Loudspeaker Driver Comparison PDF

AES STANDARDS: DRAFT FOR COMMENT ONLY
DRAFT AES-X223
STANDARDS AND
INFORMATION DOCUMENTS
Call for comment on

DRAFT
AES information document for
acoustics –
Loudspeaker driver comparison
chambers
This document was developed by a writing group of the Audio Engineering Society Standards Committee
(AESSC) and has been prepared for comment according to AES policies and procedures. It has been brought to
the attention of International Electrotechnical Commission Technical Committee 100. Existing international
standards relating to the subject of this document were used and referenced throughout its development.
Address comments by E-mail to standards@aes.org, or by mail to the AESSC Secretariat, Audio Engineering
Society, PO Box 731, Lake Oswego OR 97034. Only comments so addressed will be considered. E-mail is
preferred. Comments that suggest changes must include proposed wording. Comments shall be restricted to
this document only. Send comments to other documents separately. Recipients of this document are invited to
submit, with their comments, notification of any relevant patent rights of which they are aware and to provide
supporting documentation.
This document will be approved by the AES after any adverse comment received within six weeks of the
publication of this call on http://www.aes.org/standards/comments/, 2019-03-05 has been resolved. Any person
receiving this call first through the JAES distribution may inform the Secretariat immediately of an intention to
comment within a month of this distribution.
Because this document is a draft and is subject to change, no portion of it shall be quoted in any
publication without the written permission of the AES, and all published references to it must include a
prominent warning that the draft will be changed and must not be used as a standard.
AES STANDARDS: DRAFT FOR COMMENT ONLY

The AES Standards Committee is the organization responsible for the standards
program of the Audio Engineering Society. It publishes technical standards,
information documents and technical reports. Working groups and task groups with a
fully international membership are engaged in writing standards covering fields that
include topics of specific relevance to professional audio. Membership of any AES
standards working group is open to all individuals who are materially and directly
affected by the documents that may be issued under the scope of that working group.
Complete information, including working group scopes and project status is available
at http://www.aes.org/standards. Enquiries may be addressed to standards@aes.org
The AES Standards Committee is supported in part by those listed below who, as
STANDARDS Standards Sustainers, make significant financial contribution to its operation.
This list is current as of

2019/1/30
AES-73id-xxxx
DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY
Call for Comment on DRAFT

AES information document for
acoustics –
Loudspeaker driver comparison
chambers
Published by
Audio Engineering Society, Inc.
Copyright ©2019 by the Audio Engineering Society
Abstract
Acoustic performance measurements on loudspeaker drivers that are measured in different sites or with
different equipment or methods are subject to variations in accuracy and repeatability. Similarly, end-of-
line loudspeaker test systems often utilize small, arbitrarily shaped, driver measurement chambers that do
not correlate from line to line, or from site to site. This document is motivated by the need for engineers in
different locations to share accurate measurement data on specific or same type loudspeaker drivers and
addresses some of the factors affecting repeatability of measurements in different locations.
An AES standard implies a consensus of those directly and materially affected by its scope and provisions
and is intended as a guide to aid the manufacturer, the consumer, and the general public. The existence of
an AES standard does not in any respect preclude anyone, whether or not he or she has approved the
document, from manufacturing, marketing, purchasing, or using products, processes, or procedures not in
agreement with the standard. Prior to approval, all parties were provided opportunities to comment or
object to any provision. Attention is drawn to the possibility that some of the elements of this AES standard
or information document may be the subject of patent rights. AES shall not be held responsible for
identifying any or all such patents. Approval does not assume any liability to any patent owner, nor does it
assume any obligation whatever to parties adopting the standards document. Recipients of this document
are invited to submit, with their comments, notification of any relevant patent rights of which they are
aware and to provide supporting documentation. This document is subject to periodic review and users are
cautioned to obtain the latest edition.
Audio Engineering Society Inc., 551 Fifth Avenue, Room 1225, New York, NY 10176, US.
www.aes.org/standards standards@aes.org

2019-03-06 printing
AES-73id-xxxx
-2-
Contents
0 Introduction ...............................................................................................................................4
0.1 General.................................................................................................................................4
0.2 Patents .................................................................................................................................4
1 Scope .........................................................................................................................................4
2 Normative references ...............................................................................................................4
3 Definitions .................................................................................................................................5
4 Loudspeaker test chambers.....................................................................................................6
4.1 General.................................................................................................................................6
4.2 Definition ..............................................................................................................................6
5 General function........................................................................................................................6
5.1 Acoustic characteristics ........................................................................................................6
5.2 Usable bandwidth .................................................................................................................6
5.2.1 Calibration corrections ...................................................................................................6
5.2.2 High-frequency limit .......................................................................................................6
5.2.3 Low-frequency limit ........................................................................................................7
5.2.4 Measurement tolerance .................................................................................................7
5.2.5 Resonant modes ............................................................................................................8
5.3 Loudspeaker test chamber construction ...............................................................................8
5.3.1 Core geometry ...............................................................................................................8
5.3.2 Construction materials ...................................................................................................9
5.3.3 Acoustic Absorption .......................................................................................................9
5.3.4 Loudspeaker driver mounting.......................................................................................10
5.3.5 Microphone position .....................................................................................................10
5.4 Calibration ..........................................................................................................................11
5.4.1 Test chamber calibration..............................................................................................11
5.4.2 Frequency response ....................................................................................................11
Annex A - Theoretical Background...........................................................................................13
Boundary element verification ..................................................................................................13
Annex B – Practical Considerations.........................................................................................15
B.1 Reference Measurement....................................................................................................15
B.2 Tetrahedral measurements ................................................................................................16
Annex C – Discussion of Measurement Uncertainty...............................................................22
Annex D – Tetrahedral test chamber ........................................................................................25
Annex E – Bibliography.............................................................................................................26

2019-03-06 printing
AES-73id-xxxx
-3-
Foreword
This foreword is not part of AES-73id AES information document for acoustics – Loudspeaker driver
comparison chambers.
This document was developed in project AES-73id, in the SC-04-03 Working Group on Loudspeaker
Modelling and Measurement, under the leadership of Steve Hutt.
The members of the writing group that developed this document in draft are: Geoff Hill, Steve Hutt, David
Murphy, Ed Simon.
Steve Hutt
Chair, working group SC-04-03, 2019-02-28
Note on normative language
In AES standards documents, sentences containing the word “shall” are requirements for compliance with
the document. Sentences containing the verb “should” are strong suggestions (recommendations).
Sentences giving permission use the verb “may”. Sentences expressing a possibility use the verb “can”.

2019-03-06 printing
AES-73id-xxxx
-4-
DRAFT
AES information document for acoustics –
Loudspeaker driver comparison chambers
0 Introduction
0.1 General
In the design of loudspeaker drivers, numerous engineering acoustical tests are performed throughout the
development cycle to ascertain the loudspeaker driver’s performance characteristics and capabilities. And,
during the manufacturing phase end-of-line quality tests place each driver into a relatively compact test
chamber and measure the acoustic and electrical response based on a set of test signals. Results are
compared with stored values derived from a reference such as a sample that has typically been quantified
by the engineering acoustical tests or, with reference values derived from the mean capability of the
loudspeaker production variance. End-of-line tests are not usually set up to provide full performance
measurements however they can identify deviation and variance from the expected performance defined by
reference parameters. Such test chambers are typically built only to meet local needs. As a result, the
measurements derived from a chamber in one location do not compare well with measurements derived
from another. It would be beneficial for collaboration over different sites if the loudspeaker measurement
data was directly transferable.
This information document will set out some relevant issues and suggest loudspeaker test chamber designs
and construction methods to achieve measurement results that will be comparable between different testing
sites.
0.2 Patents
Attention is drawn to the possibility that some of the elements of this document may be the subject of
patent rights. AES shall not be held responsible for identifying any or all such patent rights.
1 Scope
This document considers factors affecting the interchangeability of measurement data from simple
loudspeaker comparison chambers and discusses some performance capabilities.
2 Normative references
The following standard contains provisions that, through reference in this text, constitute provisions of this
document. At the time of publication, the edition indicated was valid. All standards are subject to revision,
and parties to agreements based on this document are encouraged to investigate the possibility of applying
the most recent edition of the indicated standard.
ISO 266, Acoustics — Preferred frequencies for measurement. Geneva, Switzerland: International
Organization for Standardization, 1975
ISO 80000-8, Quantities and units — Part 8: Acoustics, Switzerland: International Organization for
Standardization, 2007
IEC 60268-5, Sound system equipment - Part 5: Loudspeakers, Switzerland: International Electrotechnical
Commission, 2007
IEC 60268-21, Sound system equipment - Part 21: Acoustical (output based) measurements, Switzerland:
International Electrotechnical Commission, 2018

2019-03-06 printing
AES-73id-xxxx
-5-
3 Definitions
For the purposes of this document, the following terms, definitions, and abbreviations apply.
3.1
Reference point
A point positioned with respect to some physical feature of the loudspeaker or loudspeaker system
NOTE The position shall be that specified by the manufacturer.
3.2
Reference axis
A line that passes through the reference point
NOTE The direction of the reference axis shall be specified by the manufacturer and shall
be used as the zero-reference axis for frequency response measurements.
3.3
Measuring axis
The line joining the microphone to the nominal centre of the loudspeaker under test
3.4
Frequency response
3.4.1
Magnitude response
The generated sound pressure level as a function of frequency, measured under free-field or simulated free-
field conditions, with a constant voltage source and at a stated position with respect to the reference axis
and point.
3.4.2
Transfer function
The ratio as a complex variable between generated sound pressure and loudspeaker input voltage as a
function of frequency, measured under free-field or simulated free-field conditions, with a constant voltage
source and at a stated position with respect to the reference axis and point
3.4.3
Internal response
The generated sound pressure level as a function of frequency, measured inside a test chamber, with a
constant voltage source and at a stated position with respect to the reference axis and point.
3.4.4
Correction response
The calculated or calibrated, sound pressure level as a function of frequency, such that if applied to a
measurement 3.4.3 will effectively correct such a measurement that it corresponds to a measurement 3.4.1
3.5
Amplitude smoothing
The modification of measured quantities by averaging on a mean-square basis over a band of frequencies
NOTE Amplitude smoothing can be done in successive bands by, for example, using third
octave bands per ISO 266 and their centre frequencies or in a continuous or sliding way
by using the FFT bin frequencies as centre frequencies. Enough frequency lines need to
be present within a given band for smoothing to be valid. This modification is often called
“frequency smoothing” or “bandwidth smoothing.”
3.6
Far field
The part of the radiated sound field more distant from the radiator in which the inverse-distance law (that is,
sound pressure is inversely proportional to distance) is obeyed
3.7
Very near field
The part of the radiated sound field, normally very close to the radiator typically a distance of 2 mm to
5mm maximum from the radiating surface

2019-03-06 printing
AES-73id-xxxx
-6-
4 Loudspeaker test chambers

4.1 General
Accurate and repeatable loudspeaker driver measurements require that environmental and boundary
influences have minimal impact on the measurement. This conventionally implies that loudspeaker driver
measurement methods rely both implicitly and explicitly upon the use of an anechoic environment that has
no impact on the measurement. Alternatively, measurements may employ a standard baffle or simulated
anechoic environment that is understood and can be replicated:
a. Small anechoic chambers are bandwidth limited, relative to dimensions vs. wavelength.
b. The cost or size of an anechoic chamber of adequate size may be prohibitive.
c. Anechoic chambers are impractical for line or production use.
d. Measurement methods utilizing "windowing" may restrict usage over the full audio band and are
difficult to control on a measurement-to-measurement basis, dependent on mechanical setup.
e. An anechoic chamber is often not available, often reserved for engineering development or
continuing conformance validation.
This document considers loudspeaker test chambers that produce measurements that may be replicated
predictably and repeatably.
4.2 Definition
A loudspeaker test chamber is defined in this document as a chamber that may be used for the measurement
of loudspeaker drivers. In the context of this document, the term 'test chamber' refers to an enclosure, not an
anechoic chamber. Typical enclosures may be rectangular, asymmetric with non-parallel walls, have
spherical walls, or any combination of these characteristics.
5 General function
5.1 Acoustic characteristics
Unlike a conventional anechoic chamber, the acoustic absorption in a loudspeaker test chamber is not
intended to absorb all the acoustic energies of a device under test:
(i) It is intended that the acoustic dampening be enough to minimize the medium frequency modes and the
higher frequencies produced by the device under test.
(ii) It is preferable to minimize conventional standing waves by design of the structure.
5.2 Usable bandwidth
5.2.1 Calibration corrections

Like an anechoic chamber it is envisaged that an acoustic tolerance be established. However, the
specification and tolerance of a loudspeaker test chamber may benefit from applying a calibration
correction which is normally applied by the test analyser either during the measurement cycle or may be
subsequently applied to the data.
In the same way that a microphone calibration file would be normally be applied, one could also apply a
calibration correction for a loudspeaker test chamber.
5.2.2 High-frequency limit

The high-frequency limit is determined by several factors:
(a) Microphone performance
(b) Analyzer performance
(c) Sampling rate/bandwidth
(d) Absorption and reflection coefficients
2019-03-06 printing
AES-73id-xxxx
-7-
(e) Absorption geometry

(e) High frequency scattering
(f) Bandwidth of the device under test
5.2.3 Low-frequency limit

The low-frequency limit is determined by many of the same factors plus others:
(a) Microphone performance
(b) Analyzer performance
(c) Sampling rate/bandwidth
(d) Absorption and reflection coefficients
(e) Absorption geometry
(f) Maximum SPL of the measurement system
(g) Front pressure or acoustic load provided to a loudspeaker by a test chamber
5.2.4 Measurement tolerance

Loudspeaker frequency response is often described with some tolerance in dB over a specified bandwidth
for example, +/-3 dB, 20 Hz to 20 kHz. Due to the complex nature of loudspeakers, measurement
equipment and environmental variances, repeatability of measurements with a small tolerance is difficult.
However, a deviation reference of +/-3 dB should not be considered adequate for comparison between
similar drivers measured in the same manner. An appropriate tolerance would be a range where error or
deviation does not significantly affect the results, for example +/-0.1 dB to 0.2 dB over a specified
frequency range. If mechanical setup is not strictly adhered to frequency dependent boundary influences
may increase the deviation.
Comparison of tolerance specifications for measurements made in different locations and with different
equipment should be defined by the user for specific devices. This will require careful documentation of
mechanical setup and environmental conditions, including temperature, humidity, and barometric pressure
that impact not just the measurement results but may also affect the driver under test. The user specification
should include all information necessary to replicate and process the measurements accurately.
The core requirement for any loudspeaker test chamber is in establishing whether the measurement system
is consistent, accurate and stable. Gage repeatability and reproducibility techniques and statistical tools
cover much of this in detail. However most of the analysis upon which this is based is of single dimensional
data, such as length, width and diameter. The primary concept is of statistical analysis together with normal
distribution of the dimension being measured.
Most of the Loudspeaker data measured in a test chamber is inherently multi-dimensional. For example,
amplitude (SPL) or impedance (Ohms) vs. frequency (Hz).
Data can be defined in an understandable form, with discrete and repeatable steps taking a hypothetical set
such as the example shown in table 1.
Table 1 - SPL vs Frequency data example
# 20Hz 40Hz 80Hz 160Hz 320Hz 640Hz 1,280Hz 2,560Hz 5,120Hz 10,240Hz 20,480Hz
1 50 62 74 86 87 88 89 90 95 87 80
2 45 65 75 86 88 90 88 90 91 88 77
3 55 67 79 81 83 85 87 88 94 85 79
4 47 68 83 80 79 80 81 92 95 86 76
5 51 63 75 87 88 89 90 91 96 88 79
An example method is to first overlay or superimpose all the data points and to exclude any samples whose
overall response curves significantly differ from the other samples. This may be done graphically, as
illustrated in figure 1 using the data from table 1.

2019-03-06 printing
AES-73id-xxxx
-8-
Figure 1 – Comparing data graphically to detect outlier devices
This may also be done by calculating the mean or average at each frequency and then isolating samples
based on the statistical deviation. In the example above #4 had a significantly lower level (apart from at
80Hz where it is much higher) this might indicate insufficient magnetisation. By examining the distribution
of values at each frequency, outlier data may be identified.
A useful check is to measure and plot the Impedance (Ohms) vs. Frequency (Hz) for the samples measured
at the same time and under the same conditions and to perform the same calculations and check for any
correlation between these individual measurements.
Then taking the samples without outliers calculate Min, Avg, Max, STDEV and 2 × STDEV for each
frequency.
The overall error margin can be calculated from the square root of the sum of the squares of the individual
errors at these frequencies with appropriate weighting for the octave or part octave distribution according to
Appendix C.
The key point of such an analysis should be to separate the response and errors due to the outlier
loudspeaker drive unit under test from that of the variability introduced by the test chamber and
measurement system.
By plotting both the amplitude and impedance vs frequency, along with any deviation of these
measurements, such data can then be used to reliably characterise the measurements and the measurement
system according to IEC 60268-5
5.2.5 Resonant modes

Resonant modes in a loudspeaker test chamber can be minimised by choosing geometry with few parallel
surfaces and those remaining should be covered with acoustic absorption to minimize resonant modes.
5.3 Loudspeaker test chamber construction
5.3.1 Core geometry

In the context of the scope of this document, critical attributes required to replicate measurements in
different locations require strict adherence to duplication of the loudspeaker test chamber. Volume and
shape may be arbitrarily chosen by coordinating parties, but preferred recommendations are available. For
example, in addition to the 1.35 m x 1.65 m “standard baffle”, IEC 60268-5 describes two standard
measuring enclosures essentially in the form of closed rectangular boxes:
Enclosure “A” volume ~600 l
Enclosure “B” volume ~450 l

2019-03-06 printing
AES-73id-xxxx
-9-
An alternative measurement environment which avoids the limitations of conventional enclosures is

described below (and see figure 3). The test chamber is tetrahedral in shape - the core structure of a
tetrahedral test chamber may be understood from viewing as a cube sliced diagonally. The construction and
dimensional relationships then become obvious from simple geometry.
Suggested tetrahedron sizes measured along the sides of the face, to cover the ranges of loudspeaker
diameters are shown in table 2:
Table 2 - Tetrahedron dimensions
Loudspeaker size Max. diameter Tetrahedron side

Miniature 30 mm 150 to 200 mm
Small loudspeaker sizes 100 mm 350 to 400 mm
Medium 300 mm 500 to 900 mm
Large 600 mm 1.2 to 2.0m
5.3.2 Construction materials

It is essential that the construction be strong, rigid and non-resonant in the frequency range to be tested.
Some minimum material requirements for different size loudspeakers are proposed in table 3:
Table 3 - Suggested construction material
Loudspeaker sizes Max. diameter Material thickness

Miniature 30 mm 12 mm
Small loudspeaker sizes 100 mm 18 mm
Medium 300 mm 24 mm
Large 600 mm 2 x 24 mm
Table 3 assumes the material used will be high-density plywood. Denser, stronger or heavier materials such
as aluminium may also be used. Laminates constructed of different materials may be a useful technique to
minimize panel resonances.
5.3.3 Acoustic Absorption

Unlike conventional anechoic test chambers, the intention with the acoustic absorption in this case is not to
attempt total absorption at all frequencies.
At low frequencies the wavelengths are very long - 20 m or more. This naturally sets a practical limit as to
how low a frequency can be effectively absorbed given practical material dimensions.
Conversely higher frequencies are readily absorbed and just 100 mm of material often has a typical sound
absorption coefficient of greater than 0.5 above 500 Hz.
The intention is to provide enough absorption to minimise higher order modes of standing waves but not
diminish the first mode. Absorbing material should be applied to all surfaces except the baffle on which the
test loudspeaker is mounted
As an example, a curve is shown as figure 2 below, of a sample test chamber measured without and with
approximately 50 mm of open cell foam.

2019-03-06 printing
AES-73id-xxxx
- 10 -
Figure 2 – Internal Measurement without and with damping of a sample test chamber
This clearly shows reduced effective damping at low frequencies, with simultaneously much greater effect
at higher frequencies.
Figure 3 below shows the basic shape of a tetrahedral test chamber. The measurement microphone is inside
the tetrahedron at the apex and points along the driver axis. The loudspeaker driver under test will be
mounted on the face of the tetrahedron with its open front facing inward.
Figure 3 - Tetrahedral enclosure
5.3.4 Loudspeaker driver mounting

An ideal driver mounting method will have minimal effect on loudspeaker performance. With respect to
replicating measurements on same type loudspeaker drivers it is imperative to define and describe the
mounting so that it may be duplicated in subsequent tests. For convenience, a test chamber designed for a
larger driver may use a sub-baffle to mount a smaller driver.
5.3.5 Microphone position

The measurement microphone should be positioned on an axis to the driver under test at specified position
relative to the test enclosure. The microphone shall be mounted at a fixed distance from the rear of the
baffle, with typical distances being shown in table 4.
2019-03-06 printing
AES-73id-xxxx
- 11 -
Table 4 - Microphone-to-baffle distances
Max. cone diameter Min. microphone-to-baffle distance Calibration SPL ref 1m

30 mm 31.6 mm +30 dB
100 mm 100 mm +20 dB
300 mm 316 mm +10 dB
600 mm 1000 mm 0 dB
1000 mm 3160 mm -10 dB
NOTE depending upon the particular microphone design, construction and the
performance of high frequency transducers it is possible to set up a very strong reflection
between the face of the microphone and the transducer if these are perfectly parallel. This
can result in significant, closely spaced, variations in response at high frequencies. In this
case the microphone may be deliberately moved off axis by less than 5% to alleviate the
problem.
5.4 Calibration
5.4.1 Test chamber calibration

Calibration of a loudspeaker test chamber can be performed in three primary ways:
(a) Calculation: finite element analysis (FEA) or boundary element modelling (BEM) programs may be
used to calculate required calibration data.
(b) Calibration by measurement: if a loudspeaker with a known directional response is positioned on the
loudspeaker test chamber, and the response measured at the microphone, a calibration data set may be made
to correlate the test result with that of the known directional response.
(c) Pistonic measurement: most loudspeaker transducers are pistonic at low frequencies. The center and the
periphery of the cone is therefore moving in step with each other. Pistonic means that all parts of the cone
or diaphragm move with the same displacement and velocity
The front of the cone faces into the loudspeaker test chamber, so it is effectively working in a pressure
environment at low frequencies. We can therefore measure the pressure response of the loudspeaker
together with the axial response at higher frequencies.
The rear of the cone is outside the loudspeaker test chamber, radiating into free air, and its response at low
frequencies is the response of the driver in an enclosure of an equivalent volume. Using a separate
microphone at the rear of the driver in the very near field (3.7), we can therefore make a measurement of
the near field pistonic response which, we known from Keele, is accurate at low frequencies.
A difference function of these two measurements, pressure and near field pistonic, may be combined to
form the calibration curve at low frequencies. Care must be taken to ensure that the microphone and
preamplifier responses are matched in the frequency region of interest. Once made, the calibration of the
enclosure will subsequently be stable.
This is discussed in more detail in annex B.
5.4.2 Frequency response

An overlay of normalized windowed IEC baffle data versus the same loudspeaker driver measured in a
tetrahedral shaped loudspeaker test chamber is shown in figure 4. The tetrahedral chamber is discussed in
more detail in annex B and the bibliography.

2019-03-06 printing
AES-73id-xxxx
- 12 -
Figure 4 - Comparison of windowed IEC baffle (blue) & corrected test chamber (red) data

2019-03-06 printing
AES-73id-xxxx
- 13 -
Annex A - Theoretical Background

Boundary element verification
A boundary element model of a tetrahedral test chamber model TTC 350 was made using a SEAS H1207
driver and ABEC 3 software from R&D Team Software Development (see bibliography).
Figure A.1 - Boundary element model
Simulated on the inside tetrahedral on the microphone plane. A field plot at 300 Hz shows that all regions
are at nearly the same SPL (see figure A.2).
Figure A.2 - Field plot at 300 Hz

2019-03-06 printing
AES-73id-xxxx
- 14 -
Figure A.3 - SPL response
Examination of the simulated sound pressure level plot shows that below 400 Hz the system works as a
pressure environment, above this frequency there are some modes, which are effectively controlled by the
acoustic absorption.

2019-03-06 printing
AES-73id-xxxx
- 15 -
Annex B – Practical Considerations

In the following measurements, practical verification was made using a SEAS H1207 loudspeaker driver.
B.1 Reference Measurement

A large IEC baffle was laid out to form a ground plane with the microphone - a half-inch microphone type
NTi 2010 - at 1 m directly above the loudspeaker driver’s central axis. This arrangement is shown in figure
B.1.
Figure B.1 - IEC baffle with microphone
Figure B.2 - IEC baffle, unwindowed responses
Significant reflections start after 10 to 12 ms, with the first major impulse reflection starting at around
13 ms. Frequencies lower than what corresponds to this time period should be considered unreliable.

2019-03-06 printing
AES-73id-xxxx
- 16 -
Figure B.3 - IEC baffle, windowed response
This clearly removes most of the reflection artefacts, especially above 500 Hz, where this result can
justifiably be considered as anechoic. In theory a 12 ms window should be reasonably accurate below
160 Hz and is often considered as useful down to 80 Hz.
B.2 Tetrahedral measurements

The same loudspeaker was then measured in a test chamber model TTC900. The microphone is at 316 mm
from the internal face of the test chamber.
As figure B.4 shows, there is a considerable rise in energy at low frequencies, this can also be seen in the
lower time trace. However, the response from 2 kHz upward looks to be quite close to the ground plane
measurement.
Figure B.4 - Tetrahedron, internal SPL response

2019-03-06 printing
AES-73id-xxxx
- 17 -
Most loudspeaker drivers are predominantly pistonic at lower frequencies the pressure at the front of a
loudspeaker diaphragm is 180 degrees out of phase with the rear of the diaphragm. So, in the region where
this applies, we can use this fact by measuring the diaphragm output outside the enclosure in the very near
field (2 to 5 mm from the diaphragm) as is shown in figure B.5. below. The response follows very closely a
simulated response at least up to approximately 1,700 Hz which corresponds to the theoretical pistonic
range approximating to the depth of cone 10 mm and diameter of around 100 mm. This measurement is
only required during initial calibration of the loudspeaker test chamber (see 5.3, figure B.9) and is not
needed during routine measurements.
Figure B.5 - Tetrahedron, external very-near-field response
From the work of Richard Small and Don Keele Jr. (see bibliography), we can use the low frequency
response in the near field to accurately predict the far field response of a loudspeaker in a given enclosure.
Thus, by combining the work of these we can accurately predict a low frequency correction curve.
An inverse correction curve can be generated by subtracting the External Curve (figure B.5) from the
Internal Curve (figure B.4) over this frequency range. Although the nearfield microphone and internal
microphones will be at different distances from the loudspeaker diaphragm experience with this type of test
chamber has shown that phase is not always required in the calculation.

2019-03-06 printing
AES-73id-xxxx
- 18 -
Figure B.6 - Overlaid Internal (Blue) External (Red) data
We take the difference between these two curves; the exact details differ between various measurement
systems. This can be achieved in a spreadsheet by dividing these two curves. In practice this means
subtracting the dB levels at each frequency in turn.
Figure B.7 - Rough correction curve
The resulting correction curve shown in figure B.7. is reasonably clean at low frequencies and is showing a
considerable 45 dB rise at low frequencies, however it is clearly completely inaccurate at high frequencies.
We deal with this by initially setting the levels above approximately 1,300 Hz to be equal. This may also be
used as the 0dB reference level. The exact method will depend upon the measurement system used.

2019-03-06 printing
AES-73id-xxxx
- 19 -
Figure B.8 – Final Correction Curve
This then becomes the final correction or equalization curve shown in figure B.8. This can be directly
applied to the internal SPL curve either as a part of a microphone correction or as a post process.
Figure B.9 - Final response
As seen in the frequency response plot of figure B.9, the low frequency rise has been eliminated. This is
confirmed by the corresponding impulse response which now tails off very rapidly, indicating at least
theoretically that we can use this technique to measure accurately to very low frequencies.

2019-03-06 printing
AES-73id-xxxx
- 20 -
Figure B.10 - Comparison, IEC baffle (Blue) and tetrahedral chamber (Red)
An overlay of normalized windowed IEC baffle data versus the same loudspeaker driver measured in a
tetrahedral shaped loudspeaker test chamber is shown in figure B.10. This shows correspondence
considering one measurement is made at approximately 1m but without tight control of the measurement
axis and the other is at 316mm in a TTC900.
Figure B.11 directly compares the windowed data from the IEC baffle (figure B.3) with the final result
from a tetrahedral test chamber (figure B.9). Here we can clearly see the close correspondence between
these two data sets, it also shows the low-frequency errors arising from windowing, diffraction from the
IEC baffle edges and cancellation.
Figure B.11 - Comparison of windowed IEC baffle (blue) & corrected TTC750 (red) data
The low frequency response is accurate and the high frequency performance is also reliable as this is
effectively an anechoic measurement, albeit measured at closer distance (0.316 mm).

2019-03-06 printing
AES-73id-xxxx
- 21 -
Figure B.12 - Comparison between three measurements of the same driver
Red trace IEC baffle

Green trace TTC350 (drivers up to 100 mm)
Blue trace TTC750 (drivers up to 200 mm)
Figure B.12 compares results of the same device measured on an IEC baffle, and two tetrahedral chambers
of different sizes. The three coloured traces show normalized data using the same SEAS H1207 driver.

2019-03-06 printing
AES-73id-xxxx
- 22 -
Annex C – Discussion of Measurement Uncertainty

As mentioned in paragraph 5.2.4 that environmental conditions have a significant effect on the repeatability
of electro-acoustic measurements. More details can be found in the AES Convention paper Ambient
Atmospheric Conditions and Their Influence on Acoustic Measurements: by Peter J Chapman
Clearly the atmospheric conditions, temperature, humidity and barometric pressure must be noted as part of
the data of any measurement of a device under test, so that appropriate corrections can be made when
measured in another set of conditions or another part of the world.
IEC 60268-21 Appendix B.7, uncertainty analysis, lists the critical elements in the measurement chain,
commencing with the generator accuracy, the microphone calibration uncertainty, the microphone
preamplifier uncertainty, the analysis rms detector accuracy, background noise, positioning errors, and
temperature, humidity and barometric pressure. Each of these elements has an uncertainty associated with
it, and total standard uncertainty is calculated from the square root of the sum of the squares of the
individual uncertainties.
It is unlikely that any electro-acoustic measuring system will be identical to another, so each system must
have its uncertainty calculated. One important physical uncertainty derives from the physical position of the
device under test relative to the measuring microphone. It is customary to carry out a ‘positioning test’ i.e.
three different operators each make ten measurements, removing and replacing the device under test each
time. The results are then averaged, and the standard deviation calculated. The average plus and minus this
standard deviation includes 68% of the population, and the average plus and minus two standard deviations
includes 95% of the population.
The numeric value of two standard deviations is called the expanded measurement uncertainty with a
coverage factor of 2. It is either expressed as a percentage of the measurement or as a derived unit of the
measurement such as decibels.
The figure below shows results of 100 measurements of the same driver measured without disturbance to
determine the measurement system variability. The measurement system comprised a Behringer ECM8000
mounted in a TTC900 feeding a Klippel QC analyser measuring a SEAS L12RCY loudspeaker drive unit.
The measurements and uncertainty over plotted, and a table showing the 100 values of SPL averaged
between 500 Hz and 1 kHz, and the overall average and standard deviation are shown.
In this case we can state that the measurement uncertainty is below +/-0.2 dB.
Figure C.1 – 100 Measurements of the same loudspeaker drive unit

The uncertainty analysis of the test chamber must be done for each combination of loudspeaker and sub
baffle, as the closeness of fit of the mounting conditions may vary within loudspeaker units of the same
nominal diameter depending on the shape of the mounting flange.
A small section of the spreadsheet is shown as table C.1 below:

2019-03-06 printing
AES-73id-xxxx
- 23 -
Table C.1 - Uncertainty analysis example
Frequency (Hz) MIN AVG MAX Std Dev. 2 * Std Dev.

505 112.45 112.55 112.64 0.04 0.07
523 112.48 112.57 112.65 0.04 0.07
541 112.54 112.60 112.67 0.03 0.06
S60 112.51 112.57 112.62 0.02 0.04
58O 112.44 112.52 112.55 0.02 0.04
600 112.38 112.46 112.50 0.02 0.04
622 112.32 112.39 112.45 0.02 0.04
644 112.28 112.35 112.39 0.02 0.04
667 112.34 112.38 112.44 0.02 0.04
690 112.34 112.40 112.48 0.02 0.04
714 112.22 112.29 112.39 0.03 0.05
740 112.05 112.14 112.25 0.03 0.06
766 111.93 112.03 112.11 0.03 0.06
793 112.03 112.14 112.20 0.03 0.06
821 112.27 112.35 112.42 0.03 0.05
850 112.43 112.49 112.54 0.02 0.05
880 112.48 112.55 112.60 0.02 0.05
911 112.47 112.53 112.60 0.02 0.04
943 112.46 112.50 112.56 0.02 0.04
976 112.44 112.49 112.53 0.02 0.03
1,011 112.35 112.38 112.42 0.02 0.03
An analysis was then conducted on a series of six loudspeaker drivers measured sixty-eight times over the
course of a day with the results shown in figure C.2. The temperature variation during the day was not
monitored in this series of tests and may have contributed to the variation measured.
Figure C.2 – Measurements of the six individual loudspeaker drive units

2019-03-06 printing
AES-73id-xxxx
- 24 -
This shows an increased uncertainty more than 10 times that of the 100 individual measurements, so it is
very likely that the variation shown is that of the loudspeakers under test, rather than that of the
measurement system. This example demonstrates that a significant improvement over traditional
measurement systems can be achieved.

2019-03-06 printing
AES-73id-xxxx
- 25 -
Annex D – Tetrahedral test chamber

The core of a tetrahedral test chamber is formed of three walls - right-angle triangles - together with an
equilateral triangle forming the main baffle. These three walls could be formed from many materials,
however to meet the design concept outlined here these three sides should be at a 90-degree angle to each
other.
Microphone
Baffle opening
Figure D.1 - Tetrahedral test chamber
The structural elements need to be mechanically strong to withstand the physical and acoustical stresses but
should be non-resonant to prevent intrusion on the measurements being made. Also, they should be
acoustically reflective at low frequencies.
The structural requirements of the baffle and sub baffle should allow interchangeability between baffles and
sub baffles of the various sizes.
A basic version could be formed with a triangular baffle fitted into a corner of a room with the structure
formed of brick or concrete. Great care must be taken to render the baffle completely air tight.

2019-03-06 printing
AES-73id-xxxx
- 26 -
Annex E – Bibliography
Richard H. Small, Simplified Measurements at Low Frequencies, JAES Volume 20 Issue 1 pp. 28-33;
February 1972, Audio Engineering Society, NY., US. http://www.aes.org/e-lib/browse.cfm?elib=2103
Don B. Keele, Low-Frequency Loudspeaker Assessment by Nearfield Sound-Pressure Measurement, JAES
Volume 22 Issue 3 pp. 154-162; April 1974, Audio Engineering Society, NY., US. http://www.aes.org/e-
lib/browse.cfm?elib=2774
Alan S Phillips, Measuring the True Acoustical Response of Loudspeakers, SAE Technical Paper 2004-01-
1694, 2004, doi:10.4271/2004-01- 1694. http://papers.sae.org/2004-01-1694/
HOLMImpulse software http://www.holmacoustics.com/holmimpulse.php
ABEC 3 and VACS software from R&D Team Software Development http://www.randteam.de
Geoff Hill, Consistently Stable Loudspeaker Measurements Using a Tetrahedral Enclosure Engineering
brief, presented 16 October 2013 at the 135th Convention of the Audio Engineering Society in New York.
http://www.aes.org/e-lib/browse.cfm?elib=16958
Geoff Hill, Comparative Results between Loudspeaker Measurements Using a Tetrahedral Enclosure and
Other Methods, Engineering brief, Presented on 25th April 2014 at the 136th Convention of the Audio
Engineering Society in Berlin. http://www.aes.org/e-lib/browse.cfm?elib=17152
Peter John Chapman Ambient Atmospheric Conditions and Their Influence on Acoustic Measurements AES
Convention: 136 (April 2014) Paper Number:9061 http://www.aes.org/e-lib/browse.cfm?elib=17208
Peter John Chapman, Quantifying Acoustic Measurement Tolerances and Their Importance in the
Loudspeaker Supply Chain AES Convention: 136 (April 2014) Paper Number: 9056 http://www.aes.org/e-
lib/browse.cfm?elib=17203
IEC 60268-22, Sound system equipment - Part 22: Electrical and mechanical measurements on
transducers, Switzerland: International Electrotechnical Commission, (Currently in development)

2019-03-06 printing

Loudspeaker Driver Comparison PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Loudspeaker Driver Comparison PDF

Uploaded by

Copyright:

Available Formats

AES STANDARDS: DRAFT FOR COMMENT ONLY

Call for comment on

AES STANDARDS: DRAFT FOR COMMENT ONLY

This list is current as of

Call for Comment on DRAFT

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Note on normative language

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

4 Loudspeaker test chambers

5.2 Usable bandwidth

5.2.1 Calibration corrections

5.2.2 High-frequency limit

(e) Absorption geometry

5.2.3 Low-frequency limit

5.2.4 Measurement tolerance

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Figure 1 – Comparing data graphically to detect outlier devices

5.2.5 Resonant modes

5.3 Loudspeaker test chamber construction

5.3.1 Core geometry

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

An alternative measurement environment which avoids the limitations of conventional enclosures is

Loudspeaker size Max. diameter Tetrahedron side

5.3.2 Construction materials

Loudspeaker sizes Max. diameter Material thickness

5.3.3 Acoustic Absorption

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Figure 3 - Tetrahedral enclosure

5.3.4 Loudspeaker driver mounting

5.3.5 Microphone position

Table 4 - Microphone-to-baffle distances

Max. cone diameter Min. microphone-to-baffle distance Calibration SPL ref 1m

5.4.1 Test chamber calibration

5.4.2 Frequency response

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Annex A - Theoretical Background

Figure A.1 - Boundary element model

Figure A.2 - Field plot at 300 Hz

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Figure A.3 - SPL response

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Annex B – Practical Considerations

B.1 Reference Measurement

Figure B.1 - IEC baffle with microphone

Figure B.2 - IEC baffle, unwindowed responses

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Figure B.3 - IEC baffle, windowed response

B.2 Tetrahedral measurements

Figure B.4 - Tetrahedron, internal SPL response

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Figure B.5 - Tetrahedron, external very-near-field response

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Figure B.6 - Overlaid Internal (Blue) External (Red) data

Figure B.7 - Rough correction curve

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

Figure B.8 – Final Correction Curve

Figure B.9 - Final response

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY

DRAFT FOR COMMENT ONLY - DRAFT FOR COMMENT ONLY