Chapter 1

Basic Theory
 SOUND AND VIBRATION

Sound is a vibration in an elastic medium such as air. vcaler, rnost building

materials, and the eanh.'An elastic mediurn returns to its normal state aftor a
force is removed. Pressure is a force per unh area. Sound energy progresses
rapidly, producing extremely ernall changes in atmospheric pressure, and can
travel great distances" However, each vibrating particle moves only an infinites-
imal amount to either side of its normal position. lt "bumps" adjacenl particles
and imparts most of its motion and energy to thern. A full circuit by a dis-
placed particle is called a cycle {see illustration below}. The time requirad for
ane complete cycle is called lhe periad and the number of complete cycles per
second is the freguency o{ vibration. Consequently, the reciprocal of frequency
is the period" Frequeney is measured in cycles per $econd, rhe unit for which is
called the heru {abbreviated Hz).

Vibration of Farticle in Air

The back and forth motion of a cornplete cycle is shown belcw.

?asilion ol ,
mgximrJm \
displacement \

Amp.litude
?r.*i T
uq, d isplacsmeni
during, vibrafian)

Fure Tones

A, pure tone is vibration produced at a single frequency. $hown below

is
the variation in pressure caused by srdking a tuning fork, which produces an a!
most pure tone by vibrating adjacent air molecules. symphonie music consists
cf nunrerous tones at different frequencies and pressures {e.g., a tone is cam_
pssed of a fundamental frequency with multiples of the fundamental,
called
harmonicsl. To find the period corresponding to a frequency of vibration, use
the following formula;
1

'p- f
where Io = period {s/cycle}
f = frequency {cycles/s or Hz}

'Abse is unwanled sound {e.9., annoying sound made by other peopre or very
lcud saund whieh may cau$s
heanng loss .
)

2 grsc rlEtrY

t
For example, a frequency of 63 Hz has a period Iof 1/S3::*0.02 slcycle
{roughly 30 times longer than the period at ?OO0 H:} "

dt
-{} Compreggian
5 (rnaiimum pressure)
L
6-
€
o
L &(l Normal almospheric
:a PresSure
tn
{t} €
g
6_
.t' R.arefaction
c:: (mininurn preEs$re)
$
tn
lirne,

Complox $ounds

The variation in pressure caused by speeeh, music, sr noise is shown

below. Most sounds in the everyday world are complex, consisting of a variety
of pressures which vary with time. The :hreshold of hearing for humans is one-
millionth af normal atmospheric pressure.

Hi6h so,lnd pressure

(laud Sounds')

Lorrt Sound presgure

(l (eoFt sounds)
L
:('}
lo
s,
t-
a}-
Narrnal alrnosph€ric
-(,
{ Pre6EUre
{')
dl
i rrnO

ersn ffrmv 3
FREOUENCY OF SOUND

Frequency is the rate of repetition of a periodic evenl. Sound in air con-

sists of a series of compressions and rarefactions due to air panicles set into
motion by a vibrating source. The frequency of a sound wave is determined by
the number af times per second a given molecule cf air vibrales aboul its neu-
tral position. The greater the nurnber of complete vibrations {called cyalesl,
the higher the frequency. The unit of frequency is rhe hertz {Hz}" Pirch is the
subjective response of human hearing to frequency. Low frequencies generally
are considered "boomy,'" and high frequencies "screechy" or "hissy."
Most sound $ources, except for pure tones, contain energy over a wide
range of frequencies. For measurement, analysis, and specification of sound,
the frequency rEnge is divided into sections {called bandsl. One common
standard division is into 10 octave bands identified by their center frequencies:
31"5,63, 125,25A,500, 1000, 2O0O,4000, 8000, and 16.000 Hz. .An oc-
tave band in sound analysis. like an octave on the piano keyboard shcwn
below, repre$ents a frequency ratio of 2:'l . Octave-band ranges of three other
rnusical insiruments are also shown below"

Freq"uencg (Hz)
5l. r b5 t50 et\n I 000 2000 +000
I
\l I

Violin

E,DETOAB
I octa{e

7 octave,g

Further divisions of the frequency range (e.9., one-third or one-tenth oc*

tave bands) can be used for more detailed acousticalanalyses, $ound level
meters can measufe energy within octave bands by using electronic filters to
eliminate the energy in the frequency regions outside the band of interest. The
sound level covering the entire frequency range of octave bands is referred to
as the overall level.

4 ::3:::nicpY
WAVELENGTH

As sound passss lhrough air, the to-and{ra motion of the particles alter-
nat*ly pushes together and draws apart adjacent air panicles, forming regions
of rarefaction and compression. Wavelengtlt is the distance a sound wave
travpls during one cycle of vibration. lt alsc is the distance between adjacent
regions where identicalconditions of panicle displacement occur, as shown
below by the wire spring {called a "slinky" roy } . When shaken ar one end,
the wave moves along the slinky, but the pafiicles only move back and forth
about their normal po$ition$"

ry
rWlvelEn6f h,

Sound waves in air also are analogous to the ripples {ar waves} caused
by a stone dropped into still water. The concentric ripples vividly show pat-
terns of molecules transferring energy to adjacent rnolecules along the surface
of the water. ln air, however, sound spreads in all diroctions.
To find the wavelength of sound in air at a specific frequency, use the fol-
lowing formula:
1130 .
'" f
where I = wavelength {ft}
f= frequency {Hz}
Shown below is the wavelength in air from the to-and-fro rnotion of a vi-
braring tuning fork. The movement of the prongs ahemately compresses and
rarefies adjacent air particles. This cyclical mction causes a chain reacdon be-
tween adjacent air panicles so that the waves {but nor the air parlicles } prop-
agate away from the tuning fork. Remember sound travels, but the elastic me-
dium only vibrales.
r- Ra.cl aci on ( dieot aced air
/ part,cles s?re^d'?p2&)
Wavelen<th ( r)
/ -Co^oression (d,gpla ced air
r----1 / J/ ?artictes togeti'sr)
Pr.rshed

luning forlr
(urt,h Pnon5e
viuratihf, )

n.qsc rHroRv 5
SOUI\ID SPECTRUM

Because mgst sounds are complex, flucrualing in pressure, level, and fre-
quency content, the relationships between sound pressure level and frequency
are required for meaningful analysis {data sc plotted are called a saund spec*
truml . This requirement is similar to indoor climate control, where thermal
comfort cannot be specified as a 70'F temperature alone because comfort
also depends on relative humidity. air motion, and so on. Sound spectra are
used to describe the magnitude of sound energy at many frequencies. The fre-
quency scale given below is an octaye-band scale because the ratio of succes-
sive frequencies is 2:1. the ratio for an oclave in rnusic. ln acoustics. the eX-
tent or width of octave bandS is geometric. For example, ths octave band for a
center frequency of 125 Hz contains sound en€rEy from 125 + '/2 Uzrc 125
x €He.
The line graph ar the right depicts the octave-band $pectrum for a noise
consisting of the sound energy measured within octave bands {see bar graph
at left). The line graph is plotted at the respective center frequencies of the
bands. Also shown on the graph is the sound level of a 512-Hz tuning fork.
Note thal the tuning fork produces sound energy at a single frequency only" A
tuning fork will vibrate at the same frequency if struck lightly or forcefully, but
the sound levels produced can differ greatly.

Sar $raph pictura o$ eound Line graph pictuna o{ sound

Her6ht o$ bar shoLls 6ound Dat is height'o{ bar

leve I eau iv llenl tro eodnd Sraph at a spec'fic octave-
ener(u in snadedd area oe^4 cenlgr Sraq'rency
( cal e-d'' band id! h )
I rr.'
rS
rO
-"a
-a Line connacts ddts ia ghoru
Banduidth composite pict,urg of lreorrencieE
6
n
;7 conlained in a Sound (called
q :l "gound epectrum')
{J
$
L
! - :
L
rrr
(t tt
6
$i2 Jt tunin{ iork (v;brfves
qt
!
q)
c
c*
ai sing:ie irdquancg)
{:-
-B -\t
c c
5 a
(n {o
2r0 500 1000 125 250 500 t000
FreguancS(llz)
Range lor
oLt|le'oand cenler
lreq,uencg o€ t25 t{r

lf a pianisr uses both forearms to simultaneously strike as many piano

keys as possible, the resulting nsise will be broadband because the sound pro-
duced will be spread throughout a wide range of frequencies" A graph of this
noise, therefore, would plot as a wide, flat spectrum.

6 aas*:;.€cqv
VELOCITY SF $OUND

Sound travels at a velocity that depends primarily on the elasricity and

density of the medium. ln air, at normal temperature and atmospheric pressure.
the velocity of sound is approximately 1130 feet per second {ft/s}. cr almost
S0O mi/tr. This is exlremely slow when compared to the velocity of light,
which is about 186,000 mi/s. but much faster than even hurricane winds.
ln building air distriburion systems, the air velccity at registers, diffusers.
and in ducts is so much slower than the velocity of sound rhat its effect can
be neglected. For example, an extremely high air velocity of 2000 ftlmin
{about 33 ft/s} in a duct is less than 3 percent of the velocity of sound in air.
Consequently, airborne sound travels with equal ease up$tream and down-
stream within most air ductsl
However, scund may travel at a very fast 16,00O ftls along steel pipes
and duct walls as shcwn belcw. lt is therefore important to block cr isolate
paths where sound enersy can travel through building materials {called
structure-borne sound ) to sensitive are€s great dislances away where it may
be regenerated as airborne sound.
ln buildings, the effect of temperalure on sound alsa is negligible. For ex-
ample, a 2O'F rise or drap in room air temperature is significant. but would
eause only a 2 percent change in the velocity of sound in air.

Air
ftr ))
$teel
Fan-coiIunii
Airborne
gound Pin (contccts {ioor
cauSt n8 glrucl,vre - oorag
sound )

$irucNrira - barne gotnd in

steal pipes ( l6,00tltis)

CireulaNing pump

Boi I er Air bornS gou nd

Sirr"iclure - borne
gor,ind tn tancret,e
( t? 000{tls)

6tructurel iee
Reiniorcing: sLeel .(stzoel
has tess 1amptng'.+t1an
cancreie go 6ouid uili oe
Concrete transrniiied rnore aasi lg )

ensrc rueonv 7
FREQUENCY RAIJGES OF AUDIBLE SOUND$

Hearing ranges for both young and older persons 2A years ald ) are >
shown below. ,A healthy young person is capable of hearing sound energy
from about 20 to 2O,00O Hz. Hearing sensitivity, especially the upper fre-
quency limit, diminishes with increasing age even without adverse effects from
diseases and noise-a condition called presbycusis. Long*term and repeated
exposure to intense sounds and noises of everyday living can cause permanent
hearing damage {called socrbcusr's}, and short-term expcsure can cause tern-
poraryloss. Conoequently, the extenl of the hearing sensitivity for an individual
depends on many factor$, including age, s€x, ethnicity, previous exposure to
high noise levels frorn the workplaee, gunfire, power tcols, rsck music. etc. All
other hearing losses {e.g., caused by mumps, drugs, accidents} are ealled
nosocus,'s. An audiologist should be consulted if a "ringing" sensation occurs in
ears aftsr exposure to moderately loud noise or if sounds Seem muffled or dull.
Also shown below are frequency range$ for human speech {divided into
consonants, which contain most of the information for articulation. and
vowe/s!, piano rnusic, steroo ssunds. and acoustical laboralory test$ {e.9.,
tests used to determine ab$orption and isolation properties of building mater*
ials). Human speech contains energy from about 125 to 8000 Hz' Women's
vocal cords are genetally thinner and shoner than men's, sa the wavelengths
produced are smaller. This is lhe reason the female frequency cf vibration for
speech is normally higher. Wavelengths in Sl and inglish units are indicated by
the scales at the top of the graph above the correspondinE frequency.

Wnclenglh ,ez\eg
{4n t2na 1lm 6,!n 2.S* l.4n 0.'lm 3{cln l?crn &,6cnr 4.3cn 2.14 lcn
1r+{' 7t' 3gI rg' E' 4.5' a,?5' lll&.' a/'{ a{ f,4 ?;
tt tltltr

I
& 31 5 63 tz5 25a 500 obo zo'oo qotoo aobo
sr*
Frequencg (r{r)
'Vibra*ons below 20 Hz are noa audible. bul csn be fel1.

Reference

E. H. Berger et al. {eds. } , Noise and Hearing Conservation Manual, American lndustrial
Hygiene Association, Akron, Ohio, 1986.

8 easc rl€mv
SFN$ITIVITY OF HEARING

The graph below shows the trernendous range o,f sound levels in decibels
{abbreviated dB }" and frequency in hettz over which healthy young persons
can hear" Also shown on the graph is the frequency range ior "conversational"
speech, which occurs in the region where the ear is most senshive. For com-
parison, the region where symphonic music occurs is indicated on the graph by
the large shaded area extending at mid-frequencies from below 35 dS lo sver
1OO dB {called dynarnic range}. The dynamic range for individual instrumenrs
can vary from 3O dB (woodwindsi to SO dB istringsi. The lowest level of
rnusical sound energy that tan be detected by the audience largely depends on
the baekground noise in the musie hall {see Chap. 4}, and the upper level de-
pends on the acous{cal characterisiics of the hall {see Chap. 3 } . Electroni*ally
amplified rock mu$ic in arenas and coliseums far exceeds the maximum sound
levels for a large symphonic orchestra. Rock music, purposefully amplified to
be ar the threshold of feeling {"tingling" in rhe ear}. is considered to be a sig-
nificant cause of sociocusis.

,/-Tnrgg5.ici of Teeiing, curve

( iisreners':"'
*-icKre' ne?re)

\/

-\--"/

Dgna,nic range
(g
{or sgrnphonia
..9 rissic

,tl
{i
-s
3
<)
ctr

f requencg ( t'la)
senr,t'ivitLl ocarrS
l"laxitn.rm
here a14000 ril d'ie to
ragonanca ;n ear c?nal

ThreEhold ol a*dibilit$ curve

(rninr rn,.irn so.rnci press,ireg eer
caa det,ecl)

.gecibelisthe unh u$ed ro expres$ the pressure {or intensily} level of *ound en€rgy. ln this book. sounellevelis
alway; measured tn decibels by precision sound level meters at a specilie fi*queney or weighting.

sssc rHeonv I
INVERSE-SOUARE T3W

Sound waves from a pcint source outdoors with no obstructions {called

free-field conditions l are virtually spherical and expand outlvard from the
source as shown below. A point source has physical dimensicns of size that
are far less than the distance an sbserver is away frcm the source.

Disiance dr doubled ai
gosit,ion 2 ( erea oi uate
qvadrql?d,5o ;nie.nsii! is Vq)

Porlian of gound ureve

ot positian I

lglitee {d,1
D, stance
q]' 2d,

Power is a basic quanrhy of energy flow. Although both acoustical and

electric enargies are measured in watts, they are different forms of energy and
cause different re$ponse$. For instance, 10 warts {abbreviated w} of electric
energy at an incandesent lamp produces a very dim light, whereas 10 w of
acoustical energy 6t a loudspeaker ean produce an exiremely loud sound. peak
power for rnusieal instrurnents can range from 0"o5 w for a clarinet to lb w
for a bass drum.
The intensity from a point source outdoor$ at a distance d away is the
sound power af rhe source divided by the total spherical area 4ndr of the
scund wave ar the distance of interesl. This relationship can be expressed as:

.W
'- 4nd2
where ls sound intensity {W/mr}
Wx sound power {W}
d = distance from sound source (m )

lf the distanee is measured in feet, multiply the result by 10.76. because .l ml

equels 10.76 ft2"

10 g*stc rH€oRY
 The rnversa-square hw tor sound is:

,,
tz= 14\'
\dl
Is scund intonsity {Wlmt}
d= distance from sound sourcs tft sr rnl

Note: To derive the inverse-squarg law, consider a wavefront at pnsitions 1 and ? as

shown on the abnve illu*ration. Ar position l,lN = lt4wd12, and at position 2, W =
lz4t d*. Since the ensngios are the sarne {because the source is the same } , ll4xdtt c
l"4xdf. Therefore, \/12a Qf 16ri, which is tho inverse'square lsw for sound'

an$crHEoRY 11
DECIBEL$

.dB wd
ren
- l0t'
( Threghald ol pain)

.I

(Threshold of audibilitg )

Ernst Weber and Gustav Fechner {nineteenth-century German scientisrs }

discovered that nearly all human sensations are proportional to the logarithm of
the intensity of the stimulus. ln acoustics, the belunit {named in honor of
Alexander Graham Bell) was first used t* relate the lntensriy of sound to an in-
tensity ievel corresponding to the human hearing sensation. Sound intensity
level in bels equals the lagarithm of the intensity ratio ll1o, where lo is the min-
imum sound intensity audible to the average human ear at 10OO Hz. gecibels
{prefix deci- indicates that logarithm is to be multiplied by 10} can be found
by the following formula:

l, = '10 log I,o

where Lr = sound intensity level {dB)
/ = scund inrensity {Wlmt}
/o = reference sound intensity, 'lfl"te {\41/6?}

The iilustrarisn on the following page gives the decibel level of some fa-
miliar sounds. The human hearing range frcm the threshcld of audibility at o dB
ro the threshold of pain at 130 dB represenrs a tremendous intensity ratio of
10 trillion {1O,OOO,OO0,O00.OO0) to '1. This is such a wide range of hearing
sensilivity that it may be hard to imagine ar first. For example, if a bathraom
scale had a sensitivity range comparable to that of the human ear, it would
have to be sensitive enough to weigh both a hurnan hair and a 30-story
building! Logarithms allow the huge range of human hearing sensitivity to be
conveniently represented by smaller numbers.
It is difficulr ro measure sound intensity directly" However. sound intensity
is proportionsl to ths square of sound pressure, which can more easily be
measured by sound level meters. ln air under normal atmospheric conditions.
sound intensity level and scund pressure level are nearly identical"

1 2 easc rHeonv
COMMON SOUNDS IN DECIBELS

Some commr:n, easily recognized sounds are listed below in order of in-
creasing scund levels in decibels. The sound levels shown for occupied rooms
are only example activity levels and do nor r€present crkeria for design. Note
also that thresholds vary among individuals.

9ublaciivo
Dqqibqlq * 9rrmpleg gval uetion

140 rJei engina ('l5li au.rag)

Thrashold ---{ r30 r Jet aircrrsr during rakao{r ( goo +r ,ru, :illiiri?Y
o* prin
la0 (uilh elect ronic ampiificai,lon)
J-I
Thrgshold -*{ r'l-lerd
"ocli'band
of {teling
:lH*'ffiiJJ,)",,,. ai raru {i a,rasr o"r'^;nr
'0 H
loo
lt*iahl![,(!1*],iff.1
-rv JPrialin6
an press
ls,.*, I H Ver,{ loud
o Pneumatii concraLe bncrker A "
lConrputer e+uipnent roon H
Threrhald *{ bOt o Ca$'e,e|e u,iilr so'ind-re{lecling svrlzc.zs
ol heaning loss
( long-teFm
l0 . b-757 etrcref|cabin durinr $li<trt =E Uoua
a Cracr<l ia{ of ?leslie Sood ,*r?ap[ers (2ll auag)
er.Posu re )
fl
60 aNear hi(hurau tre$f ic ( r,rhen ) 55dbA, road -a
Sgaach and rsil-irstfic 8nno3 rno€f people) E
50 o0{f ice aEiivitirs Hoae.are
H
t+0 l9E{t *l'erea nrugie in reaidenae

30 .*;lgry.t uilhout siereo prrsinf ttcte et

20 ownisper I0.,^,
---T
f Audiernatr"ic teStin6 ruorn H
l0 t,(ustia # ieaves ia brz,cte H Ver..{) Ssint
o Hurnan urcaihing'
Threshold O J
or h?arin( -{
(or aud'ba t!)

.dBA are weiglrred valuGs rneasured by r sound level rneter. See page 31 {or details oi electronic weighring net'
works which modiil the sensirivitY of met€rs.
f50 ft from a molorcycle can equal lhe noise level al less than 2000 tt frorn a ier aircraft.
lContinuous exposure to sound energy abcve 8O dg,A ean be hgzardouc to heahh afid can cause hearing loss lor
some pargons.

ansc r*onv 13
ruOfSE REDt,CTIOI\J WITH DISTANCE

Outdooro in the open and away from obstructlons, scund energy from
poinr sources drcps off by 6 dB for each doubling of the distance from the
source. {,According to ih€ inverse*square law, the intensity ratio for a doubling
of distanee is 32 = 4, and rhe corresponding decibel reduction is 10 log 4, or
6 dB.) Saund energy from line source$ {e.g., stream of autornobiles or rail-
road cars) drops off by 3 dB for each doubling of disrance. This is because
line sources consist ol successive point $ourc6s which reinforce each other.
Thus the spread of sound energy is cylindrical, nof spherical. Cylindrical surface
areas increase in prcportion to the radius {distance}, whereas sphericalsur-
face areas incrsase in proportion to ihe square of the radius. The graph below
shcws noise raduction due to di$tance for paint and line sources. Additional
reductions can bo eaused by large buildings, earth berms, trees and vegetation,
and other environmental effects {see Chap. 4i"

25
rPaint solrCe (E?lne.tcai
i\ anreaoing at.6ab ?sr doroiing
al d.1f-tnLL )
2a
rO
-1J ,-

g )/7 L;n2 6orrae, (cul indrica I

o
-rt
(, i0
,/ \'"!iii:ii{":\too r":.0:^*
t-l
3
al
ql
.9
? L.-
{t
2"0
r0 30 a0 lz0
Dist'ance {ratn source ( fl )

An ares saurce, produced by severar adjacent line sources ie.g., rows of

cheering spectstors at spOrt$ events ) or large radiating surfaces of mechanical
equipment, has little reduction of sound energy wirh distance close to the
sourc€. within distances of blrcto efx, where b is the short and c the long di_
mensisn of an area source, sound energy drops off by 3 dB for each doubiing
of distance. Beyond distsnces ot cfn, the drop-off will be 6 dB for each doub_
ling of distance outdoors {cf.. E. J. Rathe, "Nole on Two common problems
of sound Propagation." Journal of gound and vibration, November 1g6g,
pp.472-4791"

b= thort
{ imension

14 sesrc rrfonY
LOGARITHMS TUAPE rASY

tognrithm Basie*

The f*rst $tep to find the logarithm *f a nurnber is to express it as a digh

from 1 to 9 multiplied by 1O to a power. A logarithm usually consisrs cf two
parts-{h€ ctaracteristic, which is the power of 10, and the rnanfissa, which is
rhe decimal found in log tables {or from pocket calculators}. ln solving loga-
rithms, remember that

tS = 100,000
10a = 10,000
103 = 10ffi
10c = 1O0
10t = 10
l0o = 1 { * means equal ro by definition }
10-1 0.1
=
10-? = O.01
10-3 = O.0O1

and when thg decimal point is shifted to the left by n plac*s, the number is to
be muhiplied by 1O'; when the decimal is shifted tCI the right by n places, the
number is ts be divided by 10". This may seem complicated at frst, but after
reviewing a few examples it should become routine.

4,82O,OOO,O = 4.82 X 10F x E x lS 1- means approximately equaltol

\.---r'--
Numbers ending in 0.5 and greater should be rounded up as shown by the
example above. lf less lhan 0"S, the decimal should be dropped.

0.0009258 = 2.SB X 10*s :* 3 X 10-6

8,400.000,000.0 = 8.4x'tff* 8{;*L
The following shortened logarithm table can be used to quickly find the
manlissa of numbers from 1 to L

A U$EFUL !:sG rA?!9""- ,

Number Mantisea
I 0
2 0.3
3 0.48
4 o"6
5 o.7
6 0.78
7 o.85
I 0.9
g 0.95

nsserHconv 15
 ln almost all acoustisal problems. it is not necessary to work with small
fractions of decibels" Use either the log table above, or a four-place log table,
and round the final answer to the nearest decibel. A pocket calculator that
finds an entire logarithm in one srep is very handy when working with decibels.
The following examples represent logs of very large and very small num-
bers. Remember, the first *tep is to arrange the number as a digit times 1O to
a power.

log (4,82O,OOO.0) = log (5 x tOs) = 6.7 = 6.7

-;rHJ:rH:J
log {0.Ooo0?581 = log {3 x 10-5} - * los {*. ,o*
* - log {0.33 x 105) = - log {3 X 1ffl =
F4.aA
log {8,400,000,CI00.0} = log {8 x 10e} = 9"9*

Antilogarithms

The antilogarithm of a quantity, such as anrilog {x}, is the number for

which the quantity x is the logarithm. For example.
I
-1
antilog (6.7i = 5X 1S= ls x toB
L:mrrr;HJ
antilog {*4.48} * -3 x fO. = 10-{ x O.33 X = ,3 X 'lO{
*X 1O-4

whan the mantissa of a log falls between values in the log table on page
15, use the closest mantissa to find the corresponding numbei from 1 to g.

Properties of lngs

l. log xy * log y
= log x

2. logl=logx-logy
"Y
S. logx,xnlogx
4. logl=ff

'This propeny is impartam in acaucliial analysis b*c€us€ op€ning€ ln building olenrents have
no resistance to
sourd flow which then can be expressed as 0 dB el isolation.

l6 aasn THEoRY
Powers of 10 Review

Remember, the symbol lF is a shorthand notation for 1O X 1O X 1O ;

1S00. Also. the product of two powers of the sarne nurnber has an exponent
equal to the surn of the exponents of the two powers:

1Yx 1S* (10x 10| x {10x 10x 10} = !$r

or
108 X 1S = 10t2*e' = lOE

Addhional examples follow:

X 1tr = 1O{7+$} s t$la

107
10-l _ 10-€ X 10+12 * ,l0r*s+ rar s
10-rr - 108

When combining exBonent$, be careful ol the signs. Fividing by a negative

exponent such as 1O-12 is eguivalent ro multiplying by its reciprocal, 10+1?.

*1-8:'
lCI.rr = 1O*3 x 10+12 ='l${-r+ t}} =1ge

You have now lesrned to hsndle powers of 1O and logarkhms. which arc
fundamental relationships needed to describe how hurnans perceive sound and
how building materials affect sound energy. Several examples are presented
below and on the following pages.

Exanples

1. The intensity t of a rock music group is 8.93 X 10-2 Wlm''

Find the carresponding sound inlensity level L1
I
Lr = 1O log 1d1,
s 10 tos
gqe$ = 10 tos {8.e3 x 1010}

Lr = 1O { 1O.95O9} = il10 dB

ensc rusmv 17
 #,))il
?. Loud speech, msasured at 3 ft away. has a sound intensity level l, of 73 d8.
Find the conosponding intensity L

I
Lr = 10 log
10-o
73 = 10 rog
rok
Next, divide both sides nf the equarlon by 10"
I
7.3 = log 1o_1t

The above expression states that the log of a ratio ll/1Oaz1 is equal to 7.3.
l&hen the number for which the log is 7.3 {i.e., antilog ) is found, ser h equal
to ths ratio.
I
antilog (7.3)
I

= 1.995 x 107
Lfrom mantlssa taue I
ot pocket calcr-dator I

Therefore.

r.sgs x 10?
lo-t2
and by cross muhiplicatiun

l= 1.995 X 1O7 X 10-u = 1.995 X lO-a Wlmr

t8 sasn fieonv
EXAMPLE PROBLEM I INVER$8.$OUARE TAW}

lOlf ( dr)

1. A car horn outdoors prcduces a sound intensity level ly of 90 dB at 10 ft

away. To dnd the inten$ity l1 at this first lacation, use
I
Lr = 10 log
*-r2
eo= 10bg*k
t.
9.0 = log
*:A
antilos {9.? - }0 x 10s
t^
1.O x 't0e = ,g_a
d = 1.o x 1Os I 10*tz = 10-3w/me at loftaway

2. f the sound intensity / is known at a given distance in feet away from the
source, sound power W can be found by the following formula'

t= #x 10.76

By cross rnultiplication

w*4trdzx *$x
Since l, = 1O-3 Vy'/m2 at 1O ft awaY

w= 4x 3.14 X 103 = 0'12W

" *kx
10-3

BAsrcrHsoRY 1g
 3, The intensity level ir ar 80 ft away can be found by the inverse-sguare law.
First, find the sound intensity I, at the location 8O ft away.

+=er
10-3 /AO\2
r, = \10/

10-3
s4
l2

641, = 1ft3

lz= $* 10*: = &.f0 x 1fs Wlm-4 ar 80 ft away

Next, find 1,.

Lr = 1o r"g 104 - 1o rosu#C

lr = 10log {1.56 x 1O7} o 10{7.1931} = 72 d8 at SO ft
This means a listener moving from location 1 at 1O ft away to location ? ar
8O ft away would observe a change in inrensity level of 18 dB {that is, gO
dB - 73 dB ) . This reduction would be judged by most listeners as "very
much quieter' ' { see rable on the following page } . However. a car horn at
72 dB would still be considered "loud" by most people.

\
\
?eds
Sotnd lo to so
60urce 0ialance {ron sorlrac (+i)

Note; From 1o to 8o ft away is three doublings af distance (i.e., 1o to 20 ft,

?o to
4O ft, and 40 to 80 ft ) . Therefore, three doublings X 5 dg/doubling
tion and lr 3 90 18* s
72 dB at 8O ft away.
= ]g dB reduc-

2O aasr{ffiY
CHANGE$ IN $OUND LWEL

The table below is an approximation of human sensitivity to changes in

sound level. $ound intensity is not perceived directly at the ear; rather at is
transferred by the complex hearing mechanism to the brain where acoustical
sensatisns ean be interpreted as lcrudness. This makes h*aring perception
highly individualieed. Sensitivity to noise also depends on frequency content,
time of occurrence, duration of sound, and psychologieal factors such as emo-
tion and expcctadons {cf., O. L. Angevine, "lndividual Differences in the An-
noyance of Noise," $ound and Vibration, November 19751. Nevertheless, the
table is a reasonable guide to help explain inereases or decreases in sound
levels for many architectuml acoustics situations.

Change in Change in
Sound Level tdB) Apparent Loudness
'l lmperceptible {except for tones}
3 Just barely percaptible
6 ebarly noticeable "
10 About twice {or half ) as loud
20 About 4 times (or one-fourth)
as loud
*For example. distance to the point source ouldoors is halved or doubled'

The change in intensity level tor noise reduction, abbreviated NR) can be
found by:
NR=Lr-L:
and
Nn=10bsf
where NR = difference in sound levels betwe€n two conditions (dBl
Jr s souild intensity under one condirion (W/m')
lr = sound intensity under another condition {Wlm?)

fllote: By substitution of the inverse-square law expression {rom page 11 into the
above formula

NR = ro ,"- (*)
and therefore. in terms nf distance ratia drldl

NR = 2o'"r (*)
for point sources outdoors, where d's are the distanees

slsc rHeoRv 2'l

S(AMPLE P*OSLEM {MU$IC MAN }

)l \
<-**"-p

The measured sound intensity level l, of sne trombone is 8O dB. To find

the saund intensity level l, from 76 trsmbones, first find the intensity lr of ona
trombone.

t,' = 1O loq!
"lo

Bo = 1o,ogT#
B.o = los
#t
1.0x 108 =Jt-
10-12
lr=1'OX1O8x10*12
lr = 10{rVim?,lor onetrombone

To combine the intensities of 76 trombones. each producing 8O dB at a lis-

tener's position, find the inr'ensity ls ol 76 trombones. l, will be 76 X l, =
7SX 10-4W762,
t, = lolog# tO

- 1o #k ros ;c 1o los lg*-"-

= 10log {7.6 x 10e}
lr * 10 {9.8808} = 99 dS for 75 trombones
This is not a$ great an increase as might at first be expected" lt would
take '100,00$ trornbones to reach the threshold of pain at 130 dB (although
&e threshold of disgust might be reached at a much lower level) . A cornposer
is aware that a large nurnber of instruments playing the same score may fiot
produce a tremendous sound impression. Large numbers of instruments are
used to achieve lhe desired tanal texture or blend in the overall sound from the
individual in$truments. for example, one sols violin by its location and fre-
queney range may dominale portion$ of an orchestral performance.

2t s*sa THEoFY
DECIBFL ADDITION

Because decibels are logarithmic values, th€y cannot be cornbined by

normal algebraic addition. For example. wh€n the decibel values cf two
sources differ by O to 1 dB, 3 dB should be added to the higher value to find
the combined sound level. Therefore, the sound level of two violins" each
playing at 60 dB, would be 8O * 3 = 63 dB, nor60 + 60 = l?0 dS (which
wsuld be near the threshold of painl). This is similar to lighting, where two
35-W fluoressent lamps are not twice as bright as one. The following table
can be used to rapidly combine sound levels.

When Two dB Add the Following dB

Values Diff*r by to the Higher Value
0or1 3
2or3 2
4to8 1

I or mcre 0

When several decibel values are to be added, use the table to find the
combined value by adding the decibels twg at a time. For example. ro find the
combined sound level of 34 dB, 41 dB, 43 d8, and 58 dB, add as follows:

dv
nnn,

t1?
b=" '/
5g -r
o?
db

x\='lr
// -J
sb ,
\ Jbab

:,

Nodce rhat 43 * 58 3 58 dB and 34 * 58 = 58 dB because the higher

sound level {by > I dB} swamps out the lower sound
level'
To find the combined sound levelof 82 dB. 101 dB, 106 dB, 102 dB,
g0 dB, and 78 dB, add as follows:

sasc rxroRv 23
 4_9.

\ ''o'
,to

= lQbAb

using differenr orders of addition may give results that differ by 1 dB,
which is normally not too significant. Howaver, to echieve the greatest preci-
sion, combine decibels logarithmically. For example, using a po"t
*t calculator,
first compute 1o', where x is the deciber varue divided uy to,
sum 1o'an-
swers for all decibels to be combined, and then take to tog
of the sum. The
simplified table method can be used to check your
The table also can be used to subtract deciber "n**er.-
varuos. Far exampre, it is
necessary to subtract decibel values to find the noise from printer
a if the mea-
sured noise level in a computer room is go dB and the background
noise level
with the printer turned off is 73 dB. According to the table, because
the rwo
decibel values differ by ?. subract 'r dB from the higher
varue. Therefore, the
printer noise alons would be gO
- 1 79 dB.
=

Note: a number of equar deciber varues are to be cambined,

.lf add ro rog n to the dec-
ibelvalue, where n is the number of eguar decibervarues.
bones at an l,of 80 dB each. rhen the rotal l,
a;;r;;;, *
iin zo trorn_
= I
gO + rO rog 76
6|Oi fOf t.gggg)
= 99 dB for 76 trombones.

24 rasn ruronv
 HUMAN EAR

The sketch below is a general description of how the human ear functions.
The ear can delect sounds over a wide range of loudness and frequency. ln
addition, it has the ability to detect individual sounds {e.9., familiar voice} from
within a complex background of loud, unwanted sounds (in a noisy, crswdod
room, called cocktail party effectl. However, perception of speech can be
nearly impossible at noise levels above 80 dBA {cf., J. C. Webster, "Noiee
and Communication" in D. M. Jonas and A. J. Chapman {eds.), JVoise and
Societr, Wiley, New York, 1984).
$ound energy, whieh travels through the ear canal, first irnpinges on the
eardrum membrane eausing h ts vibrate. Eardrum vibrations are then trang-
mitted acro$s tha middla ear by the lever action of three small bones {called
hanlm*r, anvf, and stirrup due to their shapes) . The motion of the eardrum
ovsr a large area is thereby converted into a more forceful motion over a
smaller area of the stirrup which contacts the oval window of the snail-shaped
cochlea. The middle ear cavity contains air at atmospheric pressure dus to the
eustachian tube which connects to the throat {this is why ir helps to swallow

l'4iddla ear

9xternal ear (shapad lo Elr bonag ( hamrner, anvil,

el\ance 9.0vr'd ?redsure et' rnd etirup )
r-gcdrurn and la help
czTer raine locat i on' oi 6amicircular cana l5
*,tnd 6ourca) tlo conlrol equilibrium )

Auditorr{ nerve
l;bers (gend ira?ulEes
to braia ulhere
Ei1nals are interpreted
a3 gound)

dochl*a ( hair callr oF

S2s\iar me mbrane vibrat'e
ta aid in f reqlrencg rScognitian)
=-a- caaal ( angtilies qoiad
-esonanoe, brotecls and
=-
:-'zcts lorsnd daves to eardrom)
Eustation tube ( unnection to
ia-dnrm { vibraies to transfar enerS! lhroat, r.uhich providQs Pressure
'-'.au'h ear bo^e5 ao cocAte2 7 1atafice al eaidrum)

frttle,r aar lnner gAr

sastc rHronv 25
 I

lt
ili. or yawn when experiencing sudden pressure changes such as rapid airplane
descents). Vibrations of the stirrup are transmitted through fluid in the cochlea
to hair cells where they are converted into electrical impulses. These impulses
are transmitted by the auditory nerve fibers leading to the brain where they are
interpreted as sound,
Long-term and repeated expo$ure to loud sounds can cause permanent
damage to the inner ear {commonly called nerve deafness}. When it is neces-
sary to shout to be heard by normal-hearing persons less than 3 ft away, the
noise may ba hazardou$ to sars. ln this situation. lvear hearing-protection de-
vices sueh as earplugs {which fit snugly inta the ear canal} , earmufls (which
fit over and around the eari. or both when noise levels exceed absut 10O
dBA,

fieference
L. l. Langley et al., Oynamic Anatatny & Physialogy, McGraw-Hill, New york. '1974,
pp.33G345.

2S ansc ThEoRY
NOI$E EXPOSURF LIMITS

ln 1971, the U.$. Department of Labor established the Occupational $afety

and Health Administration iOSHA) and adopted regulatians to protect against
hearing loss caused by exposure 1o noise in the workplace. The permissible
daily upper limit of noise exposure in A-weighted decibels {abbreviated dBA)
for continuous noise is shown on the graph below for 1983 rules and regula-
tions. Single-number decibels in dBA units are measured by sound level meters
with internal electronic networks that tend 10 discriminate with frequency like
the human ear dces at lsw sound levels. Note thal amplified rock music at 12O
dBA and higher would exceed even the shortest permissible noise exposure.
Exposure to mpulsive noise such as gunfire or impaet noise from heavy ma*
chinery should not excsed 140 dBA peaft sound level.

r20
I t5*
ll0
100
t 0r?rexposura
q0 \
&
dl
380
6t^
>tu
9 PermiEEible
-rl exPosure
c 60
a
o
$50
0 4 s 1Z 1620 ?+
Or:ra?ion per drg ( hr )

.UpBer limit n6r design value } for exFosure to contifiuous noise

{
1n the warkplace withou hedring testing ptogram
or use of hearing-protectron devtces

Alrhough exposure limit* are given in dBA, only sctave-band (or nar-
rower ) analysis of noise will give a more complete picture of how severe the
frequencies' This kind of detailed information {called
'frequency are at specific
problems
analysis) is also needed to determine th* corrective measures be-
from
cause solutia,ns for high-frequency noise problems may differ considerably
those far low-frequency. COrrective mea$ures can involve redueing noise levels
processes),
at the source is.g., by redesign of noisy equipment or industrial
interrupting the pgfh {see ChaPs. ! and 4 far principles ol naise control by ab-
sorption and isolation ! , protecting the receiver {by using individual hearing-
protecticn devices). or combination$ of all lhese measures^ An effective
'l
i.rearing conservation prograrn ncrmally consists of five phases: { } education
determine
{to inform about harmfulsffects of noise}, {2} sound surveys {to
daily noise exposure of workers and identify the sources of noisel, {3i engi-
neering and administrative controls {to reduce noise levels and, if necessary,

elsn rHeonv l,?

 I

to reduce worker$' exposure to noise), {4} use of hearing-protection device$,

and {5} audiometric testing {to examine hearing at regular intervalsl.Ac-
cording to OSHA, hearing con$ervation rneasures are mandatory when workers
are exposed to scund levels of 85 dBA or more for 8 h per day. For the
basics of hearing conseryation programs, see L. H. Royster, J. D. fisyster, and
E. H. Berger, "Guidelines for Doveicping an Effective H*aring conservation Pro-
gram," .Sonnd and Vibration" May 1gBZ, pp. ?l-25.

28 pa*c rnrorv
LOUDNES$ PERCEPTION

The curves on th€ graph below {called equallaudness confours} show

condhions of sound level and frequency which listeners perceive as being
equally loud. Because variations occur from individual to individual, the curves
represent averages for many test subjects. The number on the curves is the
loudness level. called the pfion. Therefore, sounds which produce an equal
sensation of loudness will have ths $ame phon value. Note that an increase in
sound level at low frequencies is usually perceived as being much louder than
an equival*nt increase at high freguencies.
High-frequency sounds {> 2OOO Hz} are generally more annoying than
rniddle- or low-frequency sounds because human hearing is less sensitive to
low-frequency sounds. However. any noise which is abrupt. intermittent, or
fluctuates widely can be extremely annoying. lt also is hard to disregard sound
that contains information such as speech or music.

Ag sound li,vtl ine,reascs, Haxrmrim aensitivilg

eensifivitq becor^es occurs bblu?.en 2000
more unilo-rra urith and {000 l.,lr (r.rhsre
{rgouancq ( conlours mo sl agzcch,alor rnalion
arg'flat{er ) rs concenlrat ed )

L
ot
120 phon qt
+)
.:
i't
.3.E
. tl
6o
uq)
clo *rL

cq,
f .:
; o{J
l.l
-*n
q)
lou 60und 9'n-
-q levels, sensit,ivilg d-6
tll*
:
C
$
io lou trequencras
oll
drops sharpl5
cfl

Freq,rrencu ( Hr)
')
References

D. W. Robinson and L. S. Whiftl€, "The Loudness sf Octave-Bands of Nsise,"

Acustica,
vol. 14, 1964.
S. S. Stevens and H. Bavis, Hearing, Wiley, New York, 1938'

ssscrHroRv 29
SOUND LEVEL METTNS

$ound levsl meters contain micrcphones which transfsrrn sound pressura

variations in air into corresponding eleetric signals. The signals are then ampli-
fied internally and measured by electronic filters, which reject all signals outside
the selectsd frequency band. or by electronic weighting networks, with the rs-
sultg in decibels displayed on a digital readout or range indieator.
The weighting networks tend to represenr the freguency characteristics of
human hearing by rnodifuing the senshivity of the meter in pattorns similar to
specific equal loudness contours on page 29, The standard weighting net-
works are ide*tified as A to E, and the resultant decibel values are abbreviated
dBA, dBg, etc. Th6 primary difference between the standard weightings is that
the A weighting tends to considerably negleet low-frequency sound energy;
the B weighting rnoderately; and the C weighting hardly at all. $ome melers
also have a flat response, which has no frequency discrimination. The single-
number result for a flat response is celled the avaralldecibel level.
For detailed analysis of sound energy, sound level meters whh nanowband
"frequency analysis" filters should be used {e.9.. octave band for most
sources, one-third octavs band or narrower for more detailsd analyses). Fre-
quency analysis of sound {graphic plot of level vs. frequency } is extremely im*
portant because human respon$e to sound and noise control by absorption
{see Chap, 2} and isolatisn {see Chap. 4} are frequency dependent.
Although weighting nelworks do not provide direct information on the fre-
guency content of measured sound energy. they can be used to determine if
the sound €nsrgy is primarily above or below 1O0O Hz. For example, if the
dBC reading exceeds the dBA reading by 20 decibels or more. the rneasured
sound energy is primarily low frequency. lf the difference is 1o decibels or less,
the measured sound energy is primarily mid- and high frequency.

(, (to csnvert $a$nd

\ enerjg ini,o alectric signalr )
-Microphone

Oirite! read out ( to o isplauJ

soind levet in decibels)'

t-- CaEe (lo proNect'

kb \ tlzclrical

Va
iompanentr )

octrtvc - bend Filier t lo reiect,

oil eisnals
^it oiasida seiecled bdnd )

Reference

G. M. Hynes, "How To select a Low*Priced $ound Level Meter," s*und and vibntion,
May 1983.

3) eAscllGmY
A-WFIGHTEO $OUND LFVFL$

'A-weighted sound levels {in decibels, abbreviared dBA } largely ignore

low-frequency sound energy just as our ear$ do. Nevertheless, single*number
decibel values in dBA often cannot totally represent human perception of noise
and tha effects of noise on human camfort. For example, the graph of noise
levels for a residential whidpod tub shows two widely varying scund spectra
which have identical single-number dBA values {curves 1 and 2 shown
helow ) " However, the noices represented by the spectra would be perceived
by listeners as sounding vastly different: low-frequency ncise caused by ',gur-
gling" turbulenee from water jets and "humrning" pump noise {curve 1 and
},
high-frequency noise caused by "splashing"' impact from the water str*am
filling the tub (curve 2 ). single'number sound levele, therefore, should be used
with caution because human hearing perceplion dces not respond to sound in
a simple, decibeFaveraged manner. ln addition to sound level, an individual,s
sensitivity to sound varies with frequency content, duration, and psychological
factr:rs.

7 Curre i {ilhirlpeal ,araiar naLz,lag

1 ana qmp operat,n( )
//^r \-J\\-----at
/

cO

Crtrve 2 ( uihirloool
tub fillin6)
-€
f
6
tn
FreqrencS (Hz)

ln spite of these limitations, the dBA can be used to pr*dict community re-
spon$e to many kinds of environmental noise, including vehicular transportation
noise. Thgrefore, noise ordinances and regulalions use the dBA unit as a mea-
sure to specify limits on noise. Shown below is the frequency discrimination in
dB for A weighting. Note that ths curve connecting the plotted "weighting
d8s" is similar to the shape of equal loudness contours at low sound levels.

A Weighting {dB}
63 He 125 Hr 25O Hz 5OO H: 100O Hz 2000 Hz 4OOO Hz 8OO0 He

-15 - 15 *8 *3 +1 +1 *1

sastc rneonv 31
 An example computation for sound level in dBA from sound levels in oc-
tave bands at 125 to 4O0O Hz follows:

t25 '14 - l5
250 68- I
:. il:
:fi>='="
a'

6{-s \,,
500 = ?l dsh
t0CI0 65*0
ri:i
:j :>='-
2000 $?.+
:: l=='-
I

/+n00 6?+ |

$2 sasa rlicony
SOUNO SFECTRA FOR COMMON NOISES

The graph below shows sound $pectra (plotted on a standard grid typi-
cally used by acoustical consultants ) for three common hausehold appliances.
The vacuum cleaner noise is broadband because it consists of a complex mix-
lure of sound generally spread throughout a wide range of the audible spec-
trum. The mixture may include "whining" noise from rhe suction air intake,
"whistling" noise from the corrugated hose, and "screeching" impeller noise.
The electric shaver and hair dryer are complex sounds which have identical
single-number sound levals iin dBA ) . However, they would not sound the
sams to a listener because their sound energy and frequency characteristics
differ widely. The shaver noise is "whining" and the hair dryer "hushing." Re-
member the equal loudness cantours show that individuals srs mcst sensitive
to high-frequency sounds such as whining, hissing, or whistling noises.

l'liddle {requencA High f'eq,rgng3

a0
Hone vrcuurn
fl r,leaner ( at' 9Nt')
s
,70
sl
g
g
(3
€
ci 60
; L

*"rJ
9A
,6
> Eleclrrc shavor Eleclric nair druer
q'
(*t \/t lll (att+l) ,'
{t 6$ d,$Fr
t-
:rn I.n
{r
{}
L
(l*
*s
c
s,
lft 30
bt 125 2.50 500 i00a 2000 4000 4000
0c1,ave,- band ce"rigr lreq.r;0nc5 ( H l)

Note: Sound pr€s$ure level l, is rnore conveniently measured than sound intensity level
1,. For mosl architectural acouslics situations, they can be considered equivalent. The
reference value for lo in this book is 0.OOOO! N lm2 { newtons per $quare meter } , tho
minimum sound pressure at 10O0 Hz audiblc to young persons with normal hearing'
Decibel v&lues are meaningless without a reference valuel

elsrc;xeonv 33
NOISE LEVEL DATA

The table presents noiss data at octave*band center frequencies for fa-
miliar residential, autdoor, transporrarion, and building activity noise sources"
lntermittent or peak noises may exceed the data given in the table by 5 dB or

$ound Pre*$re Level {dB} dBA

Example 6ourcs 63 Hr 125 H2 ?30 lll 5OO lk 1O(n lb 2O{n flz 4O00 ll! SQOO llr
Home

Alarm clock at 4 to g ft {ringingl 46 48 55 62 63 1A 80 80

Flgctric shsver ar 'l % fi 5g 58 49 62 60 64 60 59 68
VaclBm cleaner at 3 ft 4e 66 69 73 79 73 73 72 81
Garbage dispsssl ai 2 fr s4 83 69 F6 55 50 50 49 8g
Clothes washer at 2 ta 3 ft {wa*h cycle} 59 65 59 59 58 54 S0 46 62
Torlet {refllling tank} RA 55 53 54 67 56 57 52 63
Whirlpool, six nozzles {filling rub} 68 e5 68 69 71 71 68 65 74
WindQw air-condkioning unit 64 84 65 56 53 48 44 37 59
Telephono ar 4 ro 13 ft {ringing} 41 44 56 68 73 69 83 83
TV at 10 fi 49 a2 64 6? 70 68 63 39 74
Stereo {teenager listening level} 60 72 83 8? 82 80 76 60 86
$rereo {aduh tstening level} 56 s6 75 72 1A 66 64 48 79
Violin at 5 ft {lonissimo} .. 91 St 87 83 79 66 92
Normal convercatronal speech at 3 lt 57 62 63 57 48 40 63

Outdoors

Birds at l0 ft 50 5? 54 57
Cicadas 35 5l s4 48 57
Larga dog at 50 ft (barkingl 50 58 68 7A oq s2 48 72
Lawn rnowgr at 5 ft 85 87 86 84 8'r 74 70 72 86
Rstol shot s1 250 ft {poak impulse ievelsl 83 9t g9 rc2 106 105
Suri ar 10 to 15 ft (rnod?rare sessl 71 72 7A 71 67 64 58 w78
Wind in lrees {10 mi/h} aa .F
37 37 35 43
Transprytarion

Large rrucks at 50 ft {55 milh} 83 85. 83 85 81 76 72 65 86

Passenger cars at 5O fi {55 milh} 7? 70 67 66 67 bb 59 54 71
Motorcycle at 5O ft {full ihronle, withour 95 95 91 91 91 a/ 87 8g 95
bafltel
$ngwmobrle at 50 ft 65 82 84 75 78 71 7g 69 85
Train at 1O0 ft {pu$ing hard} OK l02 94 g0 85 8? 83 7g 94
Train siren ar 50 tt 88 so 110 110 107 100 91 78 'los
q')
Carhomat 15ft
a1 tt
9S 90 w 60 s7
Corrfnercidl turbofan €irplans at 1 mile {from 82 a2 70 s6 79
takesff fbln parhl
Milirrry hclicopler a1500 ft {singl* engine, 92 89 83 8l 78 't2 82 5l 80
medium size)

lfiteriors

Amplified rock rnusic perfcrmance {large l't6 117 1r9 116 r18 115 109 lQa 121
arena l
AE
Audiovisual room 89 92 90 8S QA
so 94
Aud'torium (applausgl $0 68 75 7q 85 OA
76 65 88
Oassroom
Co.fre{ equipmenl room
60 no 72 77 74 68 60 50 78
Dog kernd
78 75 73 78 80 78 74 70 84
AfTna*m
90 t04 t0a 101 8S 73 108
72 78 84 89 86 80 7t 64S
Xrlct€n
Lfuasy
86
65
86 79 7A 72 65 57 81
i-frty
70 73 78 6s 65 61 77
lrb{tslcd sFjixr€nt roorn
60 83 6S 67 04 58 s0 4S s8
t^Jst Frtbe r6n
&7
s$
&s
34
85 84 83
g6
82 s0 78 s8
Racq.Elbd csrt o) 85
96 96
an
9t 91 90 100
Recefdr rd lobby rea
80 83 75 68 62 86
Tebcgde*rce
60
65
66
74
7Z 77 74 68 60 50 78
78 80 79 75 68 60 83

34 eAsrcrlGCFY
mors, dopending on the source or environrnent, Fsr many practical problems,
howsver, the data can be considered to be typical source levels at the given
distance and condition, or average general activity levels for interiors. The data
can be used fcr design purpose$ if proper consideration is given to especially
loud equipment or $ources which may exceed h, unusual site co*ditions, and
any other conditions that deviate from normal. For example, it is prudent to
measure transportation noise at propcsed building sites near highways, air'
pon$, etc., so design data will repre$ent existing noise sources and reflect
specific site features. Note also that many modern aircraft, trucks, and office
equipment are not as loud as examples in the table.

Note: Sources far noise level datainclude Journal of the Aroustba/ Socr'ery of ,{nerica,
Sound and Vibratian, Noise Contral Engineering Journal, and technical publications ol the
U.$. Environmental Protection Agency and National Bureau af Standards {U.S"} ,

/\.
BASIC THEffiY 35
DECIBEL SCALE$ FOR SOUND INTENSITY, FRESSURE, AND POWER

Sound pressure level L* may be considered equivalent to snlnd intensity

level [, in most architectural acoustics situations. These levels in decibel units
are affected by ttre environment in which the sound is measured. Sound power
level l*, gn ths other hand. is a more fundamental meagure of a source of
eound. lt expresses the amount of energy or powff that is radiated by a given
seurce, regardless of the space into which the source is placed. Thus lr. un-
like t and lo, is independent of the afou$tical characteristics of the space in
which the sound is heard. l* is somewhat analogous to the lumen raring of
lamps in lighting, and Lr and d to the footcandle level or illuminance in a room
{cf., M. 0. Egan, Concepts in ArchitecturalLighting, McGraw-Hill, New York,
1983) " The table below summarizes various characteristics of sound intcnsity,
sound pressure, and sound power levels. Refer to Appendix C f<rr a method to
convert l* to ln.

$ound Sound Sound

lntensity Level Pressure Level Power l,evel
Symbol Ll Le Lw
Express as 10 rog f 20 bsf 10 W
log wa

Unirs l, measured in dB L, measured in dB L* measured in dB

I measured in W/mr p measured rn N/m2 W measured in Watt
(or pascal, Pa)
fleference value* Is = 1Q-tz W/m' ft = 2x10-5 Nlmz Ws = 10-13 W
{1pW)
At reference value | * 0oB l, = OdB l* = OdB
Pain threshold value J= 10W/rnz P = 63 N/m2
At pain threshold value lr * 130 dB lo = 130 dB
'Usually taken as the lowest audible value for young persons with good hearing.

35 eAscrlGogY