Noise in Industry - Fundamentals and Assessment Tools for Occupational Ergonomics - page 612

is at 45 dBA may currently be comfortable for communication. However, if a new ventilation system
increases the noise level in the room by 10 dBA, the occupants will experience a doubling of loudness
and will likely complain about the effects of the background noise on conversation in the room. Once
again, the compression effect of the decibel scale yields a measure that does not reflect the much
larger influence that an increase in sound level will have on the human perception of loudness. Although
the sone scale is not widely used (one exception is that household ventilation fans typically have
voluntary sone ratings), it is a very useful scale for comparing different sounds as to their perceived
loudness.
It should be evident that phon levels can be calculated directly from psychological measurements in
sones, but not from physical measurements of SPL in decibels. This is because the phon-based loudness
and SPL relationship changes as a function of the sound frequency and the magnitude of this change
depends on the intensity of the sound.

31.2.3.3 Modifications of the Sone
A modification of the sone scale (Mark VI and subsequently, Mark VII sones) was proposed by Stevens 2
to account for the fact that most real sounds are more complex than pure tones. Utilizing general
formula (31.6), Steven's method utilizes octave band, 1 3 octave band noise

measurements, and adds to the sone value of the most intense frequency band a fractional portion

/ 2 octave band, or 1 /

of the sum of the sone values of the other bands ( P S). S m is the maximum sone value in any
band, and k is a fractional multiplier that varies with bandwidth (octave, k 0.3; 1 2 octave, k 0.2;

¼ / ¼

1 3 octave, k 0.15):

/ ¼

Loudness in sones ¼ S m þ k SS ð S m Þ (31 : 6)

31.2.3.4 Zwicker's Method
The concept of the critical band for loudness formed the basis for Zwicker's method of loudness quanti-
fication. 3 The critical band is the frequency band within which the loudness of a band of continuously
distributed sound of equal SPL is independent of the width of the band. The critical bands widen as fre-
quency increases. A graphical method is used for computing the loudness of a complex sound based on
critical band results obtained and graphed by Zwicker. The noise spectrum is plotted and lines are drawn
to depict the spread of masking effect (defined later in this chapter). The result is a bounded area on the
graph that is proportional to total loudness. The method is relatively complex and the reader may wish to
consult Zwicker 3 for computational details.

31.2.3.5 Noisiness Units
Loudness and noisiness are related but not synonymous. Noisiness can be defined as the “subjective
unwantedness” of a sound. Perceived noisiness may be influenced by a sound's loudness, tonality, dur-
ation, impulsiveness, and variability. 4 Whereas a low level of loudness might be perceived as enjoyable or
pleasing, a low level of unwantedness, that is, noisiness, is by definition undesirable. Equal noisiness con-
tours, analogous to equal loudness contours, have been developed based on a unit (analogous to the
phon) called the perceived noise level (PN dB ), which is the SPL in decibels of a 1 3 octave band of

random noise centered at 1000 Hz, which sounds equally noisy to the sound in question. Also, an N-
(later D-) sound level meter weighting curve was developed for measuring the perceived noise level of
a sound. A subjective noisiness unit analogous to the sone, the noy, is used for comparing sounds as
to their relative noisiness. One noy is equal to 40 PN dB , two noys are twice as noisy as one, five noys
are five times as noisy, and so on. Similar to the behavior of sone values for loudness, an increase of
about 10 PN dB is equivalent to a doubling of the perceived noisiness of a sound.

/