FAA DS-67-1

0I'• •'• NOISINESS JUDGWENTS OF

V HELICOPTER FLYOVERSTECHNICAL. REPORT

January 1967

by

Karl S. Peaons D D CBolt Beranek and Newman Inc.

15808 Wyandotte Street MAR 1 5 1967Van Nuys, California 91406

Under Contract FA65WA-1260 .

FEDERAL AVDATION AGENCY•. AIRCRAFT DEVELOPMENT SERVICE

TECHNICAL REPORTDS-67-1

Contract No. FA65WA-1260

NOISINESS JUDGMENTS OF HELICOPTER FLYOVERS

By

Karl S. Pearsons

Prepared for

THE FEDERAL AVIATION ADMINISTRATIONUnder Contract No. FA65WA-1260

By

BOLT BERANEK AND NEWMAD, INC.15808 Wyandotte Street

Van Nuys, California 91406

This report has been approved for general availability.The contents of this report reflect the views of thecontractor, who is responsible for the facts and theaccuracy of the data presented herein, and do not neces-sarily reflect the official views or policy of the FAA.This report does not constitute a standard, specificationor regulation.

ABSTRACT

Judgment tests were conducted in which 21 college studentsJudged the noisiness or unwantedness of eight recorded helicop-ter flyover noises vs a Jet transport flyover noise and ashaped band of noise. Tests were conducted in an anechoicchamber using mainly the method of paired comparisons. TheseJudgment tests indicate that the calculated perceived noiseleve' is the best predictor of noisiness, followed closely tythe N-weighted sound pressure level and the A-weighted soundpressure level, and finally, the overall round pressure level.Duration and pure-tone corrections applied to the calculatedperceived noice level did not improve the prediction accuracyof this measure.

iii

TABLE OF CONTENTS

Page No

I. INTRODUCTION . . . . . . . . . . . . . . . .. . I1

TI. TEST DESCRIPTION ................. 2

A. Subjects . . . . . . . . . . . . . . . . . . 2

B. Stimuli . . . . . . . . . . . . . . . . . . . 2

C. Equipment. . . . . . . . . .. . . . . . ... 7

D. Procedure. . . . . . . . . . . . . . .. . . 7

III. JUDGMENT TEST RESULTS. . . . . . . . . . . . . . 12

IV. CONCLUSIONS ................... 26

REFERENCES

APPENDIX A

iv

LIST OF TABLES

[1Table No., Paget1

I LEVELS OF HELICOPTER NOISE USED IN PAIRED-SCOMPARISON J UDGMEN TESTS .T.*. 6 0 .3

II HELICOPTER PAIRED-COMPARISON JUDGMENT TESTS . 15

III HELICOPTER PAIRED-COMPARISON JUDDGMENT TESTS . 16

IV AVERAGE DIFFERENCES BET-TEEN STANDARDS . . . . 20

i-V

LIST OF FIGURES

Figure No. Page No.

1 Standard Sound Stimuli for HelicopterJudgment Tests . . . . . .. . . , 0...*. , 4

2 Comparison Sound Stimuli for HelicopterJudgment Tests ............... 5

3 Comparison Sound Stimuli for HelicopterJudgment Tests............... 6

4 Time Histories of Two Helicopter NoiseSamples. . . . . . . . . . . . . . . . * 8

5 Block Diagram of Experimental Setup. . . . . 9

6 Plan View of Anechoic Chamber Showing'Speaker and Subject Locations. . . , * . . . 11

7 Sample of Results of Helicopter JudgmentT e s t s * .e o . e e o 9 * * * 9 a e 9 e * 9 e 1 3

8 Sample of Results of Helicopter JudgmentTests Plotted on Normal Probability Paper. . 14

9 Equal Noisiness Judgments of HelicopterFlyovere in Terms of ASPL A-Level,N-Level, PUL Using a Dc-8:30 Jet Flyoveras a Standard. . . .*. . . . . . . . # . . . 17

10 Equal Noisiness Judgments of HelicopterFlyovers In Tem of OAPL, A-Level,N-Level, PXlL Using Simulated Jet Flyoveras a Standard, *, e * * e e * @ e * * * e • 18

31 Time Histories of Two Sound SamplesDetermined Using N-Network . .0. . . . . . . 22

12 Equal Noisiness Judgments of HelicopterFlyovers in Terms of PHL Using a DC-8:30Flyover as a Standard. . & a , . a , a , , . 24

13 Equal Noisiness Judgments of HelicopterFlyovers in Terms of PYL Using a SimulatedJet Flyover as a Standard, . , . , . , & . 0 25

vi

I. INTRODUCTION

The recent increase in the use of helicopter, has broughtwith it additional noise exposure to many people. The heli-ports of the city have further complicated the problem bybringing the noise closer to the city dwellers. To betterunderstand people's assessment of helicopter noise, BoltBeranek and Newman has conducted Judgment tests to determinethe noisiness of typical helicopter flyovers under FAA ContractNo. FA65WA-1260. During the tests, subjects were asked tojudge the noisiness or unwantedness of helicopter flyover no)isesignals vs Jet transport flyover signals and a shaped band ofnoise.

Section II of this report deecrtbes the apparatus, stimuli,and general procedure employed during these tests. Section IIIpresents a summary and discussion of the test results. Theconclusions are presented in Section IV.

II. TEST DESCRIPTION

A. SubjectsTwenty-one college students were employed as subjects for

this test. The group included 13 males and 8 females rangingin age from 17 to 23 years with a median ige of 20. Allsubjects were audiometrically screened prior to the test. Thescreening level was held to within C0 dB )f the new ISO standardthreshold (Pef. 1).

B. StimuliEight sound stimuli representative of five eifferent types

of helicopters were chosen, as shown in Table I. The largehelicopters are represented by the Vertol CH-46 and SikorskyS-61 types while the smaller helicopters are represented bythe Sikorsky CH-34, the Kaman HH-43D and the Hughes 269A.The Sikorsky CH-34 and the Hughes 269A are powered by pistonengines; the remainder are powered by turbine engines.Included in the table are overall sound pressure levels (OASPL)determined from method of adjustment tests described later.Also presented are the equivalent measures of the calculatedperceived noise level (PNL), the N-weighted (N-level) andA-weighted sound pressure levels (A-level).

Table I also lists the time durations of the sound samplesas measured with an N-weighting network.* Time duration isdefined in Table I as the amount of time the signal is within10 dB or 20 dB of the maximum N-level.

Spectra of the sound stimuli are given in Figs. 1-3. Thesespectra wext determined by taking the maximum level of thesound samples in each one-third octave band.

Figure 1 shows the two stimuli employed as standards duringthis test; a DC-8:30 flyover and the simulated jet flyovernoise. The time pattern of the simulated jet flyover noise issimilar to that of an actual flyover. No attempt was made tosimulate Doppler shifts and the spectrum shape shown in Fig. 1remained constant throughout the sample duration.

The spread of the measurement data is also shown in Fig. 1.This spread is typical of all of the sound samples employedduring the test.

* The N-weighting network has a response equal to the inverse40-noy curve (Ref. 2).

--2-

0Cuj r- LC\ cN wO wO ' m' m 0 0

t: Cu lý Cu ý Cu Hr:_

0H4 %.00 0 t-- Q~- ' 0 co 00 N-0E

r 4 '.0 r4 r- ~ u 4 0-4

I- N - U"\ H4 CO) a' O 0

E1 P4. HT Hý H a' Lfl N, (VI) CulMJ m' a' m'C a'* m' a'

0

0aH a a% m m. aT F m

ONGi m2Y 0 0 0 0 >'a u ~C

rn~ 4)l 0 '.0m N -m Nn4 In a' a'm a, ' 0 a' a, a'

0t CO C ON 0 HOm a a\ m' m' 0 0 0 0

V H H H 0

4 4 f 4 H CO D '.0 - r- H,

Hý in N- Nl- N- - w- m- m- mZ. CO - C . . . . -

H ~ k Cu C4 I ''0 L~Lco CO co co CO N-SC O CO0

0 0 '0 uirrI -Hc

-D ___ ___ __ ___t-___ __ ___ _

oo o c o ,o o o

r4 00 00 CS M N ' eR H M HCC) CO CO CO CO CO CO C

HI 44j Mt O r4 N-ý uY 0 vCO CO CO %C 0' CO, CO COa

0~ COO

0o 4) a) a0

H~c Hu --t (' .0 H '. A+) -Hr (

-9 0 .

4.2 I'D 540a I C) 0

Col II u o0) 0 444)msVr 00 r4 .%.. '0 '0 w pt &q0 0

0P- 4-) a' a' F40i '-r- z~ 9. A'.0- H :

ca M ca co ca~~

-p-3-

I 00

C

> _4

I -- oT 2

0 z

SZ >

-4--

00-01- do* 0

------------------------------------------------

- - - 7 , Of

.9L

9-b-

.16

TW VL

R C9 8 R

Zw/*_ - oo- -( ---. Ilqopu O9 jt&dpnSpn

-5-IU

LIa

- a I- U

000

00000

70

0

zz

100

-6-0

The levels of the stimuli shown in Figs. 2 and 3 areaverage lsvels of the comparison sounds used in the method ofpaired comparison described later in the report. The stimuliwere actually presented at levels of +6, +2, -2, and -6 dB,relative to the sound pressure levels shown in the figures.

Two typical time histories of the helicopter flyovernoise samples are shown in Fig. 4. The strong modulationshown for the Vertol CH-46 helicopter noise in tis figure iscaused by the prominent blade slap present in the recordedsample.

C. EquipmentA Zenith 1lO-T audiometer was used for the audiometric

screening of the subjects. A block diagram of the equipmentused in the presentation of the test stimuli to the subjectsis shown in Fig. 5. The equipment consisted of an AmpexAG-350 tape recorder, a Daven attenuator, and an Altec Lansing165-watt power amplifier. A power attenuator, placed betweenthe amplifier and loudspeaker, was constructed to reduce theoutput of the amplifier in iO-dB steps to provide an increasedsignal-to-noise ratio.

The loudspeaker system consisted of an Altec Lansing 515-Blow-frequency uni.. coupled with an Altec Lansing 288-C driverwith a 311-30 horn. The speaker for the voice channel was asmall utility 6-in. loudspeaker driven by a Heathkit 15-wattamplifier. An electronic switch with a 100-millisecond rise-decay time was employed to reduce any audible transients inthe re-recording process during the preparing of the mastertest tape.

Measurementd of the bound stimuli presented to the subjects-were made at the subject's ear position (without the subjectpresent) for each seat location. These measurements were madewith a Bruel and Kjaer 4133 1/2-in. microphone, a Bruel andKjaer 2630 battery-operated cathode follower, a Bruel andKjaer 2203 sound level meter, and a Kudelski Nagra III-B taperecorder. The original field recordings of the helicopter andjet transport flyover noise were made using the same measurinainstrumentation. The one-third octave band analyses of thesound samples were made with a Bruel and Kjaer 2112 spectro-meter and Bruel and Kjaer 2105 graphic level recorder.

D. Procedure

The judgment tests were conducted in an anechoic chamber8' x 10' x 7.5' high, located at the Bolt Beranek and Newmanfacilities in Van Nuys, California. Two methods were employed

"-7-

LU

CL

zLU

CL.

-~ 0

-=--J

c c uo4. .0It atr

p& I r1uuiL2 2 c0

00

LU

LUJ

UO!S!A!P iad 9p 01 UO!S!A! P je W 0 1

JOASI *JflhtJd PUflOS JIIWAO

-e -

Z5

-- -- - ---- - -

0 Z

LUU

'i !

ek !: n e

za-'0 -

0 0UU

LU >

ceU

-9-

during the series of judgment tests: the method of adjustmentand the method of paired comparison.

The method of adjustment was used in preliminary tests toobtain levels for the more detailed paired-comparison tests.In the method of adjustment, subjects were asked to adjustthe level of a comparison sound until they judged that it wasjust as noisy or disturbing as the standard sound. Theactual instructions for this test are given in Appendix A.

For the ptired-comparison miethod, a tape was prepared forpresenting 4e sound samples to the subjects. In preparingthe tape, the sound samples were paired using both A-B andB-A orders to eliminate any time-error bias oicasioned by theorder in which the signals were presented. The test pairswere then randomized using a random number table and recordedon magnetic tape. During presentation of the paired-comparisontape, the subjects were asked to choose which of the two soundstimuli was the noisier and to respond by punching the appropri-ate positions on an IBM port-a-punch card. The actual instruc-tions and a typical answer card are given in Appendix A.

A plan view of the seating arrangement is shown in Fig. 6.With this arrangement three to six subjects were tested atone time until the entire group of 21 subjects completed four45-min. test sessions. To avoid the possible effect of fatigue,the 45-mtn. test sessions were conducted on separate days. Atotal of 132 sample pairs were tested.

-10-

'V-0

UL

LUUtAA

<I

-01

III. JUDGMENT TEST RESULTS

The subjects' responses, recorded on IBM cards, were enteredinto a digital computer for sorting and analysis. A computer-generated display of typical results is shown in Fig. 7. Thestendard in this case was a DC-8 flyover; the comparison, aSikorsky S-61 flyover. The dashed line represents the resultsobtained when the standard stimulus was presented first. Thesolid line represents the results obtained when the comparisonstimulus was presented first.

We considered that the two sounds were equally noisy oracceptable when 50% of the subjects stated that one sound wasnoisier than the other. Since this level of equal noisinessshould be independent of the order of presentation of thestimuli, we averaged the two values at the 50% point obtainedfrom the two different orders of presentation. For the datashown in Fig. 7, this averaged 50% level is 86.4 dB (i.e., anaverage of 83.5 dB and 89.3 dB).

To provide an idea of the spread of the judgments, theresults shown in Fig. 7 are plotted on probability paper inFig. 8. The results of this plot indicate that for the casewhen the standard is presented first, the starlard deviation is3.4 dB, and when the comparison is presented first, the standaredeviation is 4.9 dB. The standard deviations for all stimuluspairs ranged from 2.5 to 9 dB with an average of 5.4 dB. Thiscompares closely with the standard deviation of 5 dB obtainedin a related British study (Ref. 3).

It is interesting to note that the values for equal noisinessat the 50% points determined using probability paper (Fig. 8),are quite comparable to those obtained with linear paper (Fig. 7In comparing all of the results obtained using the computer'sgraphical output with those obtained using probability paper,the average difference is only 0.5 dB with a standard deviationof 1 dB.

Tabulations of the mean levels of the helicopter stimulijudged equally noisy to the two standards employed in the testare given in Tables II and III. The two tables also show themean differences in levels between the helicopter signals andthe standards in terms of the overall sound pressure levels,A- and N-weighted levels, calculated perceived noise levels,and the calculated perceived noise levels adjusted for signalduration and discrete frequency content. Signal durationcorrections were determined for both l0-dB- and 20-dB-downduration times. The differences are also plotted in Figs, 9and 10.

-12-

10051 Standard - DC-8 Flyover1 (8S i OASPL)I/I Comparison - Sikorsky

S-61 FlyoverI/

0 /ii /U

0/ Standard

Precented First

•u ComparisonS~Presented First

o% .... .....

81 85 89 93

Overall Sound Pressure Level in dB re 0.0002 dyne/era2

FIGURE 7. AMPLE OF RESULTS OF HELICOPTER JUDGMEI

-13-

100 _ _

"-a - Standard- DC-Q Flyover /S98 -- ( 88 dB OASPL)c9-

Comparison - Sikorsky4)• S-61 Flyover•_=95

r-o

E90

._a _-7 :"70

00

e0 80

570

"I- -

0

• 30

o 20___ _

U 0

•)- - Standard0 5 Presented First0

Comparisonc 2 Presented First

081 85 89 93

Overal ISound Pressure Level In dB re 0.0002 dyne/cm2

FIGURE 8. SAMPLE OF RESULTS OF HELICOPTER JUDGMENTTESTS PLOTTED ON NORMAL PROBABILITY PAPER

-14-

p * m cn-t Mko aOr-

29 1;0'; C; r- 11 0 CUC'

4t-

E--q ~~ 3~ 0 i 4)M . .. . 1. t

P a n -4C~ O0000r.4CM 0 r-I H

tol2

00

E4i C 0 E) E!~.~HC' Ma 0. H0Y'. t*- 4-)43 4=tN N I'- 0I II oV-

0 toj --.too t-0ulOO r rS ) 4) Np M MCV k 0 vt '.---a-0o P Vfr am) 0 Ca(N\ C*A M -4SM P. --) S.~~r ' 4) >3 02 02V C U\ 46) *4 r43040

0)kr 0Chjf ý CU0HCAt 44H 1W a I0 C) u0 ,0- -to zz I0 1)Cjc 4)4a)H ~ ~ ~ .H Ha 0 o. j - __ _ _ _ _ --

b. cs d '04 'a> IQ~\O '4-1 h ) r_ )

ca V+ >0 coVA MVU. M *Y Ms H 00e

t4~ 4 p4 1-4 co0 H4\3u+2t rfI %- t'-p4O' @1w O) s--i0 0 0

toI *f4 6) .* . to 0

(:4)'o gototoogM-t 4) C~

0000 4)VH

:4-e~~m I W 4t0 IV d

-P O Lf~0 IV0~ 0U d) rA

- - 0 02 k"q ýz C Cd 0-000

t-c1-1-

L0434

4) -15-' 3. s-

** t414 k ~ CO 1-CJ~ -. 1- t

$4 li II I1 -4 r41 I

W ~'0 ' ; C r ý

ft-4 r4;& 0

> ý 0 ~r-i Cýr4 '0C~j C~j 0 O

Zr4

r= .. CC)Ca" 0. WV 00CH~ (\C1 C

sC oA mO@3 rnM- -Cjn' o\0 +ECI 'PQ 01 V -

'o Vý Cý Cý iL (l;\ 0ý Ca Ca

0 >z U''\Oa 0'- 11111 I ~ L'\ @3C~ tH H\ Vo P0Q1 .- . 4..) 'H 01 CYi 00 n00 - Cj10(j - Z - C: ca0 $4 C~ U (

H F-1i V H 0 -. 0 +3I 4-1F-* .3 04000 t

E--4ri 0 0~ m "-44.. 4). 00 ~ cCa

4.) +3 *\ M0L-r D C

PO 0r -0 010 C 1 f- >r ~V-4 4) 0 4-)4)

m0 o )

04I - 00 M m4)> 0a 430 C)vC

-Hr4 Hco2 O

N CC ~jO M0ý~V "0O MM 4 ý- 1.00 M o

U ~r4 1V

M riH -. 0 0 U ()4.)

.V C~ I -'.0Ir n -0 n H4 Oj0

WCC . -

r- @4 ca Ca

ý4ýzr40 =$ :3

E-l Mko I I Vý P V t

o 0% to ~ raI2 H 0 C-4 £.. P-4 :j+z

0 toco W bMH -

4 -1 C . 0 4- -

(Z #) 9V- H:) 4JOA

LU.>cv

00-- W- (tg)9pIF-H:o104JA -

U- 0

LuJI--

0 -

V C- H) AlsJoli S u) -JILA

LUJL

U.'z

z >

ZI

U) _Z I<0

0<ZZ

UJ

00IL

(IPN1) IP u! IOAa1 Fipuo.4S sfluiW ucsijdwo:)

-17-

z- (Zo) 9V-HD) 104JOA

0l.-

-i-

(Z#) 9rC-HD) ISOIII

LU- Zc

(1#) $"C-HD) AlsjoliS .UU

-j

(L#) AIJI'CH aIJO1 (fl c4A -J W

oZn

____ ____ ___ - 9L~t-HH UiD~u)I

.In

_ij-

00 U.

+A uV 1e 2*Bn

(Wtj) p U JODIPJDPUD4S Sn'!W UOS!J~dUIOD

-18-

If the measures of the standard and comparison noise stimuliwere identical (differences were zero) when the two samples werejudged to be equally noisy, then these measures would be perfectpredictors of the judgment results for the pair of sounds.

Further, if the objective measures of noise were perfectestimates of subjective judgments, and if experimental errorsand subject variability were nil, both the mean values andthe standard deviations of the differences would approachzero. Assuming the variability introduced by experimentalerrors and subject differences is independent of the soundmeasure, the measure showing first the smallest standarddeviation and then the smallest mean difference would be thebest predictor. Obviously, a measure with a small meandifference between comparison and standard levels accompaniedby a large standard deviation is not an accurate predictor.

"From the standard deviations and means in Tables YI and III,it appears that all measures tested predict the noisiness ofhelicopters fairly well with PNL being the most accurate,followed closely by N-level and A-level and finally OASPL. Thestandard deviations for the PNL and A- and N-level differencesare comparable and are less than those for the OASPL differencesshown in Tables II and III. Further, the mean difference ofPNL is less than that for the A-level difference. This differ-ence is only significant, however, at the 90% level of confid-ence using the students' t-test (Ref. 4) for the simulated jetflyover standard. In comparing the results tabulated in TablesII and III, we note that the mean differences and standarddeviations of Table II are always less than those for Table III.This is probably due to the more complex nature of the DC-8:30Jet flyover (standard for Table III) compared to the simulatedjet flyover (standard for Table II).

.,n order to check the possible differences in judged noisi-ness between the two standards, we first note the level of bothstandards relative to that of a comparison sample; then, wetake the difference in these relative levels. The mean andstandard deviation of the eight differences (one for eachcomparison) for the four objective measures are shown in TableIV. As expected, the standard deviations are all the same1.5 dB) since they represent the same variation aroundifferent means. The mean differences are significantly small-

er at the 90% level of confidence for A- and N-level andcalculated PNL than for OASPL. In other words, if judgmentshad been made comparing the noisiness of the two standards, PNLand A- and N..level would have been more accurate In predictingthe noisiness than OASPL.

-19-

TABLE IV

AVERAGE DIFFERENCFS BETWEEN STANDARDS*

Level (Simulated Jet Flyover)Minus Level (DC-8:30)

OASPL A-Level N-Level PNL

(dB) (dB) (dB) (PNdB)

Standard Deviation 1.5 1.5 1.5 1.5

Mean Difference 2.5 1.3 1.6 1.5

* The differences were determined by taking differencesbetween each standard Judged equally noisy to one com-parison sound.

-20-

These conclusions are in agreement with those of someBritish investigations. One study (Ref. 5) found PNdBpreferable to OASPL as the measure of the noisiness ofsimulated rotor noise of helicopters. Another study inEngland (Ref. 3) using 4-sec. samples of recorded helicopternoise, found that the various objective measures ranked intheir accuracy for predicting the noisiness of helicopterflyovers as follows: loudness level (Zwicker), PNL, A-level,loudness level (Stevens), and OASPL, with OASPL being theleast accurate predictor.

Recent studies (Refs. 6 and 7) suggest additional adjust-ments for duration and pure-tone content to increase theaccuracy of the calculated perceived noise level. The resultsof applying these adjustments are indicated in the last twocolumns of Tables II and III. The time duration adjustmentswere made relative to the duration of the standard.

The pure-tcne adjustment changes the results very little,since the pure-tone content of most of the samples was quitesmall. However, the duration adjustment affects the resultsquite markedly. As shown in the next to the last column ofTables II and III, this duration adjustment for durationsmeasured 10 dB down from the maximum level increases thecalculated difference between the standard and comparisonperceived noise level when they are judged to be equally noisy.The duration adjustment also increases the standard deviationof the difference to values comparable to those for the OASPL.A possible explanation for this increased error may involvethe shape of the time history for the flyover samples undertest. If we leok at the time history for one of the helicopterflyovers shown in Fig. 11, we note that the helicopter timehistory is quite different in shape from the jet aircraft fly-over noise time history also shown in the figure. The jetaircraft flyover noise level increases at a near uniform rateto a maximum level and decays at about the same rate. Thenoise level of the helicopter flyover time pattern on theother hand, appears to increase at two different rates. Itincreases at one rate then abruptly changes to another untilit reaches the maximum level. During the decay portion, theprocess is reversed.

The choice of the lO-dB-down points as a measure of durationwas not critical during early tests of the development of theduration adjustment (Ref. 7), since the stimuli possessedregular time patterns and could be represented by durationsmeasured at 10 or 20 dB down from the maximum level. However,since the iO-dB-down duration for the helicopter flyover

-21-

u zu

'.0

0

2 0

0 I0

2-

- 22-

samples may not be representative of the time history, the20-dB-down points were tried in an effort to determine abetter measure of the subjected assessment of the helicopterflyover noise samples.

Since the duration adjustment is given only as a functionof the 10-dB-down duration, the measured duration was convertedto a 10-dB-down duration by assuming a triangular shaped timehistory. With this assumption, the 20-dB-down duration wasconverted to a 10-dB-down duration by dividing it by two.

The results shown in the final column of Tables II and IIIand Figs. 12 and 13 indicate that this 20-dB-down durationadjustment provides some improvement over the lO-dB-downadjustment. However, based upon size of the standard deviation,the adjusted PNL is still not as accurate as the PNL withoutduration and pure-tone adjustment.

This suggests that more basic work might be carried outinvolving additivity of spectral content and duration factorsin calculating perceived noise level. This research wouldinclude development of better measures of time duration insofaras they relate to noisiness assessments.

-23-

(Zj) 9$,-H:) 104JOA 'LU

> in

I.-

LL. >

00 00

- i~~9-S AliJO)I!S L

V~Z Z

a 0

nz _ _ _

7*00

()V 69E 04BnH

0 0 0

+ +

G'Nd u! I*AS- pJDpUDIS IflU!W UOSIJ~D&'JOD)

-24-

II

* *--.(Z#) 9$"-H:) (OJAlo! i

CL.i

UJ

.2 .0 D

(~g) C-HD ~sjo1 15 In__________ !l (A

4A, L-

72 cja'aZ

000

(9j) V 69Z SOON 0

A. *L I.

o U-

+ +

SPNd u! J*Aal pJopuo15 sfluiW uosijJdWuoD

- 25-

IV. CONCLUSIONS

As a result of investigations carried out under thiscontract, the following conclusions were reached:

1) As a predictor of the noisiness of helicopter flyovers,the calculated pprceived noise level provides the mostaccurate measure of the four objective measures includedin this investigation. The N-level and A-level,although slightly less accurate, were also reasonablepredictors, followed finally by the overall SPL.

2) Duration and pure-tone corrections did not improve thepredictability of the noisiness of the helicopter fly-over noise samples under test, possibly due to inadequateduration measures or a factor in the additivity of theduration adjustment not previously tested.

-26-

REFERENCFS

1. International Standards Organization, "Standard ReferenceZero for the Calibration of Pure-Tone Audiometers," ISORecommendation R- 389.

2, D. E. Bishop, Federal Aviation Agency SRDS Feport No. RD-65-130, Part II, December 1965.

3. D. W. Robinson, J. M. Bowsher, "A Subjecti.ve Experimentwith Helicopter Noises," J. Roy. Aeron. Soc. 6'j, 635-637September 1961.

4. W. J. Dixon, F. J. Massey, Jr., Introductloq to StatisticalAnalysis, McGraw-Hill Book Company, Inc., New York, 1951.

5. G. W. Crosse, I. M. Davidson, T. J. Hargest, M. J. Porter.,"A Controlled Experiment on the Perception of HelicopteiRotor Noise," J. Roy. Aeron. Soc. 6., 629-C32, October 1960.

6. K. D. Kryter, K. S. Pearsons, "Judged Noisiness of a Bandof Random Noise Containing an Audible Pure Tone,"J. Accust. Soc. Am. L, n06-112, 1965.

7. K. S. Pearsons, "The Effects of Duration and BackgroundNoise Level on Perceived Noisiness," Tech. Report FAA-ADS-78, April 1966

APPENDIX A

Instructions for Judgment Tests

INSTRUCTIONS

Judgments of Noisiness (Method of Adjustment]

The purpose of these tests is to determine the relativenoisiness ol" various sounds.

When you move the control switch to "standard" the lightwill glow and you will hear a noise; this noise will repeetitself over and over umtil you move the switch. When youmove the switch to "comparison" you will then hear a differ-ent nol!e. The overall intensity of the comparison noisemay iýe controlled by turning the knob on the "level control."

Mooj job is to listen first to the standard noise, then toiiste.,° to the comparison noise, and then to adjust the inten-sity of the comparison noise until it sounds as noisy to youas the standard. By eaually noisyk we mean that you wouldJust as soon have one as the other in or outside your oeperiodically 20 to 30 times during the day and n•ight. Statedanother way, we mean by eouallv noisy that the comparisonnoise would be no more nor no less disturbing to yo in 2zoutside your home than the standard noise.

You may turn back and forth between the two noises as oftenas you wish and listen to each as long as you wish. It issuggested that before you proneed to equate the comparisonnoise to the standard noise you make the comparison noisemuch more intense than the standard; then make the compari-son noise much less intense than the standard. With thoselimits established, adjust the intensity of the comparisonnoise until it would be just as noisy as the standard noisein or outside your home.

You will notice that you can switch from the standard tocomparison and vice versa only during the brief pause thatexists between the end and the beginning of each noise. Whenyou feel the two noises are equally noisy, press the "finished"button and turn the switch to comparison. Leave the switchin tht. comparison rosition until the light goes out. Thenswitch to the neutral position. Proceed with the next judg-ment when the standard light goes on.

A-I

IN. I2RUCTIONS

Judgments of Noisiness [Method of Paired Comparison]

The purpose of these tests is to determine the relativeno.siness of different sounds. The tests are part of a pro-gram of research designed to obtain information that will beof aid in the planning of military and civilian airports andfor noise control purposes in general.

When the tests start, you will hear a number followed bytwo noises presented in quick succession. The number repre-sents a pair of sounds. Your job is to punch a hole inColumn 1 or 2 corresponding to the noise (the first or second)which you feel is noisier or more objectionable. Please makea judgment for each pair of noises, even though you feel youmay be guessing.

In making this judgment, assume that the noise would occurat your home 20 to 30 bimes during the day and night. Pleaseremember to include in your judgment the total effect of thesound which may include intensity level, duration, and typeof sound, rather than maximum intensity level alone.

Please write on the back of your answer card your name,age, occupation, sex, seat number, and the date. Pleaseremember to use the same seat location each time you take thetest.

A-2

!I

21 2 21 21 m 41 m 81 w 121 m2 • m 42 82 m =1223 m 43 m 83 n 123 m4w 4 4 84 m 12 4 m5m 45r 85m w125

i 46 86m 126i7 7 47m 8 7 m1278 m 48 m m 8 8 128 m3 9m 49 mm 8 9 12 9 m

m10 mO 90l 130 w1 m wl 51 91 131 l

Sin1 2 52- m m 2 13 2 wm13 33 s 93 m 133in 14 54•w 94 m 134

s15 53 e m in 9 5 135m 16 m 56 96 136

1 7 m 5 7 w 97 m 137i 8 m 5 8 w 9 8 w 138

I 19 m m 59 m 99 13920 m m 100 m 140

m 21 61 m 101 m 14122 m 62 m= 10 2 142

m 23 6- 103 m 14324 m 64 104 14425 65. 105w 14526w n 66 106 m 14627 o 67 = o107 14728 m a 68 10Sw 1482 m m 69 109 m 14930m 70 110 m 150

m31 m 71 111 m 15132 w 72 m112 152

m33 n 73 113w 153m 34 n 74 m114 154

35 m 73 w 115 m 155m 36 76 w 116 m 156m 37 m 77 m117 157

38 = 78 : 1118 156m 39 m 79 =119 159m40 0 120, 160

Fl

FIGURE A-i. SUBJECT'S ANSWER CARD1