Professional Documents
Culture Documents
Published by
Audio Engineering Society, Inc.
Copyright 2001 by the Audio Engineering Society
Abstract
This document provides guidelines for characterizing the uniformity of scattering produced by surfaces from
measurements or predictions of scattered polar responses.
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. An AES information document
is a form of standard containing a summary of scientific and technical information; originated by a technically
competent writing group; important to the preparation and justification of an AES standard or to the understanding
and application of such information to a specific technical subject. An AES information document implies the same
consensus as an AES standard. However, dissenting comments, if any, may be published with the document.
The existence of an AES standard or AES information document 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 conforming to the standard. 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.
This document is subject to periodic review and users are cautioned to obtain the latest edition and printing.
AES-4id-2001
–2–
Contents
Foreword................................................................................................................................................................3
1 Scope...................................................................................................................................................................4
2 Normative references..........................................................................................................................................4
3 Definitions ..........................................................................................................................................................4
4 Measurement of polar response ..........................................................................................................................7
4.1 Measurement environment ..............................................................................................................................7
4.2 Measurement field ...........................................................................................................................................9
4.3 Test sample ....................................................................................................................................................11
4.4 Measuring techniques ....................................................................................................................................11
4.5 Polar response processing ..............................................................................................................................13
5 Prediction techniques........................................................................................................................................16
6 Extraction of coefficients..................................................................................................................................16
6.1 Directional diffusion coefficient ....................................................................................................................16
6.2 Calculation of area factors .............................................................................................................................17
6.3 Diffusion coefficient ......................................................................................................................................18
7 Presentation of results .......................................................................................................................................18
8 Test report.........................................................................................................................................................18
Annex A Qualification of a measurement space..................................................................................................19
A.1 Space criteria.................................................................................................................................................19
A.1.1 Sound pressure level deviation ..................................................................................................................19
A.1.2 Signal-to-noise ratio...................................................................................................................................19
A.2 Environments ................................................................................................................................................19
A.2.1 Anechoic chamber .....................................................................................................................................19
A.2.2 Non-anechoic space ...................................................................................................................................19
A.2.3 Placement...................................................................................................................................................19
Annex B Informative references..........................................................................................................................20
2001-02-19 printing
AES-4id-2001
–3–
Foreword
[This foreword is not a part of AES information document for room acoustics and sound reinforcement systems—
Characterization and measurement of surface scattering uniformity, AES-4id-2001.]
This document was prepared by a writing group of the SC-04-02 Working Group on Characterization of Acoustical
Materials of the SC-04 Subcommittee on Acoustics in partial fulfillment of project AES-X06, Measuring and
Modeling Acoustical Materials: Specification of Measuring Methods, Computer Models, and Transportable
Computer Data Files Compatible with Room Simulators and Auralizers. The project was initiated in 1994.
Trevor Cox headed the writing group for the final draft in cooperation with the International Organization for
Standardization (ISO) TC 43 SC 2 Building Acoustics working group (WG25) under AES project AES-X81.
2001-02-19 printing
AES-4id-2001
–4–
1 Scope
This document provides guidelines for characterizing the uniformity of scattering produced by surfaces from
measurements or predictions of scattered polar responses. In this context, the surface scattering is quantified in
terms of a single diffusion coefficient. The diffusion coefficient is a measure of quality designed to be used by
producers and users of surfaces that, either deliberately or accidentally, diffuse sound. It is also intended for use
when needed by developers and users of geometric room acoustic models. The diffusion coefficient is not intended,
however, to be blindly used as an input to current diffusion algorithms in geometric room acoustic models. The
diffusion coefficient characterizes the sound reflected from a surface in terms of the uniformity of the scattered
polar distribution. The information document details a free-field characterization method.
2 Normative references
No standards contain provisions that, through reference in this text, constitute provisions of this document.
3 Definitions
For the purposes of this standard, the following definitions apply.
3.1
reference flat surface
plane, rigid, and thin surface, with the same projected shape or footprint as the test surface
3.2
reference normal
outward-pointing vector perpendicular to the front face of the reference flat surface
3.3
reference point
geometric center of gravity of the reference flat surface
3.4
sound ray
line following one possible direction of sound propagation from a source point
3.5
specular reflection
incident sound ray that undergoes specular reflection such that Snell's law (that is, the angle of reflection equals
the angle of incidence) is obeyed, when the wavelength of sound is small compared to the dimensions of the
reference flat surface
NOTE Following Fermat’s principle the actual path between source and receiver via the panel will be
the one that is traversed in the least time.
2001-02-19 printing
AES-4id-2001
–5–
3.6
specular zone
area contained by imaginary lines constructed from the image source, which is created about the plane of the
reference flat surface via the edges of the reference flat surface to the receiver arc
NOTE The position at which an imaginary line from the image source to a receiver crosses the diffuser
is the specular reflection point. See Figure 1.
SOURCE
SPECULAR ZONE
α α
β β
REFERENCE FLAT
RECEIVER ARC SURFACE
IMAGE SOURCE
3.7
far field
part of the reflected sound field more distant from the test surface in which the inverse-distance law (that is, sound
pressure is inversely proportional to distance) is obeyed, as opposed to the near field, where the angular field
distribution is dependent on the distance from the radiator
3.8
single-plane diffuser
surface that displays distinct anisotropic behavior, as can be the case for a cylinder or a one-dimensional Schroeder
diffuser
NOTE For these surfaces the diffusion is measured in the plane of maximum diffusion. See figure 2.
2001-02-19 printing
AES-4id-2001
–6–
3.9
hemispherical diffuser
surface that is expected to display more approximately isotropic behavior, where hemispherical evaluation is
needed with a single diffusion coefficient
3.10
semicircular polar response
sound pressure level created by energy scattered from the surface as a function of angle measured about the
reference normal, generated under free-field or pseudo-free-field conditions, in a specified plane, on a semicircle
centered at the reference point, at a radial distance consistent with guidelines in 4.2
3.11
hemispherical polar response
sound pressure level scattered from the surface as a function of spherical coordinates measured about the reference
normal, generated under free-field or pseudo-free-field conditions, on a hemisphere centered at the reference point
3.12
complete diffusion
a) condition that occurs with a single-plane diffuser when all receivers on a semicircular polar response receive the
same reflected sound pressure
b) condition that occurs with a hemispherical diffuser when the same sound energy is scattered from the surface
into all solid angles
NOTE Complete diffusion occurs when the polar responses are completely uniform.
3.13
directional diffusion coefficient
dθ
measure of the uniformity of diffusion produced by a surface for one source position
NOTE dθ is bounded between 0 and 1. When complete diffusion is achieved by the surface, the diffusion
coefficient is 1. If only one receiver receives non-zero scattered sound pressure, the diffusion coefficient
is 0. The subscript θ is used to indicate the angle of incidence relative to the reference normal of the
surface.
2001-02-19 printing
AES-4id-2001
–7–
3.14
diffusion coefficient
dθ1, θ2, …, θn
measure of the uniformity of diffusion for more than one source position
NOTE A mean of the directional diffusion coefficients for the different source positions is used to
calculate the diffusion coefficient. The subscripts θ1, θ2, … θn are used to indicate the angles of incidence
relative to the reference normal of the surface used.
3.15
random incidence diffusion coefficient
d
measure of the uniformity of diffusion for a representative sample of sources over a complete semicircle for a
single-plane diffuser, or a complete hemisphere for a hemispherical diffuser
NOTE A mean of the directional diffusion coefficients for the different source positions is used to
calculate the diffusion coefficient. A guideline to achieve a representative sample of sources is given in
4.2.1. The lack of a subscript for d indicates random incidence.
3.16
diffusion bandwidth
frequency range over which significantly more diffusion is achieved when compared to the reference plane surface
NOTE In terms of diffusion coefficient, the value should exceed that achieved by the reference flat
surface by at least 0,1.
3.17
transfer function
ratio of the sound pressure at the receiver to the sound pressure produced by the source measured at 1 m under
free-field conditions
NOTE 1 Other source references are appropriate, provided there is a fixed linear relationship between the
sound pressure at 1 m and the reference, for example, the input voltage to a moving-coil loudspeaker
driven by a constant-voltage source.
NOTE 2 An inverse Fourier transform is applied to the transfer function to obtain the impulse response.
3.18
physical scale ratio 1:N
ratio of any linear dimension in a physical scale model to the same linear dimension in full scale
2001-02-19 printing
AES-4id-2001
–8–
Large non-anechoic spaces may be used to simulate a reflection-free environment according to annex A.
Boundary measurements may also be carried out to remove the necessity for a space to be anechoic in one plane,
provided conditions in annex A are satisfied. An implementation of such a setup is illustrated in figure 5.
2001-02-19 printing
AES-4id-2001
–9–
Scale models may be used to evaluate the diffusion from test surfaces. The sound wavelength used in a scale model
shall conform to the physical scale ratio. If the speed of sound in the model is the same as in full scale, then the
frequencies used for the model measurements shall be a factor of N higher than in full scale. For scale models the
absorption properties shall be the same for both the full-scale surface at full-scale frequencies and the test surface
at the equivalent model-scale frequency. When considering absorption from samples, losses due to viscous
boundary layer effects shall be included. This inclusion can limit the usable model scales.
4.2.1.2 An exception to 4.2.1.1 is the application of the diffuser only for far-field sources and receivers, in which
case diffusion coefficient measurements may be undertaken only in the far field.
4.2.1.3 When comparing test surfaces, the same geometry shall be used in each case to avoid errors. Full geometry
information, source locations, receiver positions, and test-surface dimensions and construction shall be quoted in
reports.
1) r >> Dmax
2) r / Dmax >> Dmax / λ
3) r = 2r1r2 / (r1 + r2)
where
NOTE 4.2.3 is a far-field criterion for on-axis scattering. The near field can extend significantly further
for oblique sources and receivers.
4.2.3.1 True far-field conditions can require very large measuring distances, further than can be realistically
achieved for many test geometries. If true far-field conditions cannot be achieved, at least 80 % of the receiver
positions shall be outside the specular zone. See figure 6. If achievable within the above constraints, the source-to-
reference-point distance should be 10 m and the receiver’s semicircle or hemisphere should have a radius of 5 m.
2001-02-19 printing
AES-4id-2001
–10–
1,00
80 % of receivers
outside specular zone
0,75
Diffusion
0,50
0,25
0,00
0,1 1 10 100
Receiver arc radius (m)
Plane surface Rectangular battens
4.2.3.2 Measurements shall be made with a maximum receiver angular resolution of 5°. This may be achieved using
either a discrete fixed position system or a continuous moving system. Figure 7 illustrates the effect of changing
receiver angular resolution.
1,0
0,8
Diffusion
0,6
5 deg
1 deg
0,4
0,2
0,0
100 1000 10000
Frequency (Hz)
Figure 7 — Effect of changing spacing between receivers on the single-plane diffusion from a
plane surface
4.2.3.3 To obtain a random-incidence diffusion coefficient, source positions should be measured with a maximum
angular separation of 10°, covering a semicircle or hemisphere measured about the reference normal. Where time
is limited by the conditions of the measurement, the directional diffusion coefficients should be obtained for normal
and 55° angles of incidence.
2001-02-19 printing
AES-4id-2001
–11–
4.3.2.2 For large sample test surfaces, the choice in 4.3.2.1 can be impossible due to the measurement distance
required as outlined in 4.2.1. In such cases, the following scheme may be applied for reducing the test surface size.
For a periodic test surface at least four complete repeat sequences and if possible more periods should be included,
so that the lobing effects of repetition can approach reality as closely as possible. Figure 8 illustrates the effects that
increasing periodicity has on the diffusion coefficient. For aperiodic or random surfaces, representative samples
of the surface roughness should be tested. These samples should be large enough so that surface effects rather than
edge effects are more prominent in the scattering. The geometry recommended for this measurement in this
information document is applicable to frequencies up to and including the 5-kHz third-octave band full scale.
1,0
0,8
Autocorrelation
0,6
0,4
0,2
0,0
0 2 4 6 8 10
Number of periods
2001-02-19 printing
AES-4id-2001
–12–
4.4.2 Source
A loudspeaker that illuminates the entire diffuser as if it were an omnidirectional source should be chosen. The
loudspeaker shall create a sound pressure distribution equivalent to what would be expected from a true
omnidirectional source, to within ± 2 dB in magnitude and ± 20° in phase, over the reference flat surface.
Microphones should have the same sensitivity to all conceivable reflection paths from the diffuser directly to the
microphone, to within ± 1 dB in magnitude and ± 10° in phase. Figure 9 illustrates the effect of changing the source
from being omnidirectional.
1,0
0,8
Diffusion
0,6 Omnidirectional
3 dB, 30 deg
0,4 2 dB, 20 deg
0,2
0,0
100 1000 10000
Frequency (Hz)
4.4.3 Tests
For each source and receiver pair the following measurements shall be made:
The transfer function may be obtained as an equivalent measure to the impulse response.
2001-02-19 printing
AES-4id-2001
–13–
2001-02-19 printing
AES-4id-2001
–14–
2001-02-19 printing
AES-4id-2001
–15–
4.5.1 Reflections
The influence of background reflections shall be removed by subtracting h1(t) from h2(t).
where
4.5.3 Windowing
A rectangular window shall be applied to the impulse response. The window has unity gain where the test surface
reflections are present, and is zero elsewhere. The window removes residual reflections clearly separated from the
test surface reflections in time. The size of the window shall be determined by considering the shortest and longest
times for reflections from the surface. The shortest time may be calculated from simple geometric consideration.
For the longest possible reflection path, more than the first-order reflections from the surface shall be considered
to ensure that the full test surface response is obtained. Visual inspection of the impulse response may be used to
ascertain the window location and size.
Within the windowed section of the impulse response, a signal-to-noise ratio of at least 40 dB should be obtained
for the reference flat surface. In such a case, noise not only refers to random measurement error, but also to bias
due to unwanted reflections from objects other than the surface being measured. This signal to noise ratio may be
obtained by calculating the following:
[h1 (t ) − h2 (t ) ]2dt
∫
S / N = 10log10 T
∫T h2 (t )dt
2
where
The signal-to-noise ratio should be tested in each of the third-octave bands being measured.
4.5.5 Corrections
If a range of source and receiver distances r1 and r2 are used, then the measured levels should be corrected to allow
for spherical or cylindrical spreading.
2001-02-19 printing
AES-4id-2001
–16–
In terms of the combined distance r from 4.3, and normalizing to standard distances of r1 = 10 and r2 = 5, the ith
correct measured level is:
r
Li = L'i + ε 10 log10
6.67
where
However, where possible, measurements should be made over an arc to remove distance variation, because
the distance correction in 4.5.5 is approximate because test surfaces do not reflect sound to form exact spherical
or cylindrical waves.
5 Prediction techniques
The polar responses may be generated by the use of computer models such as boundary element or finite element
methods, or by quicker methods such as Fresnel diffraction, provided these models have been verified against
measurement results conforming to this information document or another accurate calculation model.
6 Extraction of coefficients
2
n Li /10
( )
n
∑ 10 − ∑ 10
Li /10 2
dθ = i =1 i =1
( )
n
(n − 1)∑ 10 Li /10
2
i =1
This equation is valid only when each receiver position samples the same measurement area. This sampling is
automatically achieved for single-plane measurements on an arc with an even angular spacing between receivers.
A diffuser that scatters sound completely has a diffusion coefficient of 1. When the scattered level is concentrated
in one measurement location, the diffusion coefficient approaches zero.
NOTE The fact that receiver positions at ± 90° actually sample half the area of the other receivers can
be ignored, since a correction makes no significant difference to the diffusion coefficient.
2001-02-19 printing
AES-4id-2001
–17–
y (Reference normal)
Receiver
θ
ϕ x
For some measurements, the area sampled by a receiver will vary with the receiver position. In that case, the
following formulation should be used:
2
n Li /10
( )
n
∑ ∑
2
10 N i − N i 10 Li /10
dθ = i =1 n i =1
n
∑ i ∑ N i 10
N − 1 ( )
Li /10 2
i =1 i =1
where
Ni is proportional to the area sampled by receiver point i and should be calculated from 6.2
This will arise for hemispherical measurements using an even angular spacing between receiver positions in
azimuth (φ) and elevation (θ). See figure 12. An uneven sampling of the hemisphere with respect to area results,
with more receiver points closer to the reference normal. The equation adds additional dummy samples into the
calculation to make the sampling linear with respect to area.
4π 2
Ai = sin ( ∆θ / 4) θ = 0°
∆φ
Ai = 2sin(θ ) sin(∆θ / 2) θ ≠ 0°, θ ≠ 90°
Ai = sin(∆θ / 2) θ = 90°
where
∆φ and ∆θ are the angular spacings, each usually 5°, in azimuth and elevation between adjacent
receivers.
Ai
Ni =
Amin
2001-02-19 printing
AES-4id-2001
–18–
where
7 Presentation of results
The diffusion coefficient shall be presented in a table for each third-octave band. The diffusion coefficient should
be rounded to two decimal places. In a graphical representation, the points of the measurement should be connected
by straight lines, the abscissa giving the frequency on a logarithmic scale and the ordinate showing the results from
0 to 1 on a linear scale. The abscissa should be in equivalent full-scale values with a statement of the scale ratio 1:N.
The ratio of the ordinate distance from 0 to 1, to the abscissa distance of five octaves should be 2:3. In addition,
although not necessary to conform to this document, the diffusion coefficient or coefficients for the reference flat
surface should be included for additional information.
8 Test report
The test report shall state:
a) conformity to AES-4id;
b) name and address of the testing laboratory;
c) manufacturer's name and product identification;
d) name and address of the organization or person who ordered the test;
e) date of the test;
f) description of the test sample with sectional drawings;
g) scale of the test sample relative to the full manufactured product;
h) coordinates of the source and receiver positions;
i) whether the sample was measured in a single plane, multiple planes, or over a hemisphere;
j) for single or multiple plane measurements, a statement of the planes used;
k) details of the measurement or prediction technique used to obtain polar responses;
l) if applicable, a table or formulation showing the distance corrections used for each receiver (see 4.5.3);
m) a table and graph of the diffusion coefficients obtained versus third-octave-band center frequency.
2001-02-19 printing
AES-4id-2001
–19–
Annex A
(normative)
A.2 Environments
A.2.3 Placement
Microphones, loudspeakers, and the test surface may be placed on a fully reflective surface to enable the
measurements to be made in a hemianechoic chamber or a pseudo-hemianechoic environment. The surface should
have a sound pressure reflection coefficient of at least 0,99. Sources and receivers shall be sufficiently close to the
reflective surface so that all frequencies being measured do not suffer from destructive interference due to grazing
reflections. This condition can be achieved by ensuring that measurement elements are no further than 1/4
wavelength from the fully reflective surface. The fully reflective surface shall be sufficiently large so that edge
diffractions from this surface are at least 40 dB below the signal within the rectangular window.
2001-02-19 printing
AES-4id-2001
–20–
Annex B
(informative)
Informative references
COX, TJ., and LAM, YW. Evaluation of Methods for Predicting the Scattering from Simple Rigid Panels. Applied
Acoustics, 1993, vol. 40, p. 123-140.
COX, TJ., and LAM, YW. Prediction and Evaluation of the Scattering from Quadratic Residue Diffusers. J. Acoust.
Soc. Amer., 1994, vol. 95, no. 1, p. 297-305.
D’ANTONIO, P. The Directional Scattering Coefficient: Experimental Determination, J. Audio Eng. Soc., 1992,
vol. 40, no. 12, p. 997-1017.
D’ANTONIO, P., KONNERT, JH., and KOVITZ, P. The Disc Project: Experimental Measurement of the
Directional Scattering Properties of Architectural Acoustic Surfaces. Wallace Clement centennial symposium, New
York, US: Acoustical Society of America, 1994, p. 1PAAD2.
HARGREAVES, TJ., COX, TJ., LAM, YM., and D'ANTONIO, P. Surface Diffusion Coefficients for Room
Acoustics: Free Field Measures. J.Acoust.Soc.Am., 2000, vol. 108 (4), p. 1710-1720.
ISO 266:1997, Acoustics – Preferred frequencies. Geneva, CH: International Organization for Standardization.
IEC 61260 (1995-08) Electroacoustics – Octave-band and fractional-octave-band filters. Geneva, CH:
International Electrotechnical Commission.
2001-02-19 printing