Designation: E 1130 – 02e1

Standard Test Method for

Objective Measurement of Speech Privacy in Open Offices
Using Articulation Index1
This standard is issued under the fixed designation E 1130; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (e) indicates an editorial change since the last revision or reapproval.

e1 NOTE—Footnote 4 was added editorially in March 2002.

INTRODUCTION

This is one of a series of test methods for evaluating the acoustical characteristics of open office
environments and the performance of acoustical components. Other proposed test methods in this
series deal with the laboratory measurement of interzone attenuation of partial height space dividers
and ceiling systems.

1. Scope 1.6 This standard does not purport to address all of the
1.1 This test method describes a means of measuring speech safety concerns, if any, associated with its use. It is the
privacy objectively between locations in open offices. This test responsibility of the user of this standard to establish appro-
method relies upon acoustical measurements, published infor- priate safety and health practices and determine the applica-
mation on speech levels, and standard methods for assessing bility of regulatory limitations prior to use.
speech communication. This test method does not measure the 2. Referenced Documents
performance of individual open office components which affect
speech privacy; it measures the privacy which results from a 2.1 ASTM Standards:
particular configuration of components (1, 2).2 C 384 Test Method for Impedance and Absorption of
1.2 This test method is intended to be a field test for the Acoustical Materials by the Impedance Tube Method3
measurement of speech privacy in actual open offices. How- C 634 Terminology Relating to Environmental Acoustics3
ever, this test method could be used in an environment E 1041 Guide for Measurement of Masking Sound in Open
arranged to simulate an open office. Offices4
1.3 This test method could be adapted for use in other open E 1179 Specification for Sound Sources Used for Testing
plan spaces such as open plan schools. It could also be adapted Open Office Components and Systems3
for measuring the speech privacy between open plan and 2.2 ANSI Standards:5
enclosed spaces or between fully enclosed spaces. S1.4 Specification for Sound Level Meters
1.4 This test method relies upon the Articulation Index S1.6 Preferred Frequencies and Band Numbers for Acous-
which predicts the intelligibility of speech for a group of tical Measurements
talkers and listeners. While both the Articulation Index and this S1.11 Specification for Octave, Half-Octave and One-Third
test method can be expected to reliably predict average speech Octave Band Filter Sets
privacy, neither predicts the specific degree of speech privacy 3. Terminology
afforded to particular open office occupants.
1.5 The values stated in SI units are to be regarded as the 3.1 Definitions—The acoustical terminology used in this
standard. The inch-pound units in parentheses are for informa- test method is consistent with Terminology C 634. Of special
tion only. importance are the terms average sound pressure level and
arithmetic mean sound pressure level, both of which are
defined in Terminology C 634, and source point which is
1
This test method is under the jurisdiction of ASTM Committee E33 on defined in Specification E 1179.
Environmental Acoustics and is the direct responsibility of Subcommittee E33.02 on
Open Plan Spaces.
3
Current edition approved Jan. 10, 2002. Published February 2002. Originally Annual Book of ASTM Standards, Vol 04.06.
published as E 1130 – 86. Last previous edition E 1130 – 90 (1994)e1. 4
Discontinued. See 2001 Annual Book of ASTM Standards, Vol 04.06.
2 5
The boldface numbers in parentheses refer to the list of references at the end of Available from American National Standards Institute, 11 W. 42nd St., 13th
this test method. Floor, New York, NY 10036.

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.

1
 E 1130
4. Summary of Test Method 4.9 Calculate the signal-to-noise ratio in each one-third
4.1 Select two locations in an open office environment, octave band by subtracting the measured ambient and sound
between which the speech privacy is to be measured. Designate pressure levels from the calculated speech levels at the listener
one location as the talker or source location and the other as the location.
listener or measurement location. These locations should typify 4.10 Calculate the Articulation Index in accordance with
the usual working positions of office occupants. Section 11 using the one-third octave-band signal-to-noise
4.2 All masking sound systems and HVAC systems shall be ratios.
operating in their usual manner. 4.11 Report the Articulation Index to two decimal places as
4.3 At the listener location, measure the ambient sound the measure of speech privacy.
pressure levels in each one-third octave-band from 200 to 5000 5. Significance and Use
Hz and the A-weighted sound level.
5.1 The speech privacy between open offices is determined
4.4 Locate a qualified sound source at the talker location by the degree to which intruding speech sounds from adjacent
and orient it toward the listener location. Drive the source with offices exceed the ambient sound pressure levels at the listen-
pink or white noise at a level sufficient to increase the one-third er’s ear; a classic signal-to-noise ratio situation.
octave-band sound pressure levels at the measurement location 5.2 The sound pressure levels at the listener’s ear from
by at least 10 dB above the ambient over the entire frequency speech in adjacent offices depend upon:
range of interest. The sound pressure levels produced by the 5.2.1 The individual vocal effort and orientation of the
sound source at a 0.9-m (3-ft) reference position for a known talker,
electrical input will have been previously established (3). 5.2.2 The attenuation of speech signals due to distance or
4.5 Measure the sound pressure levels in one-third octave intervening barriers, and
bands at the listener location with the source on. 5.2.3 The reinforcement of speech signals due to reflections
4.6 Calculate the level reduction in each one-third octave from office surfaces such as the ceiling, furniture panels, light
band, that is, the difference in sound pressure levels produced fixtures, walls, or windows.
by the sound source at 0.9 m (3 ft) and at the listener location. 5.3 The ambient sound pressure levels will usually be
4.7 Determine the speech spectra to be used. The normal controlled to mask intruding speech. This is accomplished by
voice spectrum of male speech peaks from Table 1 must be means of a masking sound system. However, in certain
used; optionally, additional spectra may be used. positions and frequency ranges, heating, ventilating, or air
4.8 Calculate the one-third octave-band sound pressure conditioning equipment (HVAC) may contribute significantly
levels for the speech spectrum at the listener location. This is to ambient sound pressure levels. Guide E 1041 may be used to
carried out by subtracting the measured level reductions from measure masking sound.
the speech spectrum. 5.4 The primary purpose of this test method is to measure
the speech privacy for an average speech spectrum using the
standard Articulation Index method. This requires measure-
TABLE 1 Speech Peaks for MalesA,B ment of the relevant acoustical characteristics discussed in 5.2
One-Third One-Third and 5.3 for a pair of offices and calculation of the Articulation
Octave-Band Octave-Band
One-Third
Sound Pressure Sound Pressure
Index using an average speech spectrum. The average speech
Octave-Band spectrum is for male talkers speaking with normal voice effort.
Levels of Levels of
Center
Frequency,
Speech Peaks Speech Peaks 5.5 The Articulation Index ranges from 0.00, where speech
for Normal for Raised
Hz
Voice Effort Voice Effort
is unintelligible, to 1.00, where all individual spoken words can
(dB re: 20 µPa) (dB re: 20 µPa) be understood. Caution should be exercised in interpreting the
200 60 63 numerical results of this test method. There is a need for further
250 64 68 research to establish the relationship of Articulation Index to
315 63 67 speech privacy. One purpose of this test method is to encourage
400 65 70
500 66 72 the measurement of data and further research on this topic
630 64 70 perhaps leading to development of well-documented speech
800 58 66 privacy categories and criteria.
1000 58 65
1250 59 67 5.6 This test method can be used to:
1600 56 63 5.6.1 Compare the relative privacy afforded between differ-
2000 52 59 ent pairs of open offices.
2500 53 60
3150 53 58 5.6.2 Evaluate how changes in open office components
4000 50 56 (barriers, furniture, ceilings, masking sound, or wall panels)
5000 46 52 affect speech privacy.
A
Speech peaks calculated from rms values given in Ref (4) by adding 12 db, in 5.6.3 Measure speech privacy objectively for correlation
accordance with (7). Values given in Ref (4) are normalized to 1 m and are used
in this test method as representative values for 0.9 m. with subjective responses.
B
The A-weighted sound level and linear sound pressure level of the normal voice 5.7 This test method could be one element of a performance
effect spectrum given in Table 1 are 70 dBA and 73 dB, respectively. The or acceptance test procedure. However, many additional items
corresponding levels for the raised voice effort spectrum are 76 dBA and 78 dB.
These may be compared to the levels for the “idealized speech peak spectrum” would need to be specified to use this test method for
given in Table 8 of ANSI S3.5, 74 dBA and 77 dB.(7) performance testing of an open office environment, such as, the

2
 E 1130
number of office pairs to be tested and method of selecting 7.1.4 Calibrated Sound Source, of known sensitivity and
those offices, and the method of averaging the results. Speci- directivity, completely described in Specification E 1179 and
fying a numerical criterion in terms of the Articulation Index is calibrated in accordance with the Annex of this test method.
also necessary for acceptance testing; however, the selection of 7.2 On-Site Data Acquisition Instruments, that may consist
such a criterion and permissible deviations should be under- of a microphone with calibrator, a precision sound level meter
taken with care in view of the present state-of-the-art as capable of measuring average sound pressure level, a one-third
discussed in 5.5. octave-band analyzer, and a strip chart recorder (optional). Test
5.8 Recent data on speech levels as discussed in Ref (4) are data may be read on-site or tape recorded for later analysis.
normalized to 1 m (3.3 ft). However, recently developed test Recording instruments may consist of a microphone with
methods for evaluating open office components and systems, calibrator, precision sound level meter (optional), and an
including this test method, have been developed using dis- instrument quality tape recorder. Laboratory data reduction
tances standardized in U.S. customary units (feet). This test instruments may consist of a tape recorder, a one-third octave-
method assumes that speech levels at 0.9 m (3 ft) and 1 m are band analyzer and a computer.
equivalent. This is a reasonable and conservative assumption 7.3 The background noise of measuring instruments in each
considering that the standard deviation of normal voice speech one-third octave band must be at least 10 dB below measured
levels is several decibels or more as discussed in Ref (4). Using sound pressure levels.
a point source model of spherical spreading, the difference in
8. Test Signal
sound pressure levels between 0.9 and 1 m (3 and 3.3 ft) would
be at most 1 dB. 8.1 The test signal shall be contiguous, one-third octave
bands of pink or white noise, from 200 to 5000 Hz, and may be
6. Test Space extended.
6.1 The test space shall be an actual or a mock-up open 9. Speech Spectra
office environment. 9.1 The normal voice effort spectrum (2) of speech peaks
6.2 The ceiling of the test space shall be complete, including given in Table 1 shall be used for the calculations. Additionally,
ceiling board, light fixtures, and air diffusers. other spectra may be used such as the raised voice spectrum of
6.3 The floor covering and wall finishes shall be completely speech peaks in Table 1.
installed prior to testing.
10. Procedure
6.4 Office furnishings shall be in place.
6.5 Any masking sound system shall be set as intended for 10.1 Measurement Location:
use in the occupied space. 10.1.1 Select a listener (measurement) location at the typi-
cal ear-height of a seated listener, that is, the microphone shall
NOTE 1—If the masking system is adjusted to produce significantly be located 1.2 m (4 ft) above the floor.
higher sound pressure levels, greater speech privacy is obtained; however, 10.1.2 The measurement location should be at least 1 m (3.3
the higher levels may be found unacceptably annoying. ft) from any vertical surfaces such as walls, columns, desks, or
6.6 The open office space should be unoccupied during the office furniture. In the event that this cannot be met, the
tests. selected location shall be at the usual worker’s position.
10.2 Ambient Measurements:
NOTE 2—If the purpose of testing is to evaluate the change in speech 10.2.1 With the test signal off, the average sound pressure
privacy which results from component changes (for example, installing
level shall be measured in each one-third octave band over a
wall finishes or adjusting the masking system) the different conditions for
each test result shall be carefully documented. time period of at least 4 s at four positions at 90° intervals
around a circle of 0.3-m (1-ft) radius centered on the location.
7. Apparatus The arithmetic mean sound pressure level shall be calculated
for each band from the four measured values. The A-weighted
7.1 The minimum instrumentation required for this test
sound level shall also be measured and calculated in this
method is as follows:
manner. All data shall be obtained with the microphone
7.1.1 Microphone and Amplifier, that meet or exceed the diaphragm parallel to the floor and facing upwards.
requirements of ANSI S1.4 for Type 1 sound level meters. A 10.2.2 Verify that the measured ambient sound pressure
random incidence microphone shall be used, that is, one that levels exceed the background noise of the instruments by at
has its flattest frequency response for sounds arriving at least 10 dB in each band. If not, the data shall not be used.
random angles. 10.3 Source Location and Orientation:
7.1.2 One-Third Octave-Band Filter Set, meeting the re- 10.3.1 The sound source shall be placed at the talker
quirements of ANSI S1.11 for Order 3 or higher, Type 1 or location and the source point shall be located 1.2 m (4 ft) above
better. The nominal center frequencies of the filters shall the floor.
include all the preferred one-third octave bands from 200 to 10.3.2 The loudspeaker axis should be oriented so that its
5000 Hz conforming to ANSI S1.6. This range may be projection in the horizontal plane is along a straight line from
extended. the talker to the listener location.
7.1.3 Sound Level Meter, Graphic Level Recorder, or other 10.3.3 In the vertical plane, the loudspeaker axis should be
device from which the sound pressure level can be read. oriented to be 25° above horizontal.

3
 E 1130
10.4 Level Reduction Measurements: 12. Report
10.4.1 Turn the test signal on and measure and adjust the 12.1 Report the following information:
electrical signal to be the same as used for the measurement of
12.1.1 Statement, if true in every respect, that tests were
the reference levels.
conducted in accordance with the provisions outlined in this
10.4.2 Repeat the measurements described in 10.2.1.
test method. Any options or exceptions to this test method shall
10.4.3 Verify that the measured sound pressure levels ex-
be noted.
ceed the ambient values measured in 10.2 by at least 10 dB in
each band. If not, the data shall not be used. 12.1.2 Description of the test space including the ceiling
10.4.4 Calculate the level reduction due to the environment material, the type of ceiling suspension system, light fixtures,
in each one-third octave band, by subtracting the values ceiling height, the partial height space dividers, and the floor
measured at the measurement location from the source refer- covering. Optionally, descriptions may be useful for ceiling
ence levels (see Annex A1). diffusers, construction details above and below the ceiling, the
plenum depth, wall finishes, and interior furnishings.
11. Calculation 12.1.3 Description of the sound source location and orien-
11.1 Select a sound pressure level spectrum of speech peaks tation and the measurement location including distances from
from Table 1 or elsewhere. One set of calculations shall use the office boundaries and furniture. Optionally, sketches may be
normal voice spectrum from Table 1, as required by 9.1. used to illustrate positions and distances.
11.2 Calculate the sound pressure levels of peak speech 12.1.4 Description of any masking system including the
signals in each one-third octave band at the listener location by means of generating and distributing sound.
subtracting the level reductions from the spectrum selected in 12.1.5 Description of the air handling system or other
11.1. ambient noise sources that contribute to masking sound.
11.3 Calculate and round to the nearest decibel, the signalto- 12.1.6 Complete description of the test signal and the sound
noise ratio in each one-third octave band: the amount by which source.
the peak speech levels calculated in 11.2 exceed the ambient 12.1.7 Description of the instruments used to acquire acous-
values measured in 10.2. Whenever the signal-to-noise ratio is tical data including manufacturer, type, and model.
zero or less, use a value of zero; whenever the signal-to-noise 12.1.8 Statement identifying data that were not acquired at
ratio is 30 dB or more, use a value of 30. least 1 m (3.3 ft) from the nearest vertical surface.
11.4 Calculate the Articulation Index by multiplying the 12.1.9 Statement that “The calculated Articulation Indices
band weighting factors from Table 2 by the signal-to-noise given in this report are a function of the specific environment
ratios and summing as follows: and should not be construed as applying to other environ-
15 ments.”
AI 5 ( Wi · R i
i51
(1)
12.1.10 Listing, and optionally a graph, of the following
one-third octave-band sound pressure level spectra rounded to
the nearest decibel:
where:
AI = Articulation Index, 12.1.10.1 Reference levels for the sound source,
Wi = weighting factor (from Table 2) for band i, and 12.1.10.2 Ambient values measured in accordance with
Ri = signal-to-noise ratio for band i. 10.2.1, including the A-weighted sound level,
NOTE 3—Fig. 1 presents a data sheet that may be helpful in performing
12.1.10.3 Values measured with the sound source on in
these calculations. accordance with 10.4.1,
12.1.10.4 Level reductions calculated in accordance with
10.4.3,
TABLE 2 Articulation Index Weighting Factors for Preferred One-
Third Octave Bands in Accordance with ANSI S3.5 (7) 12.1.10.5 Spectra of speech peaks used (one of which must
One-Third
be the normal voice spectrum),
Octave-Band
Weighting Factor 12.1.10.6 Levels of speech peaks calculated in accordance
Center Frequency,
Hz
with 11.2, and
12.1.10.7 Signal-to-noise ratios calculated in accordance
200 0.0004
250 0.0010 with 11.3.
315 0.0010 12.1.11 Calculated Articulation Index rounded to two deci-
400 0.0014
500 0.0014
mal places.
630 0.0020 12.2 If testing is undertaken to compare the resulting speech
800 0.0020
1000 0.0024
privacy under different conditions such as with different
1250 0.0030 masking spectra, the test report should indicate the Articulation
1600 0.0037 Indices for each condition.
2000 0.0038
2500 0.0034
3150 0.0034 13. Precision and Bias
4000 0.0024
5000 0.0020 13.1 The precision and bias of this test method have not
been established.

4
 E 1130

NOTE 1—Lp denotes sound pressure level, dB re 20 µPa. Example: (B-C) denotes arithmetic subtraction of values in Column C from Column B.
Column H: If less than 0 use 0, if more than 30 use 30.
FIG. 1 Sample Data Sheet for Speech Privacy Measurements

14. Keywords
14.1 architectural acoustics; articulation index; field test;
open office; open office system; open-plan space; speech
privacy; system test

ANNEX

(Mandatory Information)

A1. CALIBRATION OF THE SOUND SOURCE

A1.1 Measurement of Reference Levels A1.1.2 Measure and document the electrical signal fed to
A1.1.1 The reference sound pressure level in each one-third the source. This can be accomplished by measuring the voltage
octave band is defned as the arithmetic average of the levels fed to the loudspeaker using a precision sound level meter set
measured on the loudspeaker axis at 0.6, 0.9, and 1.2 m from to A-weighting. A precision sound level meter can be used as
the source point. Obtain these levels in an anechoic room with a wide-band voltmeter by removing the microphone and
surfaces that have a minimum normal incidence sound absorp- driving it directly with an electrical signal.
tion coefficient of 0.990 at all frequencies above 175 Hz as
measured in accordance with Test Method C 384.

5
 E 1130
APPENDIXES

(Nonmandatory Information)

X1. RELATIONSHIP OF ARTICULATION INDEX TO SPEECH PRIVACY

X1.1 The Articulation Index ranges from 0.00 to 1.00. X1.1.2 At Articulation Index values between 0.05 and 0.20,
Values increase as speech privacy becomes progressively normal speech privacy is indicated. In this range concentrated
worse. There is a need for further research to establish the effort is required to understand intruding speech (6).
relationship of Articulation Index to speech privacy. The X1.1.3 Speech becomes more readily understood at Articu-
following general guidelines are provided to give the user of lation Index values greater than 0.20. Some describe unaccept-
the test method some means to interpret the numerical results: able privacy as values above 0.30. At Articulation Index values
X1.1.1 Speech privacy may be described as confidential above 0.40, there is essentially no privacy.
when speech cannot be understood. This degree of speech
privacy is indicated at Articulation Index values at or below
0.05 (5, 6).

X2. ALTERNATE SINGLE NUMBER RATING

X2.1 Another useful scale for speech privacy is called X2.2 Consequently, confidential speech privacy is achieved
Privacy Index, (PI). It is expressed in percent and can be at PI values of 95% or more. Normal speech privacy corre-
calculated from the Articulation Index, (AI) as follows: sponds to PI values between 80 and 95%.
PI5~1–AI! 3 100% (X2.1)

REFERENCES

(1) Warnock, A. C. C., “Studies of Acoustical Parameters in Open-Plan Protection Agency, May 1977, p. 39. Available from National Techni-
Offices,” Journal of the Acoustical Society of America, Vol 63, No. 3, cal Information Service as No. PB-270 053.
March 1978, pp. 832–840. (5) Cavanaugh, W. J., Farrell, W. R., Hirtle, P. W., and Watters, B. G.,
(2) Warnock, A. C. C., “Acoustical Privacy in the Landscaped Office,” “Speech Privacy in Buildings,” Journal of the Acoustical Society of
Journal of the Acoustical Society of America, Vol 53, No. 6, June 1973, America, Vol 34, No. 4, April 1962, pp. 475–492.
pp. 1535–1543.
(3) Herbert, R. K., “Use of the Articulation Index to Evaluate Privacy in (6) Pirn, R., “Acoustical Variables in Open Planning,” Journal of the
the Open Office,” Noise Control Engineering, Vol 11, No. 2, Acoustical Society of America, Vol 49, No. 5, May 1971, pp.
September-October 1978, pp. 64–67. 1339–1345.
(4) Pearsons, K. S., Bennett, R. L., and Fidel, S., “Speech Levels in (7) ANSI S3.5–1969, “Method for the Calculation of Articulation Index”,
Various Noise Environments,” EPA-6001-77-025, U. S. Environmental American National Standards Insitute, New York, NY 10036.

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or [email protected] (e-mail); or through the ASTM website
(www.astm.org).