the
graph corresponds to a frequency of 3340 Hz.
The useful range of the array extends up to a value of d/ equal to
about 1. Above that point the directional pattern shows the development of numerous off-axis lobes, although the on-axis directivity factor
remains fairly uniform. The useful range of the array can be extended
by shortening it with rising frequency. This involves having numerous
elements and crossing over from one wider-spaced set to another with
smaller spacing with increasing frequency. With careful attention to
matters of frequency division, response as shown in Figure 19–4 is
The frequency range of a directional array using logarithmically spaced
elements is more efficient in the use of a fixed number of transducers
than an array consisting of equally spaced elements. Van der Wal et al.
3:04 PM
Page 325
19: Overview of Microphone arrays and Adaptive Systems
A logarithmically spaced
line array of microphones
(Van der Wal et al.,
Polar response estimates
for the array shown in
Figure 19–5 (Van der Wal
et al., 1996).
(1996) propose an array as shown in Figure 19–5. The spacing of
elements is more dense toward the center of the array, and this allows
a decreasing array size by filtering with rising frequency such that the
number of active transducers is optimized in all frequency ranges. The
directivity of the array is shown in Figure 19–6. Here, the measured
directivity over the four-octave range from 500 Hz to 8 kHz shows a
remarkably uniform central lobe, with minor lobes largely in the 20 dB
Mahieux et al. (1996) describe an array consisting of 11 elements placed
in a slight arc above the display screen of a multimedia workstation. The
operating distance from the microphones to the operator is about 0.7 m,
well in the near field. The purpose of the array is to pick up the voice
of the operator with a fairly uniform directivity index of 10 dB over
the frequency range from about 500 Hz to 8 kHz. The elements of the
array are grouped into four subarrays as shown in Figure 19–7. Through
appropriate filtering into four frequency bands a net directivity index is
produced as shown in Figure 19–8.
3:04 PM
Page 326
Top view of line array
built into a multimedia
workstation (Mahieux
et al., 1996).
Directivity index of
the array shown in
Figure 19–7 (Mahieux
et al., 1996).
The German Microtech Gefell model KEM970 microphone is an example of a fixed vertical array that produces a broad lateral cardioid pickup
pattern and a 30 (6 dB down) vertical pattern over the frequency range
3:05 PM
Page 327
19: Overview of Microphone arrays and Adaptive Systems
Microtech Gefell KEM970
microphone. (Photo
courtesy of Cable Tek
Electronics Ltd.)
FIGURE 19–10
Microtech Gefell KEM970
microphone polar data:
horizontal polar family (A);
vertical polar family (B);
typical directivity balloon
(C). (Data courtesy of
Cable Tek Electronics Ltd.)
above 800 Hz, resulting in a directivity index of about 10 dB at MF and
HF. The length of the array is 0.35 m (14 in). Figure 19–9 shows a side
view of the microphone, and Figures 19–10A and B show families of
polar response. A typical “directivity balloon” for the microphone is
shown at C.
3:05 PM
Page 328
The Audio-Technica model AT895 microphone is shown in Figure 19–11,
and a view of its elements is shown in Figure 19–12. The center element
is a short-section line microphone of the type discussed in Chapter 6, and
the four capsules at the base are cardioids. In normal operation, opposite
pairs of cardioids are subtracted to produce a pair of figure-8 patterns at
MF and LF which are oriented 90 to each other. These resulting patterns
are perpendicular to the line element and thus do not pick up sound along
the main axis of the microphone. They do of course pick up sound at
MF and LF arriving at right angles to the line element. The signal flow
diagram of the system is shown in Figure 19–13.
The adaptive portion of the system, which is digital, derives signal
correlation values among the three outputs from the microphone elements and determines, on a continuing basis, the amount of interfering
sound level at MF and LF. The interfering signals are added to the output of the line element in reverse polarity so that cancellation takes place
at MF and LF. Much of the art in development of the system was in
determining the range of operation, both in level manipulation and frequency, and also the time constants necessary for smooth, unobtrusive
FIGURE 19–11
Photo of Audio-Technica
model AT895 adaptive
microphone. (Photo courtesy
of Audio-Technica US, Inc.)
FIGURE 19–12
Positions of microphone
elements. (Figure courtesy
of Audio-Technica US, Inc.)
3:05 PM
Page 329
19: Overview of Microphone arrays and Adaptive Systems
FIGURE 19–13
Signal flow diagram
for AT895 adaptive
microphone. (Figure after
Audio-Technica US, Inc.)
There are three basic modes of system operation:
1. All elements operating. This produces maximum rejection of interfering signals.
2. Line element plus one figure-8 set. This produces rejection only in
one pickup plane; useful under conditions where interfering sources
are largely confined to the ground plane.
3. Line element operating alone, with optimum filtering.
The microphone is intended primarily for activities such as news
gathering and sports events, where interfering sounds may be constantly
changing in directional bearing.
Modern conference systems allow users in one location to communicate
easily with another group at some distance. Duplex operation of the
systems permits conversation to proceed freely in both directions and is
normally a basic requirements of these systems. A fundamental problem
exists in a simple system, such as that shown in Figure 19–14. If talkers
in one location are communicating with a distant location, they will hear
their own voices delayed by the cumulative characteristics of the transmission path as the signal passes through the loudspeaker-microphone
pair at the receiving end and is returned to the sending location.
These “echoes” may occur in a fraction of a second – or they may,
as in the case of satellite links, be in the range of a second or longer. In
addition, the return signal will exhibit noise as well as the acoustic
3:05 PM
Page 330
FIGURE 19–14
A basic duplex (two-way) conferencing system: signals originating at the sending location
are reproduced at the receiving location via a loudspeaker; the return signal heard by the
original talker will be delayed by the transmission path and further contaminated by
room reflections at the receiving end.
FIGURE 19–15
Adaptive filtering can be used to cancel the “echo” of the original signal from the
sending end; the adaptive filter system intercepts the original signal and compares it with
the return signal; any differences are forced to zero via a negative feedback path, and the
signal returned to the sender is thus free of the echo.
signature of the receiving location. There may be equalization changes
due to a variety of other transmission effects and, taken as a group, these
effects will be quite disturbing to anyone attempting to communicate
over the system.
Adaptive filtering can be used to alleviate these problems, and
operation of the filter is shown in Figure 19–15. Because it is a generic
filter, it can provide delay functions along with equalization. The filter is
located at the sending end of the system and actually “models” the downstream transmission and return paths, comparing their output with the
original signal from the sending location. The filter is placed in a negative
feedback loop that actively “seeks” to reduce the difference between the
original and received signals to zero, thus cancelling the echo.
In modern systems, the initialization of the system is an automatic
process and usually takes place in a matter of a second or so after the
3:05 PM
Page 331
19: Overview of Microphone arrays and Adaptive Systems
FIGURE 19–16
The original signal is
returned with delay (echo)
and further acoustical
contamination; the
adaptive filter cancels these
effects to a large degree.
(Data presentation after
system is turned on. Once initialized, the system remains fairly stationary; updating itself only when there is a substantial change in the overall
signal path, such as moving the microphone, opening or closing doors,
and the like, at the receiving end.
Figure 19–16 shows typical action of the adaptive filter. The originating signal is modeled as an impulse at A, and the cumulative effects
of round-trip transmission are shown at B. The cancelled signal is shown
at C. Note that the direct-through path from the microphone at the
receiving end is outside the filter’s feedback loop and is fed directly back
to the loudspeaker at the sending end. Thus, the communication path
remains open at both ends at all times.
Modern systems are quite stable, provide more than adequate gain,
and can handle a number of microphones through automatic gating.
Woolley (2000) provides an excellent overview of the subject.
Kyriakakis et al. (2000, 2002) describe the concept of a virtual microphone
in the context of a classical orchestral recording in which the outputs of
selected microphones may be encoded at a low bit rate and reconstructed
later as needed. Figure 19–17 shows a partial layout for an orchestral
recording session; there are two main microphones (an ORTF pair) and a
single accent microphone for timpani pickup. (In actuality there would be
many more microphones, but for our discussion here we will consider
only the main pair and a single accent microphone.)
We will record the primary ORTF microphone pair at the full data
rate, but the timpani microphone is to be recorded as a virtual microphone. The timpani microphone signal will be a subset of the left ORTF
3:05 PM
Page 332
FIGURE 19–17
Basic layout for a classical
recording session; only the
main microphone pair and
a single accent microphone
are shown for clarity.
FIGURE 19–18
A normal digital (PCM)
recording path shown at
the left in the figure; the
path shown at the right
stores the virtual timpani
track as a time-varying set
of adaptive equalization
coefficients, which can later
be convolved with the left
ORTF signal to produce a
reconstruction of the
original timpani signal.
channel signal. This subset consists of filter coefficients that define, on a
moment by moment basis, the output of the timpani microphone in
terms of the overall spectrum picked up by the left ORTF microphone.
The left ORTF channel has been chosen for this purpose since its pattern
and orientation contain substantial signal from the timpani.
On playback, the virtual output of the timpani microphone will be
recovered by time-varying equalization of the left ORTF signal via
an adaptive filter. The filtering generates low data rate coefficients which
are stored during the recording operation, resulting in a continuous
“re-equalization” of the left ORTF signal so that its spectrum is identical to that of the actual timpani microphone signal. This process is
shown in Figure 19–18.
3:05 PM
Page 333
19: Overview of Microphone arrays and Adaptive Systems
The effectiveness of the entire process depends largely on instantaneous
musical balances as they are picked up by the ensemble of all microphones.
If thresholds of the virtual signal are high enough in the main microphone
channels, then the recovery can be excellent. The signal recovered in this
manner can then be used for postproduction rebalancing of the program.
Even in this age of high data rate delivery media, the requirements
of extended multichannel systems will at some point strain the available
hardware resources. The virtual microphone principle offers many
attractive solutions: for example, large venue surround sound presentations can be outfitted with a large number of ambience channels, each
one corresponding to a microphone positioned in the original recording
venue at various distances, near and far, from the orchestra. If these
signals are recorded via virtual microphones the overall data rate can be
vastly reduced.
In this chapter we have covered only a few of the many areas where
microphone arrays and adaptive signal processing are useful. The interested reader will want to explore the following application areas:
1. Hearing aids: Taking advantage of binaural hearing, basic beamforming techniques can lock in on nearby sources, providing the
hearing-impaired user with stronger localization cues and better
speech intelligibility. Adaptive beam-forming can be used to alter the
directivity patterns at each ear, resulting in similar improvements.
2. High-directivity beamforming: As with loudspeakers, large arrays
of microphones can be designed for high directivity over a fairly
wide frequency range. If the arrays are adaptive they can be used
for tracking individual sources over large angular ranges in difficult
3. Blind deconvolution: In situations where multiple signals of unknown
source location are presented simultaneously, advanced techniques
can be used, within broad limits, to sort out the signals.
An excellent starting point for further studies is the book
Microphone Arrays (Brandstein and Ward, 2001).
3:05 PM
Page 334
There is relatively little that sound engineers and technicians can do in
the way of actual maintenance of microphones, outside of minor external repairs. Far more important are the routine things that can be done
on a daily basis to care for them and to ensure that they are working
properly. All technical personnel in a given broadcast, recording, or
sound reinforcement activity should be taught the rudiments of proper
microphone handling, and if these rules are followed a microphone
should last virtually indefinitely with little or no departure from its
original response. In addition, there are number of easy measurements
and performance comparisons that may determine if a microphone is
performing according to its specifications.
When not in actual studio use, microphones should be stored in the
boxes in which they were shipped. A good alternative here is to construct
a cabinet with felt-lined sections for each microphone, and it is recommended that the microphone clip, or mount, normally used with the
model be stored in the same slot. Fussy managers will demand that the
cabinet be locked at all times and that operating personnel be given keys.
Many older models of dynamic and ribbon microphones have open magnetic structures with a considerable stray magnetic field that can attract
ferric dust particles. Over the years this dust can pile up and possibly
impair performance, so be on the lookout for it.
Handheld microphones take the greatest abuse in any location.
Inexperienced users may inadvertently drop them, and prolonged close-in
use will result in breath moisture, even saliva, on the protective screen. It
is a good idea to wipe off all handheld microphones after use with a
3:05 PM
Page 335
20: Care and Maintenance of Microphones
moistened lint-free cloth. Studio quality capacitor microphones are
normally stand-mounted and protected with a pop screen or slip-on foam
screen. Keep enough of these devices on hand for backup replacement.
Modern dynamic microphones are surprisingly rugged and can
sustain multiple drops. Much the same can be said for small format
capacitor models, but large format capacitors will usually end up with
dented screens if dropped. In most variable pattern models, the dual
diaphragm element is suspended on a small post. Some of the vintage
models have been known to break at this point when dropped, so
proceed with caution.
If there is any question about the frequency response of a microphone, it
can be checked one of two ways. Figure 20–1 shows a method for comparing a questionable microphone with one known to be in excellent
operating condition. First, use the reference microphone to arrive at a
rough equalization for flat response as shown on the RTA. (This will
facilitate seeing any clear differences between the response of the two
microphones.) Then, using the same microphone location, substitute the
questionable microphone and compare the two. A test such as this is not
as rigorous as would be carried out in the manufacturer’s laboratory, but
it will enable you to identify a problem. You should also be aware that
not all microphones of a given model are identical in frequency response.
The published specifications will usually make this clear.
The test should be done in a fairly dead acoustical space and as isolated as possible from early reflections. Do not be overly concerned that
Comparing frequency
response of two
microphones using a
real-time analyzer.
3:06 PM
Page 336
neither of the two response curves will look flat. What you are looking
at here is a combination of both loudspeaker and microphone response,
and your concern should be only with the differences between the two
microphone curves.
An alternate method is shown in Figure 20–2. In this test the two
microphones are placed as close together as possible, and a talker is
placed about 1 m (40 in) away. Feed both microphones into a console
and set both trims and faders to the same positions. Then, invert the
polarity of one of the microphones. If the microphones are well matched
the sound level produced by the control room monitors will drop considerably. You may need to make a fine adjustment to one of the trim
controls in order to null out the response to the maximum degree. This
test is not quite as rigorous at the first, but it will isolate any serious differences between the microphones.
You can get a rough idea of relative microphone sensitivities using the
test setup shown in Figure 20–3. The test should be carried out in a studio
and well away from any reflective boundaries. You will need a sound level
meter (SLM) to set the level 1 m from the small loudspeaker to approximately 94 dB LP as measured on the C or flat scale. In the laboratory a
precision SLM would be used for this purpose, but our concern here is
primarily with the relative sensitivities between pairs of microphones.
Both microphone and SLM should be located 1 m from the loudspeaker, and for most purposes the measurement should be made using
Comparing frequency
response of two
microphones by nulling
their outputs.
Comparing microphone
3:06 PM
Page 337
20: Care and Maintenance of Microphones
an octave-wide band of pink noise centered at 1 kHz. Rough sensitivity
comparisons may be read directly from the meter settings on the console
that are necessary to match their levels.
The relative self-noise levels of microphones may be compared by
using an extension of the test we have just discussed. Adjust the two
microphones so that their outputs, as measured in the control room and
using a noise signal of 94 dB LP, are equal. Now, without making any
changes in the gain structure, turn off the noise source and move the
microphones to a very distant and quiet location far away from the control room. A good place might be an isolation room or closet at the far
end of the studio. The intent here is to prevent any audible acoustical
feedback as you proceed with the rest of this test.
Progressively raise the gain of both microphones equally. Do this by
rerouting each microphone through a second set of console inputs so
that you will have enough reserve gain to raise the noise level to a point
where you can clearly hear it. You will progressively have to lower the
control room monitor loudspeakers as you do this, and it is also a good
idea to roll off excess LF response of both microphones via the in-line
equalizers. The key point here is to maintain exactly the same gain in
both microphone channels. Be very careful! With all of this excess gain
you could, with just a little carelessness, damage your monitor loudspeakers. You may need 50 to 60 dB of added gain in order to hear the
self-noise levels clearly.
When carefully carried out this test will let you make subjective comparisons of the noise floors of a pair of microphones. The procedure is
outlined in Figure 20–4.
Comparing microphone
self-noise floors.
3:06 PM
Page 338
This chapter was written as a result of numerous conversations with
both microphone users and manufacturers following publication of the
first edition of The Microphone Book. While the list reflects my personal
view of the subject, it should come as no surprise that it contains virtually all of the celebrated models of the last 75 years. Most engineers
would agree that the German and Austrian capacitor (condenser) models
introduced into the United States shortly after the Second World War
virtually reinvented both popular and classical recording in America, and
those models form the core of this presentation. To these I have added
the American contributions in dynamic and ribbon technology that
became the mainstay of broadcasting and communications in this country,
along with a group of fine omnidirectional capacitors that were also
distinctively American.
There are several independent criteria I have set for inclusion in
the list:
1. The microphone should be at least 30 years old and generally
regarded as an exemplar in its field. Some models are still in demand,
often at collectors’ traditionally high prices.
2. Microphones that have been at the vanguard of a particular design
or usage technology are also eligible, whether or not they have
attained wide renown.
3. Microphones that at an earlier time attained a unique cult following.
What I have in mind here are the aforementioned high-performance
American capacitor omnis that flowered in the early 1950s as part of
the burgeoning audiophile tape recording movement.
3:06 PM
Page 339
21: Classic Microphones: The Author’s View
You might think that gathering photos and specifications of all of
these entries would have been a simple task. Not so. While I want to
thank many people for their permission to use photos of some of these
models, the bulk of the graphics in this chapter come from older publications and manufacturers’ original specification sheets, many of them
requiring all of the touchup wizardry that modern computer graphics
programs can provide. The presentation order is essentially chronological, with occasional grouping of similar models that may be slightly out
of sequence.
Many of the entries show more than just a microphone. Response
curves, circuit diagrams, and control functions are also presented where
they underscore the performance of a given model.
This is as good a time as any to acknowledge the many avid authors,
designers, and historians who have embraced the subject of microphones
over the years. Their dedication to this fascinating field has kept a unique
history alive and vital for many of us – Abaggnaro, 1979; Bauer, 1987;
Knoppow, 1985; Paul, 1989; Sank, 1985; Webb, 1997; and Werner, 2002.
The stunning industrial design of this microphone takes its cue from the
German Bauhaus movement, and its realization in fine metal casting and
finishing is truly amazing for 1928 (Figure 21–1). While Wente’s omnidirectional capacitor design (Western Electric 1917) predates Georg
Neumann’s design by a decade, we must remember that American capacitors were virtually all omnidirectional until well after the Second World
War, whereas Neumann’s design included bidirectional and cardioid
elements based on the work of Braunmühl and Weber. Neumann’s work
in thin plastic material development and gold sputtering techniques pointed
the way for low mass diaphragms with extended response and high sensitivity. You can hear a stereo pair of these microphones on a CD titled “Das
Mikrofon” (Tacet 17) in a performance of excerpts from a Haydn string
quartet. There is nothing “old” about the sound of these microphones.
Largely because of Neumann’s pioneering work, the European
broadcasting and recording communities adopted the capacitor microphone early on in their pursuit of sonic excellence – a state of affairs that
would later position them favorably in the world market that developed
after the war.
Neumann Model CMV3A.
(Photo courtesy of
When broadcasting was introduced in 1919, carbon button microphones
were the mainstay of transmission in the studio. A few capacitors had also
made the grade, but their expense and technical requirements had pretty
much ruled them out for general use. The introduction of the Western
Electric 618 dynamic microphone was a breakthrough in that is was rugged
3:07 PM
Page 340
Western Electric Model 618
(A and C from Read, 1952;
B from Frayne and Wolfe,
and, thanks to new cobalt-based magnet materials, had good sensitivity
and did not require a power supply. In essence it was the first dynamic
omnidirectional microphone, complete with all of the features that are
inherent in that design today. It had a midband-tuned highly damped
moving system complete with a resonance chamber for extended LF and
a small resonator behind the diaphragm for extended HF response.
Figure 21–2A shows the basic unit, and a cutaway view of the system is
shown at B. Typical on- and off-axis response curves are shown at C.
Because of their extensive manufacturing of microphones for broadcasting, motion picture sound, and recording, Western Electric had a need
for a highly stable calibration microphone. The introduction of the
640AA in 1932 provided this reference unit, and it soon became a standard for all electroacoustical manufacturing in the US. Even after Western
Electric ceased microphone manufacturing in the early 1950s, it retained
3:07 PM
Page 341
21: Classic Microphones: The Author’s View
Western Electric Model
640AA. (A and B from
Read, 1952; C from
National Bureau of
Standards, 1969.)
the manufacture of this model until the Danish Bruël & Kjær company
had attained sufficient stature in the instrumentation microphone field to
become the new standard. Figure 21–3A shows the 640AA unit mounted
on the RA-1095 preamplifier, and a close-up detail of the capsule is
shown at B. Typical response curves are shown at C.
Introduced in the mid 1930s, the model 630, dubbed the “Eight Ball,”
improved on the performance of the 618. The new model was designed
to be used with its principal axis facing upward so that it had uniform
response over the entire 360 azimuthal plane. The spherical body design
3:07 PM
Page 342
Western Electric Model
630. (A and C from
Tremaine, 1969; B from
Read, 1952.)
contributed to the overall smoothness of response. Note that the 90
off-axis response is maintained within an envelope of 3 dB from 50 Hz
to 10 kHz.
This famous model from the late 1930s represents Western Electric’s
only use of ribbon technology. In it, there are two elements: an omni
dynamic and a ribbon mounted directly above it. (See Figure 5–8 for
internal construction details.) While the RCA ribbons were corrugated
from top to bottom, the ribbon in the 639 was corrugated only at the top
and bottom, with the middle section crimped so that it moved as a unit.
Via external switching, the two acoustic elements could be operated
independently or combined in several ways to produce the family of firstorder cardioids. Figure 21–5A shows a view of the famous “bird cage”
design, typical on- and off-axis response curves are shown at B, C, and D
for three modes of response.
The model shown here bears the Altec brand name. By the mid
1950s, Western Electric ceased manufacturing commercial microphones,
giving that segment of their business to the Altec Corporation, which
3:07 PM
Page 343
21: Classic Microphones: The Author’s View
Western Electric
Model 639. (Data from
Tremaine, 1969.)
had earlier assumed the manufacture and sale of Western Electric loudspeakers and amplifiers in the late 1930s.
The original A-version of this most famous of all ribbon microphones
dates from 1931, with the BX-version following in the mid 1930s. Few
microphones are as distinctive in overall shape, and its “box-kite” shape
has been copied many times. Olson’s dedication to the engineering simplicity of the ribbon was a hallmark of his illustrious 50-year career at
RCA Laboratories, and a quick survey of Olson’s writings will underscore the variety of concepts, models, and design challenges that he
solved using only ribbons. The microphone became a broadcast favorite
during the 1930s, lasting well into the 1960s. It was also a staple in the
recording studio from the 1930s onward and more than held its own
against the onslaught of German and Austrian capacitor models as the
LP era got under way.
You will find 44s, lovingly maintained, in large recording complexes
everywhere, where they are used primarily for close-in brass pickup in
3:07 PM
Page 344
large studio orchestras and big band jazz groups. Many engineers feel
that the directivity and proximity effects of the 44 provide a warmth and
mellowness on loud, hard-driven brass passages that can’t be easily
attained using capacitors, no matter how many equalizers you may
throw at the problem!
The rugged look of these microphones belies their essential fragility.
The ribbon elements themselves are low-frequency tuned and as such can
be deformed under conditions of mechanical shock. The perforated protective metal screening is also very prone to denting if the microphone is
dropped. A front view of the microphone is shown in Figure 21–6A, and
typical on-axis response is shown at B. RCA ceased manufacture of this
model in the early 1970s.
RCA Model 44-BX.
(A from Tremaine, 1969; B
from RCA specification
3:07 PM
Page 345
21: Classic Microphones: The Author’s View
Virtually everything we have said about the RCA 44 applies as well to the
77. The microphone was designed by Olson’s group in the 1930s as a
unidirectional (cardioid) model for those applications where the figure-8
response of the 44 was not appropriate. The microphone provided a far
more varied design test bed than the 44, and the 77 ultimately went
through 7 iterations (Webb, 1997). A rear view showing the pattern adjusting control is shown in Figure 21–7A, and three switchable polar pattern
families are shown at B, C, and D.
Earlier models of the 77 used the technique shown in Figure 5–9 to
attain cardioid response. One portion of the ribbon was exposed on both
RCA Model 77-D: back
view of microphone
showing pattern
adjustment screw control
(A); polar response (B, C,
and D) (Data from
Tremaine, 1969.)
3:07 PM
Page 346
sides and responded with a figure-8 pattern. The other half of the ribbon
was shrouded on one side via a damping tube and responded as a pressure element. The two combined to produce a cardioid pattern. Later
models used a variable approach as shown in Figure 5–15A. Here, the
shroud extended over the entire back side of the ribbon, and an aperture,
adjustable by a movable vane, provided the amount of gradient component required to attain a desired first-order pattern. As we see at B, C,
and D, the patterns were fairly accurate up to about 4 kHz; above that
frequency the back-side pattern response was only approximate.
This 1947 model was based on the later 77 designs. It had a nonadjustable shroud tailored to produce a nominally flat cardioid pattern
aimed downward 45. It was intended for boom use on motion picture
RCA Model MI-10001:
view of microphone (A);
response curves (B) (Data
from Tremaine, 1969.)
3:07 PM
Page 347
21: Classic Microphones: The Author’s View
sound stages and was produced in limited quantities. It became the first
microphone to be supported in field literature by a comprehensive threedimensional family of polar response curves, of which only two are
shown in Figure 21–8.
Neumann Model U47.
(Photo courtesy of
This multiple pattern microphone was introduced to the US market in
1948 under the trade name of Telefunken, a German distribution company, and gained a high reputation for excellence in both popular and
classical recording. Its introduction coincided pretty much with the
development and advent of the LP record, and it was highly touted (often
as a choice item for LP cover art) for its extended high frequency response.
At a price tag of about $500 (in 1948 a great deal of money), it soon
became a favorite of the recording elite.
The microphone, which is shown in Figure 21–9, provided both
cardioid and omnidirectional patterns. A close relative, the U-48, was
introduced in the mid-1950s and offered cardioid and figure-8 patterns.
By the late 1950s Neumann began marketing its products directly in
the US, and the Telefunken brand name was replaced by Neumann. It
remains one of the most sought-after classic European capacitor models.
These two models are completely different but share the same outer
packaging envelope. The M49, introduced in 1949, was the first capacitor microphone to offer remote pattern switching via a control on the
power supply, as can be seen at B in Figure 21–10. The M50, which is
described in detail in Chapter 3 under On-axis Versus Random Incidence
Response, was designed to solve orchestral pickup problems when positioned in the transition zone between direct and reverberant fields of the
ensemble. Its capsule is mounted on a plastic sphere and behaves as a
pressure element up to about 2.5 kHz. Above that point the response
becomes gradually more directional and rises in output, attaining a
6 dB shelf in response relative to mid and lower frequencies. The intention was to maintain a certain amount of sonic intimacy when recording
at normal distances in spaces with moderate reverberation.
The M50 became the microphone of choice for both British Decca
and EMI classical engineers, and the so-called Decca tree, plus outriggers,
made use of five M50s. Today, it is a mainstay of classical orchestral
recording in large spaces, and it is also widely used for pickup of stereo
ambience in large studios and soundstages.
AKG was founded in Vienna in 1947 and quickly became a major player
in the field of recording and broadcasting microphones. Details of the
3:07 PM
Page 348
FIGURE 21–10
Neumann Models M49/50.
(Photos courtesy of
C-12 capacitor model, which was introduced in 1953, are shown in
Figure 21–11. Electrical pattern switching was facilitated by the unique
dual backplate capsule design used by AKG. Note the pattern switching
unit, which is in line with the microphone and the power supply (detail
shown at C). Switching was noiseless and could be carried out during
recording. The C-12 is highly sought after today for studio work, most
notably for pop vocals.
When Neumann took over its own foreign distribution, Telefunken
turned to AKG for replacement models and in 1959 contracted for what
became the ELAM 251. The basic elements in this model were the same
3:07 PM
Page 349
21: Classic Microphones: The Author’s View
FIGURE 21–11
AKG Model C12. (Data
courtesy of AKG
as in the C-12; however, the in-line pattern changing box was incorporated as a three-way switch on the microphone body. In the overall
redesign process the microphone body was made somewhat larger. The
ELAM 251 remains today one of the most prized, and expensive, vintage
tube microphones items in any studio.
Those readers interested in more technical details of the C12 and
ELAM 251 designs are referred to Paul (1989) and Webb (1997).
3:07 PM
Page 350
Introduced in the late 1930s, the Shure M-55 Unidyne was the first
dynamic microphone with a cardioid pattern. It became the industry prototype for all dynamic vocal microphones both in the US and in Europe.
It was invented by the ever-resourceful Benjamin Bauer, who reasoned
that a good cardioid pattern should exist somewhere between omni and
dipole. His simple solution was to introduce a two-element acoustical
phase shift network instead of a fixed delay path in one branch of a basic
dipole. The time constant of the phase shift network provided excellent
front-back rejection over a broad portion of the midrange, while forward
directionality provided the necessary front-back discrimination at higher
frequencies. The M-55 is shown in Figure 21–13A, and on-axis response
is shown at B.
The M-55 has been in the Shure catalog in one form or another since
its inception in the late 1930s. It is truly one of the icons of the industry
and has been documented in press and news activities over the years.
Who can forget those photos of Elvis Presley and President Kennedy
with the ubiquitous M-55?
FIGURE 21–12
Telefunken Model
ELAM 251. (Data
from early company
FIGURE 21–13
Shure Model M-55:
photo (A); typical on-axis
response (B). (Data
courtesy of Shure Inc.)
3:08 PM
Page 351
21: Classic Microphones: The Author’s View
Wiggins (1954) designed the Variable-DD® dynamic cardioid at ElectroVoice. Up to that time, all dynamic cardioid designs had a single rear opening and a mass-controlled diaphragm in order to maintain flat, extended
LF response. The penalty paid here is handling noise and susceptibility to
mechanical shock. By stiffening the diaphragm and providing three rear
openings (low, medium, and high frequency), each of a different length,
the net forces on the diaphragm remain as before, and the response is effectively flat. (Design details are given in Chapter 5 under The Electro-Voice
Variable-D® Dynamic Microphone.)
The model 666 is shown in Figure 21–14A. That microphone, in
combination with the equalizer shown at B, comprise the Model 667
system. In addition to normal speech applications, Electro-Voice recommended the 666 for more distant sound stage applications, inasmuch
as the microphone had good noise immunity. The electronics package
FIGURE 21–14
Electro-Voice Model
666/667. (Data courtesy of
3:08 PM
Page 352
FIGURE 21–15
Electro-Voice Model 643
rifle microphone. (Data
from Tremaine, 1969 and
company advertising.)
provided added gain and a family of EQ curves to correct for various
distance effects. Details of the EQ contours are shown at C.
Over the years, rifle microphones have rarely exceeded about 0.5 m (20 in)
in length, but in the late 1950s Electro-Voice designed the largest commercial model ever, the 643, which measured 2.2 m (86 in). The model
was designed for coverage of sports events and other large venue activities in an era long before wireless microphones were commonplace. The
microphone is shown in Figure 21–15A, and polar response curves are
shown at B. While the 643 did a remarkable job at moderate distances,
ambient noise, much of it at middle and low frequencies, took its toll on
intelligibility. Additionally, the microphone was expensive and clumsy to
set up and operate. With the coming of wireless microphones the distant
coverage challenges could be far more easily met.
The D-12 dynamic cardioid microphone was introduced in 1952 and with
some modifications remained in the AKG catalog until 1985. It was notable
3:08 PM
Page 353
21: Classic Microphones: The Author’s View
FIGURE 21–16
AKG Acoustics Model
D-12: photo of microphone
(A); various response
curves (B). (Data courtesy
of AKG Acoustics.)
in studio work for its abilities to handle high levels and became a favorite,
both in Europe and the US, for kick drum pickup. The design of the microphone included a very linear moving system and a LF resonance chamber
which extended the response down to about 30 Hz. Figure 21–16A shows
a photo of the model, and frequency response curves are shown at B. You
will find many D-12s still in use, and the model is one of very few single
diaphragm dynamic cardioids to have attained such vintage status.
The C-37A quickly became a lower cost alternative to the Telefunken
U47 when it was introduced during the mid 1950s, and it is easy to
3:08 PM
Page 354
see why. As is clear from Figure 21–17B, the response was remarkably
flat to 20 kHz, lacking the rise in the 10 kHz range which characterized
many of the European dual diaphragm capacitors. Notable also is the
integrity of the cardioid pattern, which retains its 90 target response
of 6 dB remarkably well to just below 10 kHz. By any measure this is
excellent response from a nominal 25 mm (1 in) diameter diaphragm.
The only negative design comment that can be made is the susceptibility
of the perforated metal screen to show dents – much the same way the
RCA ribbons do.
FIGURE 21–17
Sony Model C-37A dual
pattern capacitor. (Photo
courtesy of Eric Weber;
data at B from Sony
specification sheet.)
3:08 PM
Page 355
21: Classic Microphones: The Author’s View
The promise of stereo in the mid 1950s gave rise to a number of capacitor
microphone models in which rotatable, variable pattern elements were
stacked vertically. Notable models here were made by Neumann, AKG
Acoustics, and Schoeps, and they were all quite similar in overall design.
Only the capsule at the end of the stem was rotatable, but both capsules
had variable patterns. All were tube types, but the designs changed
rapidly with the coming of solid state electronics and phantom powering
during the 1960s. The Neumann and AKG models used dual diaphragm
capsules while the Schoeps model used their hallmark single diaphragm
capsule with acousto-mechanical pattern changing.
FIGURE 21–18
Neumann Model SM2
stereo microphone.
(Data courtesy of
3:08 PM
Page 356
FIGURE 21–19
AKG Acoustics Model
C-24 stereo microphone.
(Data courtesy of AKG
As the 1950s got under way it seemed that high-end microphone technology was rapidly becoming a European specialty. In all other respects,
including basic recording technology, loudspeakers, and the electronic
chain, American companies were very much in the running – and in fact
had led in the early development of the post-war art. But in the microphone area, American companies had invested heavily in dynamic and
ribbon technology while the recording industry had firmly endorsed the
newer capacitor microphones as the leading edge in design.
3:08 PM
Page 357
21: Classic Microphones: The Author’s View
FIGURE 21–20
Schoeps Model CMTS-501
stereo microphone. (Data
courtesy of Schoeps GmbH.)
A handful of American companies, some of them quite small, responded
to this challenge. Noting that omni capacitors were relatively straightforward in concept, they decided to make their individual contributions
in this direction. Those companies were: Altec, the most prominent
manufacturer of systems for motion pictures and sound reinforcement;
Stephens Tru-Sonic, a manufacturer of high-end loudspeakers; Capps and
Company, a manufacturer of disc recording styli; and a small California
company, Stanford-Omega.
The Altec M21 system is shown in Figure 21–21A. The basic design
was begun in the late 1940s, and a number of capsule designs were experimented with. The version shown here was unique in that sound reached
the diaphragm indirectly through a circular opening around the rim, which
can be seen in the cutaway view at B. The diaphragm itself was a piece of
optical quality glass ground to a thickness of approximately 0.013–0.02 in,
3:08 PM
Page 358
FIGURE 21–21
Altec Model M21 capacitor
microphone: photo of 150A
base with M21B capsule
(A); cutaway view of
capsule (B). (Photo courtesy
of William Hayes and
Micromike Labs.)
depending on the desired sensitivity, and gold plated on one side. No
tensioning of the diaphragm was necessary inasmuch as the stiffness of
the thin glass was sufficient to maintain a high resonance frequency.
A more conventional capsule, the 11B was also used with the same base.
3:08 PM
Page 359
21: Classic Microphones: The Author’s View
The Stephens Tru-Sonic company was founded in Los Angeles in the
1940s by Robert Stephens, who had worked with Lansing, Shearer, and
Hilliard on the academy award winning MGM loudspeaker system for
motion pictures. Like Lansing, he absorbed the manufacturing philosophies and techniques that had been pioneered by Western Electric, and
his company was regarded during the 1950s in the same professional
class as Altec and JBL. In the early 1950s he introduced the first RF
microphone intended for recording and reinforcement activities, the model
C2-OD4. Figure 21–22A shows the C2 capsule assembly mounted on a
narrow stem, which was connected to the OD4 oscillator-demodulator
through a coaxial cable of any length – as long as the overall cable length
(transmission line) was a multiple of 37.5 inches – a requirement for efficient signal transfer from the capsule to the demodulator. The schematic
of the oscillator-demodulator is shown at B, and it can be seen that the
capsule portion contained only the half-inch (12 mm) diaphragm closely
coupled to an oscillating tank circuit. The system operated on the FM
FIGURE 21–22
Stephens Model C2-OD4
RF microphone. (Photo A
from Read, 1952; data at B
from Tremaine, 1969.)
3:08 PM
Page 360
principle in the range of 9 MHz, and signal demodulation took place in
the associated circuitry.
The system had excellent performance characteristics but exhibited
certain reliability problems due to shifts in the operating characteristics
of the vacuum tubes in the electronics unit. The microphone was a
favorite of such notables as Ewing Nunn, producer of Audiophile recordings, and Paul Klipsch, noted loudspeaker manufacturer and recording
enthusiast. Perhaps because of the long term stability problem, relatively
few of these microphones were built.
Frank Capps had worked for Edison in his phonograph division, and
when that company closed in 1929 Capps founded a company whose
FIGURE 21–23
Capps Model CM2250
microphone. (Data from
company advertising.)
FIGURE 21–24
Stanford-Omega Condenser Microphone: view of microphone (A); frequency response
curves for a stereo pair (B and C). (Data provided by Lowell Cross.)
3:08 PM
Page 361
21: Classic Microphones: The Author’s View
single product was cutting styli for disc recording. In the early 1950s the
company developed the model CM2250 capacitor omni shown in
Figure 21–23. The design was elegant and the tapered capsule assembly
was easily spotted at some distance. Emory Cook, of Cook Laboratories,
used a pair of these microphones in virtually all of his early audiophile
and ethnic music recordings.
Marketed under the name Stanford-Omega, the Thompson Omega
company in southern California manufactured the microphone shown in
Figure 21–24A. The microphone carried no model number and was
known only as the Stanford-Omega Condenser Microphone. The manufacturer did something that very few companies do today – they provided actual calibration curves on each microphone. Curves for a typical
stereo pair are shown at B and C.
The Danish Bang&Olufsen company is primarily noted for high performance home high fidelity equipment elegantly representative of modern
Scandinavian industrial design. As far back as the early 1950s the company was experimenting with stereo, and a variety of ribbon microphones,
both mono and stereo, were built by the company. Compared with the
relatively large ribbons of the mid 1950s, the B&O models were fairly
small, and response beyond 10 kHz was maintained. The microphone was
made famous by the many articles written by Erik Madsen (1957) in both
the technical and popular press in which a spaced pair of these microphones were deployed with a directional baffle (see Figure 12–9A).
Figure 21–25A shows a view of the BM-3 model; response curves,
with and without the protective mesh grille, are shown at B. Note the
smooth response and HF extension beyond 10 kHz. The exposed flanges
FIGURE 21–25
B&O Model BM-3 ribbon
microphone: Photo (A);
frequency response (B).
(Data from B&O
specification sheet.)
3:09 PM
Page 362
on each side of the grille structure are actually magnetic return paths
around the ribbon.
The first capacitor microphones for studio applications were fairly large
models, most of them with switchable pickup patterns. During the early
to mid 1950s three companies, AKG, Schoeps, and Neumann introduced
smaller format designs with capsules in the diameter range 15–18 mm
FIGURE 21–26
The earliest small format
capacitor microphones
from AKG, Schoeps
and Neumann. (Photos
courtesy of the
3:09 PM
Page 363
21: Classic Microphones: The Author’s View
(0.6–0.7 in). Many of these had families of replaceable capsules that
could operate on a common preamplifier. Figure 21–26 shows three such
early models, one from each company, that typified this design approach.
With the advent of phantom powering in the 1960s these designs proliferated and, with their lower prices, eventually dominated the professional market, at least in terms of numbers.
Taking advantage of solid state electronic components and remote
powering, Sennheiser Electronics introduced its RF microphones in the
early 1960s. The designs have been refined over the years, and today they
are the flagship products of the company, exhibiting performance
characteristics that are at the leading edge of the art. The MKH 404,
shown at Figure 21–27A, was the first solid state cardioid microphone
FIGURE 21–27
Sennheiser MKH 404 RF
microphone: photo of
microphone (A); polar
response (B); off-axis
response curves (C);
circuit diagram (D). (Data
courtesy of Sennheiser
Electronics and Tremaine,
3:09 PM
Page 364
that operated on the RF principle. The design had been preceded by the
omnidirectional model MKH 104. Performance data is shown at B and C,
and a schematic drawing of the electronics is shown at D. The overall
uniformity of response is notable, but the polar response is exemplary,
even by today’s standards. In the earliest days of these models, phantom
powering was not yet in general use; Sennheiser manufactured an in-line
battery module that provided 9 Vdc powering for these microphones.
The famous 421 was introduced in 1960 and has been in the Sennheiser
catalog ever since. While it was intended as a general-purpose dynamic
cardioid, its robustness was evident from the start. It eventually gravitated, along with the AKG D-12, into its present role as a kick drum
microphone par excellence. Today it is the most sought-after dynamic
microphone for that purpose.
FIGURE 21–28
Sennheiser Model MD 421
dynamic. (Data courtesy of
Sennheiser Electronics.)
3:09 PM
Page 365
21: Classic Microphones: The Author’s View
Dating from the early 1970s, this is the “youngest” microphone presented
in this survey. The Crown PZM pressure zone microphones differ from
other omni boundary microphones in that they provide a short indirect
path to the microphone capsule, thus ensuring that the amplitude pickup
will not be materially influenced by the bearing angle of the sound source.
The microphone is shown in Figure 21–29A, and details of the actual
path to the capsule are shown at B. The PZM-30 has been in the Crown
catalog since it was introduced and remains the classic reference for
boundary layer recording.
FIGURE 21–29
Crown International
PZM-30 boundary layer
microphone: photo (A);
detail of sound pickup (B).
(Photo courtesy of Crown
3:09 PM
Page 366
3:13 PM
Page 367
L. Abaggnaro, ed., AES Anthology, Microphones. New York: Audio Engineering
Society, 1979.
Anon., “The Telephone at the Paris Opera.” Scientific American, 31 December
1881 (reprinted in Journal of the Audio Engineering Society 29, no. 5 (May
M. Barron, “The Subjective Effects of First Reflections in Concert Halls – the
Need for Lateral Reflections.” Journal of Sound and Vibration 15 (1971),
pp. 475–494.
B. Bauer, “Uniphase Unidirectional Microphone.” Journal of the Acoustical
Society of America 13 (1941), p. 41.
B. Bauer, “A Century of Microphones.” Proceedings, IRE (1962), pp. 719–729
(Also, Journal of the Audio Engineering Society 35, no. 4 (1987).
A. Benade, Fundamentals of Musical Acoustics. New York: Oxford University
Press (1976).
L. Beranek, Acoustics. New York: J. Wiley, 1954.
L. Beranek, Concert and Opera Halls: How they Sound. Acoustical Society of
America, New York (1996).
W. Bevan et al., “Design of a Studio Quality Condenser Microphone Using
Electret Technology,” Journal of the Audio Engineering Society 26, no. 12
(1978). (Included in AES Microphone Anthology.)
J. Blauert, Spatial Hearing. Cambridge, MA: MIT Press, 1983, pp. 209–210.
G. Boré, Microphones. Neumann USA, Sennheiser Electronics, 6 Vista Drive,
Old Lyme, CT, 06371 (1989).
M. Brandstein and D. Ward (eds), Microphone Arrays, New York: Springer,
H. Braünmuhl and W. Weber, “Kapacitive Richtmikrophon,” Hochfrequenztechnic
und Elektroakustic 46 (1935), p. 187.
D. Cooper and T. Shiga, “Discrete-Matrix Multichannel Sound,” Journal of the
Audio Engineering Society 20, no. 5 (June 1972).
J. Cooper, Building a Recording Studio, 5th edn. Synergy Group, 23930
Craftsman Road, Calabasas, CA 91302 (1996).
3:13 PM
Page 368
References and Bibliography
S. Dove, “Consoles and Systems,” Chapter 22 in Handbook for Sound System
Engineers. Indianapolis: Sams & Co., 1987.
J. Eargle, Microphone Handbook. Commack, NY: Elar Publishing, 1981.
J. Eargle, Electroacoustical Reference Data. New York: Van Nostrand Reinhold,
J. Eargle, Music, Sound & Technology. New York: Van Nostrand Reinhold,
J. Frayne and H. Wolfe, Sound Recording. New York: John Wiley & Sons, 1949.
M. Gayford, ed., Microphone Engineering Handbook. London: Focal Press, 1994.
R. Glover, “A Review of Cardioid Type Unidirectional Microphones,” Journal of
the Acoustical Society of America 11 (1940), pp. 296–302.
M. Hibbing, “High-quality RF Condenser Microphones,” Chapter 4 in Microphone
Engineering Handbook, ed. M. Gayford. London: Focal Press, 1994.
F. Hunt, Electroacoustics. New York: Acoustical Society of America, 1982.
D. Josephson, “Progress in Microphone Characterization – SC-04–04,” presented at the 103rd Audio Engineering Society Convention, New York,
September 1997, preprint number 4618.
F. Khalil et al., “Microphone Array for Sound Pickup in Teleconference
Systems,” Journal of the Audio Engineering Society 42, no. 9 (1994).
L. Kinsler et al., Fundamentals of Acoustics. New York: J. Wiley & Sons, 1982.
R. Knoppow, “A Bibliography of the Relevant Literature on the Subject of
Microphones,” Journal of the Audio Engineering Society 33, no. 7/8 (1985).
Y. Mahieux et al., “A Microphone Array for Multimedia Workstations,” Journal
of the Audio Engineering Society 44, no. 5 (1996).
J. Meyer, Acoustics and the Performance of Music, translated by Bowsher and
Westphal. Frankfurt: Verlag Das Musikinstrument, 1978.
J. Monforte, “Neumann Solution-D Microphone,” Mix Magazine (October
H. Nomura and H. Miyata, “Microphone Arrays for Improving Speech
Intelligibility in a Reverberant or Noisy Space,” Journal of the Audio
Engineering Society 41, no. 10 (1993).
H. Olson, “The Ribbon Microphone,” Journal of the Society of Motion Picture
Engineers 16 (1931), p. 695.
H. Olson, “Line Microphones,” Proceedings of the IRE, vol. 27 (July 1939).
H. Olson, Acoustical Engineering, New York: D. Van Nostrand and Company,
1957. Reprinted by Professional Audio Journals, 1991, PO Box 31718,
Philadelphia, PA 19147–7718.
H. Olson, “Ribbon Velocity Microphones,” Journal of the Audio Engineering
Society 18, no. 3 (June 1970). (Included in AES Anthology, Microphones.)
H. Olson, Modern Sound Reproduction, New York: Van Nostrand Reinhold,
P. Parkin, “Assisted Resonance,” Auditorium Acoustics. London: Applied
Science Publishers, 1975. pp. 169–179.
S. Paul, “Vintage Microphones,” parts 1, 2, and 3, Mix Magazine (Oct, Nov, and
Dec 1989).
C. Perkins, “Microphone Preamplifiers – A Primer,” Sound & Video Contractor
12, no. 2 (February 1994).
S. Peus, “Measurements on Studio Microphones,” presented at the 103rd Audio
Engineering Society Convention, New York, September 1997, preprint
number 4617.
3:13 PM
Page 369
References and Bibliography
S. Peus and O. Kern, “TLM170 Design,” Studio Sound 28, no. 3 (March
J. Pierce, The Science of Musical Sound. New York: Scientific American Books,
O. Read, The Recording and Reproduction of Sound. Indianapolis: H. Sams,
A. Robertson, Microphones. New York: Hayden Publishing, 1963.
J. Sank, “Microphones,” Journal of the Audio Engineering Society 33, no. 7/8
(July/August 1985).
G. Sessler and J. West, “Condenser Microphone with Electret Foil,” Journal of
the Audio Engineering Society 12, no. 2 (April 1964).
D. Shorter and H. Harwood, “The Design of a Ribbon Type Pressure-Gradient
Microphone for Broadcast Transmission,” Research Department, BBC
Engineering Division, BBC Publications, London (December 1955).
H. Souther, “An Adventure in Microphone Design,” Journal of the Audio
Engineering Society 1, no. 2 (1953). (Included in AES Microphone
J. Steinberg and W. Snow, “Auditory Perspective – Physical Factors,” Electrical
Engineering 53, no. 1 (1934), pp. 12–15 reprinted in AES Anthology of
Stereophonic Techniques (1986).
R. Streicher and W. Dooley, “The Bidirectional Microphone: A Forgotten
Patriarch,” AES 113rd Convention (Los Angeles 5–8 October 2002).
Preprint number 5646.
H. Tremaine, Audio Cyclopedia. 2nd edn. Indianapolis: H. Sams, 1969.
M. van der Wal et al., “Design of Logarithmically Spaced Constant-Directivity
Transducer Arrays,” Journal of the Audio Engineering Society 44, no. 6
Various, Condenser Microphones and Microphone Preamplifiers: Theory and
Application Handbook, published by Brüel & Kjær, 185 Forest Street,
Marlborough, MA 01752 (reprinted May 1977).
Various, A History of Engineering & Science in the Bell System, The Early Years
(1875–1925), Bell Laboratories (1975).
Various, “Shields and Grounds,” Journal of the Audio Engineering Society 43,
no. 6 (June 1995).
J. Webb, “Twelve Microphones that Changed History,” Mix Magazine 21,
no. 10 (October 1997).
B. Weingartner, “Two-Way Cardioid Microphone,” Journal of the Audio
Engineering Society 14, no. 3 (July 1966).
E. Werner, “Selected Highlights of Microphone History,” AES 112th Convention
(Munich, 10–13 May 2002). Preprint number 5607.
R. Werner, “On Electrical Loading of Microphones,” Journal of the Audio
Engineering Society 3, no. 4 (October 1955).
A. Wiggins, “Unidirectional Microphone Utilizing a Variable Distance between
Front and Back of the Diaphragm,” Journal of the Acoustical Society of
America 26 (Sept 1954), pp. 687–602.
G. Wong and T. Embleton, eds, AIP Handbook of Condenser Microphones. New
York: American Institute of Physics, 1995.
A. Zukerwar, “Principles of Operation of Condenser Microphones,” Chapter 3
in AIP Handbook of Condenser Microphones, Wong, G. and Embleton, T.,
eds. New York: AIP Press, 1994.
3:13 PM
Page 370
References and Bibliography
B. Bartlett, Stereo Microphone Techniques. Boston: Focal Press, 1991.
A. Benade, “From Instrument to Ear in a Room: Direct or via Recording,”
Journal of the Audio Engineering Society 33, no. 4 (1985).
A. D. Blumlein, “British Patent Specification 394,325 (Directional Effect in
Sound Systems),” Journal of the Audio Engineering Society 6, (reprinted
1958), pp. 91–98.
J. Borwick, Microphones: Technology and Technique. London: Focal Press,
J. Bruck, “The KFM 360 Surround Sound – A Purist Approach,” presented at
the 103rd Audio Engineering Society Convention, New York, November
1997. Preprint number 4637.
R. Caplain, Techniques de Prise de Son, (in French). Paris: Editions Techniques
et Scientifiques Françaises, 1980.
C. Ceoen, “Comparative Stereophonic Listening Tests,” Journal of the Audio
Engineering Society 20. no. 1 (1970).
H. Clark et al., “The ‘Stereosonic’ recording and reproduction system,” Journal
of the Audio Engineering Society 6, no. 2 (April 1958).
E. Cohen and J. Eargle, “Audio in a 5.1 Channel Environment,” presented at the
99th Audio Engineering Society Convention, New York, October 1995.
Preprint number 4071.
D. Cooper and J. Bauck, “Prospects for Transaural Recording, Journal of the
Audio Engineering Society 37, no. 1/2 (1989).
J. Culshaw, Ring Resounding. New York: Viking, 1957.
J. Culshaw, Putting the Record Straight. New York: Viking, 1981.
N. Del Mar, Anatomy of the Orchestra, Berkeley and Los Angeles: University of
California Press, 1983.
M. Dickreiter, Tonemeister Technology (translated by S. Temmer) New York:
Temmer Enterprises, 1989.
W. Dooley, Users Guide, Coles 4038 Studio Ribbon Microphone, Pasadena, CA:
Audio Engineering Associates, 1997.
W. Dooley and R. Streicher, “MS. Stereo: A Powerful Technique for Working in
Stereo,” Journal of the Audio Engineering Society 30 (1982), pp. 707–717.
J. Eargle, Handbook of Recording Engineering. Boston: Kluwer Academic
Publishers, 2003.
N. V. Franssen, Stereophony. Eindhoven: Philips Technical Bibliography, 1963.
F. Gaisberg, The Music Goes Round. New York: Macmillan, 1942.
R. Gelatt, The Fabulous Phonograph. New York: Lippincott, 1955.
H. Gerlach, “Stereo Sound Recording with Shotgun Microphones,” Product
usage bulletin, Sennheiser Electronic Corporation, 6 Vista Drive, Old Lyme,
CT (1989).
M. Gerzon, “Periphony: With-Height Sound Reproduction,” Journal of the
Audio Engineering Society 21, no. 1 (1973).
M. Gerzon, “Optimum Reproduction Matrices for Multichannel Stereo,”
Journal of the Audio Engineering Society 40, no. 7/8 (July/August 1992).
M. Hibbing, “XY and MS Microphone Techniques in Comparison,” Sennheiser
News Publication, Sennheiser Electronics Corporation, PO Box 987, Old
Lyme, CT, USA 06371 (1989).
T. Holman, “The Number of Audio Channels,” parts 1 and 2, Surround
Professional Magazine 2, nos. 7 and 8 (1999).
3:13 PM
Page 371
References and Bibliography
T. Holman, 5.1 Channel Surround Up and Running. Woburn, MA: Focal Press,
U. Horbach, “Practical Implementation of Data-Based Wave Field Reconstruction,”
108th AES Convention (Paris, 2000).
D. Huber and R. Rundstein, Modern Recording Techniques, 4th edn. Boston:
Focal Press, 1997.
ITU (International Telecommunication Union) document ITU-R BS.775.
J. Jecklin, “A Different Way to Record Classical Music,” Journal of the Audio
Engineering Society 29, no. 5 (May 1981).
J. Klepko, “Five-Channel Microphone Array with Binaural Head for
Multichannel Reproduction,” presented at the 103rd AES Convention,
1997. Preprint no. 4541.
W. Kuhl, Acustica 4 (1954), p. 618.
C. Kyriakakis and C. Lin, “A Fuzzy Cerebellar Model Approach for Synthesizing
Multichannel Recording, 113th AES Convention (Los Angeles, 5–8 October
2002). Preprint no. 5675.
C. Kyriakakis and A. Mouchtaris, “Virtual Microphones for Multichannel
Audio Applications,” ICME 2000, New York (October 2000).
E. Madsen, “The Application of Velocity Microphones to Stereophonic
Recording,” Journal of the Audio Engineering Society 5, no. 2 (April 1957).
D. Malham, “Homogeneous and Non-Homogeneous Surround Sound Systems,”
AES UK Second Century of Audio Conference, London (7–8 June 1999).
J. Meyer and T. Agnello, “Spherical Microphone Array for Spatial Sound
Recording,” 115th AES Convention (New York, 10–13 October 2003).
Preprint no. 5975.
D. Mitchell, “Tracking for 5.1,” Audio Media Magazine, October 1999.
C. O’Connell, The Other Side of the Record. New York: Knopf, 1941.
L. Olson, “The Stereo-180 Microphone System,” Journal of the Audio
Engineering Society 27, no. 3 (1979).
A. Previn, Andre Previn’s Guide to the Orchestra. London: Macmillan, 1983.
O. Read and W. Welch, From Tinfoil to Stereo. Indianapolis: Sams, 1959.
E. Schwarzkopf, On and Off the Record: A Memoire of Walter Legge, New
York: Scribners, 1982.
W. Snow, “Basic Principles of Stereophonic Sound,” Journal of the Society of
Motion Picture and Television Engineers 61 (1953), pp. 567–589.
R. Streicher and F. Everest, The New Stereo Soundbook. Pasadena, CA: Audio
Engineering Associates, 1999.
G. Theile, “Main Microphone Techniques for the 3/2-Stereo-Standard,” presented at the Tonmeister Conference, Germany, 1996.
G. Theile, “Multichannel Natural Recording Based on Psychoacoustical Principles,”
presented at the 108th AES Convention (Paris 2000). Preprint no. 5156.
G. Theile, “On the Naturalness of Two-Channel Stereo Sound,” Proceedings of
the AES 9th International Conference, Detroit, Michigan (pp. 143–149).
M. Thorne, “Stereo Microphone Techniques,” Studio Sound 15, no. 7 (1973).
M. Tohyama et al., The Nature and Technology of Acoustical Space, Academic
Press, London (1995).
Various, Stereophonic Techniques, an anthology published by the Audio
Engineering Society, New York (1986).
M. Williams, “Microphone Array analysis for Multichannel Sound Recording,”
presented at the 107th Audio Engineering Society Convention, New York,
September 1999. Preprint no. 4997.
3:13 PM
Page 372
References and Bibliography
M. Williams, “Unified Theory of Microphone Systems for Stereophonic Sound
Recording,” Audio Engineering Society. Preprint no. 2466 (1987).
W. Woszczyk, “A Microphone Technique Applying to the Principle of SecondOrder Gradient Unidirectionality,” Journal of the Audio Engineering
Society 32, no. 7/8 (1984).
T. Yamamoto, “Quadraphonic One-Point Pickup Microphone,” Journal of the
Audio Engineering Society 21, no. 4 (1973).
“Das Mikrofon,” Compact Disc, Tacet 17; Nauheimer Strasse 57, D-7000
Stuttgart 50, Germany (1991).
AKG Acoustics WMS300 Series Wireless Microphone System, AKG Acoustics,
1449 Donelson Pike, Nashville, TN 37217 (1997).
G. Ballou, ed., Handbook for Sound Engineers. Indianapolis: H. Sams, 2002.
B. Beavers and R. Brown, “Third-Order Gradient Microphone for Speech
Reception,” Journal of the Audio Engineering Society 16, no. 2 (1970).
C. P. Boner and C. R. Boner, “Equalization of the Sound System in the Harris
County Domed Stadium,” Journal of the Audio Engineering Society 14,
no. 2 (1966).
C. P. Boner and R. E. Boner, “The Gain of a Sound System,” Journal of the Audio
Engineering Society 17, no. 2 (1969).
D. and C. Davis, Sound System Engineering, 2nd edn., Boston: Focal Press,
D. Dugan, “Automatic Microphone Mixing,” Journal of the Audio Engineering
Society 23, no. 6 (1975).
Y. Ishigaki et al., “A Zoom Microphone,” Audio Engineering Society preprint
no. 1713 (1980).
D. Klepper, “Sound Systems in Reverberant Rooms for Worship,” Journal of the
Audio Engineering Society 18, no. 4 (1970).
H. Tremaine, The Audio Cyclopedia, 2nd edn. Indianapolis: H. Sams, 1969.
Various, Sound Reinforcement, Volumes 1 and 2 (ed. D. Klepper); reprinted
from the pages of the Journal of the Audio Engineering Society (1978 and
T. Vear, Selection and Operation of Wireless Microphone Systems, Shure, Inc.,
222 Hartrey Avenue, Evanston, IL 60202 (2003).
S. Woolley, “Echo Cancellation Explained,” Sound & Communications (January
3:12 PM
Page 373
A-B powering (See T-powering)
AB stereo (spaced microphone
pickup), 177–179
Absolute polarity, 117
Absorption coefficient, 15–16
Accent microphones, 204–206
Acoustic bass pickup, 226–228
Acoustical power, 10–11
Active acoustics
Delta Stereophony, 321
LARES (Lexicon acoustical
reverberance enhancement
system), 320–321
Parkin’s Assisted Resonance, 320
Adaptive arrays, 328–329
ADP (ammonium dihydrogen
phosphate), 48
Air dielectric, 26
Air particle velocity, 11
Altar microphones, 309–310
Altering room acoustics, 210–212
Altering the piano 225–226
Ambience in recording, 245–247
Amplitude, 7
Analysis, cardioid microphones,
Analysis, gradient microphones,
Analysis, pressure microphones,
Anti-noise (noise canceling) microphones, 291–292
Anti-phase signals, 9
Armature microphone, 1–2
Assisted Resonance (Parkin), 320
Automatic microphone mixing,
Average value of waveform, 11
Azimuthal stereo microphone,
Baffles (goboes), 222–223
Barometric pressure, 30
Barron’s data, 177–178
Bass equalizing tube, dynamic
microphones, 44
Beamwidth, definition, 96
Biasing resistance, capacitor
microphone, 28–29
Binaural recording, 187, 191
Binaural synthesis, 267
Blumlein array, 169–171
Blumlein, early stereo, 184–185
Boundary layer microphone,
Brass instruments, 232–233
Braunmühl-Weber dual diaphragm
microphones, 81–84
Braunmühl-Weber electronic pattern
control, 82–84
Broadcast studios, 297–298
Cable testing, 133–136
Cables, electrical characteristics,
Calculation of microphone directivity,
Capacitor biasing resistance, 28–29
Capacitor gradient microphones,
Capacitor microphone padding of
output, 29–33
Capacitor microphone sensitivity to
barometric pressure, 30
Capacitor microphone sensitivity to
temperature, 29–30
Capacitor microphone
bias requirements, 26–27
electrical details, 26–33
electrical loading, 33
electronics, 32–33
noise spectra, 35–38
polarizing, 24
RF operation, 35
self-noise floor, 37
shielding, 27
diaphragm materials, 24
Capillary tube, capacitor microphones, 24
Carbon microphone, 3–5
Cardioid family
Braunmühl-Weber type, 81–84
capacitor, 77–78
cardioid, 67–69
dynamic, 78–79
hypercardioid, 68–70
proximity effect, 76
subcardioid, 67–69
summary, 70–72
supercardioid, 67–69
Variable-D, 84–87
Chamber music recording, 202–204
Chamber orchestra recording,
Coincident arrays, 168–173
Concert sound reinforcement,
Condenser (See Capacitor)
Console input circuitry, 123–128
Creating a center channel, 257–258
Critical distance (DC), 15–16, 305
Crossed cardioid arrays, 172–173
3:12 PM
Page 374
dBu, definition, 136
DC (critical distance), 15–16, 305
DC-to-DC conversion, 120–121
Decca tree, 182
Decibel (dB), 7, 11–13
Delayed signals in stereo, 168
Delta Stereophony, 321
Desk and surface mounting, 152
Desk stands, 152
Dialog pickup, 297–298
Diaphragm materials, 24
Dielectric, air, 26
Diffraction, 19–21
Digital microphones, 140–141
Direct field, 14–16
Directivity and polar response of
microphones, 106–108
Directivity factor, 16
Directivity of musical instruments,
Direct-to-reverberant ratio in
recording, 214–215
Discrete line array theory, 322–324
Distance factor (DF), 70–71
Distortion in microphones, 112
Drum set pickup, 218–222
Dual-element directional
microphones, 72–75
Dynamic pressure microphones,
Dynamic range of microphones, 113
Dynamic ranges of musical
instruments, 198–199
EAD (equivalent acoustic distance),
Echo cancellation, 329–331
Effects of humidity, 17–18
Eigenmike, 244, 262–263
Electret microphones, 5–6, 38–39
electret backplate, 38
foil electret, 38
polytetrafluoroethelene electret
material, 38
Electrical loading, capacitor
microphone, 33
Electrodynamic microphones, 4, 42
Electromagnetic microphones, 42
Electronic control of feedback,
Electronic microphones, 47
Equivalence of MS and XY
techniques, 175
Estimating line microphone
performance, 94–96
Estimating system intelligibility,
Extensive quantities, 10
Faulkner array, 180
Feedback in speech systems, 303
Fiberoptic transmission, 135
Figure-8 response, 20
First-order cardioid microphones,
First-order cardioid patterns in three
dimensions, 72, 74
Forward acceptance angle 70–71
Free field, 14–15
Frequency, 7
Frequency shifting, 314
Gain analysis of speech systems,
Gain structure, 136–138
Geometrical mean, 43
Gradient microphone, 50–51
Gradient microphone response,
Gradient, pressure, 16
Grounding and ground loops,
Guitar recording, 200–201,
Handheld booms, 154–155
Hand-held MS microphone,
Harpsichord recording, 199
Head tracking, 266
Headset-microphone combination,
HF nulls, gradient response, 54–55
High-directivity microphones,
Higher-order microphone patterns,
History of microphones, 1–6
Hole-in-the-middle, 176–177
Holophone, 252
HRTFs (head-related transfer
functions), 265–266
Hum pickup in microphones,
Hum-bucking coil, 47, 113
Humidity, effects of, 17–18
Hypercardioid pattern, 68–70
IACC (interaural cross-correlation),
Impedance, 108–110
Impulse response of microphones,
In-line accessories, 163–164
Instrumentation microphones, 39–40
bandwidth, 41
diaphragm displacement, 41–42
operational ranges, 39–40
sensitivity, 41
Intensity, 10–11
Intensive quantities, 10
Interconnecting electronic hardware,
Interference-type microphones,
Internal delay paths, 77–79
ITU (International
Telecommunications Union),
Jecklin disc, 193
Johnston-Lam seven channel array,
Klepko frontal array, 255
LARES, 320–321
Large studio orchestra, 238–239
Lectern and pulpit microphones, 308
Level, power, 12
Levels of typical sound sources, 13
Line losses and electrical interference,
Line microphones, 93–97
Loading of microphones, 126–128
Localization (Franssen), 169
Loss in dB over distance, 14
Madsen baffle array, 193
Magnetic flux density, 42–43
Magnetic induction, 11–43
Magnetostriction, 48
Maintenance of microphones
noise level, 337
checking frequency response,
checking sensitivity, 336–337
Mallet instruments, 224
Mass-controlled diaphragms and
ribbons, 53
3:12 PM
Page 375
Maximum SPL handling capability of
microphones, 106
Microphone array for narrow vertical
coverage, 326–327
Microphone array for work stations,
Microphone emulator circuit,
Microphone gating, 292–293
Microphone HF response extension,
Microphone history, 1–6
Microphone preamplifier noise
spectra, 37–38
Microphone snakes, 133
Microphone specifications, 105–166
Microphone splitters, 129–130
armature, 1–2
Braunmühl-Weber type, 81–84
capacitor, 4–6
carbon type, 3–5
cardioid, 5
diaphragm materials, 24
diffraction, 19–21
directivity and polar response,
distortion, 112
dynamic range, 113
electret type, 5–6, 38–39
electrodynamic, 4
figure-8 pattern, 20
gradient type, 50–51
handheld booms, 154–155
hum pickup, 113
in-line accessories, 163–164
instrumentation type, 39–40
loading, 126–128
maximum SPL capability, 106
moving coil, 4
output padding, 123–125
output sensitivity, 105
output source impedance, 110
output transformers, 125–126
permanent installations, 157–158
piezoelectric type, 5
pop screens, 161
power output sensitivity, 108–110
power supplies, 164–165
powering of capacitors, 117–120
pressure type, 22–49
proximity effect, 17
ribbon type, 4
self-noise, 106
self-noise weighting curves, 111
shock mounts, 162–163
source impedance, 105
stands and booms, 154–155
stereo mounts, 154–157
unbalanced inputs, 128–129
variable contact, 2–5
velocity type, 50–51
windscreens, 158–161
Microphones for teleconferencing,
294, 298–299
Midband damping, dynamic
microphones, 44
Mike mouse, 152–154
Mixed microphone arrays, 181–183
Modern studio acoustics, 239–241
Moving coil (dynamic) microphone,
4, 42
MS (mid-side) technique, 173–175
Multiple tube microphones, 92
Narrow-band equalization, 314–315
Near-coincident techniques, 179–180
Neodymium magnets, 45–46
Noise canceling microphone,
Noise floor, capacitor microphone, 37
Noise spectra, capacitor microphone,
NOM (number of open microphones), 316
NOS array, 180
Olson Stereo-180 array, 180
Optical microphones, 48
ORTF array, 179–181
Output levels of musical instruments,
195, 198–199
Output sensitivity of microphones,
Output source impedance of
microphones, 110
Padding of preamplifier inputs, 127
Padding of microphone outputs,
29–33, 125
PAG (potential acoustical gain), 303
Paging in high noise areas, 300
Paging systems, 299–300
Panpots, 172
Parabolic reflectors and lenses, 96–98
Parallel operation of microphones,
Pascal (Pa), 12
Peak value of waveform, 11
Percussion instruments, 223–224
Period, 7
Permanent installations, 157–158
Perspectives in recording, 213–214
Phantom (Simplex) powering,
Phantom images, 166–168
Phase differences in stereo, 166
Phase relationship, 8–9
Piano recording, 198–200, 225–226
Piezoelectric bimorph crystals, 49
Piezoelectric microphones, 48–49
Piezoelectric principle, 5
Pipe organ recording, 201–202
Plane wave, 7, 16
Polarflex system (Schoeps), 89–90
Polarizing, capacitor
microphones, 24
Pop screens, 161
Power output sensitivity nomograph,
Power output sensitivity of
microphones, 108–110
Power supplies, 164–165
Power, acoustical, 10–11
Powering of capacitor microphones,
Pressure gradient, 16
Pressure gradient microphones,
Pressure microphones, 22–49
Proximity effect, 17, 64–65
Proximity effect in cardioid
microphones, 75–77
PZM microphones, 291–292
Quadraphonic microphones, 257
Random efficiency (RE), 70–71
Ranges of microphone sensitivity,
Rayl (SI), 11
Reciprocity calibration of
microphones, 113–114
accent microphones, 204–206
chamber music, 202–204
chamber orchestra, 204–206
guitar, 200–201
harpsichord, 199
piano, 198–200
pipe organ, 201–202
solo instruments and voice,
3:12 PM
Page 376
Recording (continued)
soloists with orchestra, 210
symphony orchestra, 206–210
Reflections from lecterns, 310–311
Rejection of off-axis sound, 70–71
Resistance controlled diaphragm,
Reverberant field, 14–16, 305
Reverberation time (R60), definition,
RF (radio frequency) microphone
technology, 138–139
RF operation, capacitor microphones,
Ribbon microphone
damping, 55–57
design details, 55–56
dimensions, 55–56
ribbon sag and deformation, 56
response, 58–60
sensitivity, 57–58
transformer, 56, 58
Ribbon stereo microphone, 186–187
Rifle (shotgun) microphones, 91–96
RMS (root-mean-square) value of
waveform, 11–12, 23
Rochelle salts (potassium sodium
tartrate), 48
SAM array, 253–254
Schoeps KFM 360 array, 251–255
Schoeps OCT array, 255–256
Schroeder-Atal crosstalk cancellation,
Second-order frontal array,
Self-noise of microphones, 106
Self-noise weighting curves for
microphones, 111
Sensitivity to temperature, capacitor
microphones, 29–30
Shielding, capacitor
microphones, 27
Shock mounts, 162–163
Simplex powering (See Phantom
Sine wave, 8
Single-diaphragm cardioid microphones, 77–79
SLM (Sound Level Meter), 39–49
Solo instruments and vocal recording,
Soloists with large orchestra, 210
Sound level meter, 39–40
Sound pressure, 11
Sound pressure versus level, 14
Sound reinforcement,
automatic microphone mixing,
electronic control of feedback,
estimating system speech
intelligibility, 316–318
loudspeaker types, 311–314
Soundfield microphone, 244,
Source impedance of microphones,
Spaced pickup techniques, 174–179
Specific acoustical impedance, 11
Speech reinforcement
altar microphones, 309–310
feedback, 303
gain analysis, 302–307
head-worn microphone, 307
intelligibility, 316–318
lectern microphones, 308
reflections from lecterns, 310–311
requirements, 301–302
reverberant field, 305
Speed of sound, 9
Spherical wave, 7, 16–176
SPL array, 253–254
Spot microphones, (See Accent
Stage preamplifiers, 135
Stand-alone microphone
preamplifiers, 130–131
Stands and booms, 154–155
Starquad cable, 132
Stereo listening conditions, 183
Stereo microphones
baffled pairs, 192–193
Crown SASS (Stereo Ambient
Sampling System), 198, 192
dummy heads, 187, 191
Jecklin disc, 193
Madsen microphone baffle, 193
remote control, 187, 189
spheres, 187, 192
stereo mounts, 154–157
Blumlein array, 169–171
coincident arrays, 168–173
crossed cardioids, 172–173
Decca tree, 182
delayed signals, 168
Franssen localization, 169
MS technique, 173–175
panpots, 172
phantom images, 166–168
phase differences, 166
spaced pickup, 174–179
XY technique, 172–173
Stereosonic technique, 170
Stiffness controlled diaphragm,
String instruments, 232–233
Studio acoustics, 238–241
Studio recording
acoustic bass, 226–228
brass instruments, 232–233
drum set, 218–222
guitar, 230–231
large orchestra, 238–239
mallet instruments, 224
percussion instruments, 223–224
piano, 225–226
string instruments, 232–233
synthesizer, 231
vocals and vocal groups, 228–230
woodwind instruments, 232
Subcardioid pattern, 66–67, 70
Subtractive mixing, 271, 274–277
Summing multiple outputs,
Super Phantom Powering, 121–122
Supercardioid pattern, 66–67, 70
Surround sound playback
configurations, 247–249
Surround sound
ambience, 245–247
creating a center channel,
dipole loudspeakers, 248
Eigenmike, 244, 262–263
frontal second-order array,
general requirements, 272–273
Holophone, 252
ITU recommendations, 248–249
Johnston-Lam seven channel array,
Klepko frontal array, 255
microphone technology, 243–270
parallax, 244
quadraphonic microphones, 257
SAM (surround ambience
microphone) array, 253–254
Schoeps KFM 360 array, 251–255
Schoeps OCT (optimum cardioid
triangle) array, 255–256
Soundfield microphone, 244,
SPL (Sound Performance Lab)
array, 253–254
subtractive mixing, 271, 274–277
3:12 PM
Page 377
TMH 10.2 playback array, 250
transaural technology, 264–267
video formats, 5.1 243
wavefront reconstruction,
Symphony orchestra recording,
Synthesizer recording, 231
Synthesizing first-order patterns,
Systems with parallax, 244
Telephone handset, 290–291
Temperature, effects of, 9
THD (total harmonic distortion), 112
Theater systems, stereo
reinforcement, 318–319
Thermal microphones, 48
Three-to-one rule, 241–242
Tie-tack microphone, 295–296
T-powering, 118
Transaural technology, 264–267
Two-way microphones, 87–89
Unbalanced microphone inputs,
Variable contact principle, 2–5
Variable pattern cardioid
microphones, 79–81
Variable-D dynamic microphones,
Vector representation, line
microphone, 93–94
Velocity microphone, 50–51
Virtual microphones, 331–333
Vocal microphones, 78
Vocals and vocal groups, 228–230
Watts, 10–11
Wavefront reconstruction, 267–270
Wind screens, 158–161
Wireless microphones, 142–151
frequency allocations, 142–143
power management, 149–150
pre and post-equalization,
signal companding, 145–146
usage protocols, 150–151
Woodwind instruments, 232
XY technique, 172–173
Zoom microphone, 100–102
3:12 PM
Page 378
Download PDF