These deﬁnitions are intended simply
as an aid to understanding and for this reason, some
are not strictly scientiﬁcally accurate. For correct
formal deﬁnitions, consult IEC standards 60651, 60804, 60942, 61252 and IEC 61672:
2003 part 1.
Sound: Sound is simply Pressure. The units are Pascal (symbol
Pa). Look at your Hi-Fi
speakers at home,
the cones vibrate and
these vibrations are sending out
waves of pressure
(imagine them as being similar
to the ripples
on a pond, but in 3-D).
These waves spread
out in all directions from
the original source;
getting weaker the further they go.
picks up these
vibrations and converts these pressure waves
into electrical impulses that are carried
to the brain to be "interpreted" as sounds.
The Decibel: (symbol dB). This is a logarithmic scale of pressure
produced to allow
a better correlation between what we hear
and the scales
we use. Being
logarithmic, you can't
add decibel values
you need to anti-log
them to get actual
pressure in Pascal,
add these values
together and then
log the result.
Needless to say
this is time consuming when done by hand. To give
some scale, '0' dB is a pressure
of 20 micro Pascal, while
94dB is 1 Pascal and 140dB is 200 Pascal.
computers, a scale
from 20 mPa to 200 Pa (a ratio of 10,000,000:1) was difﬁcult to read, but 0 to 140dB was simple to understand and display.
Frequency Weighting: Usually A or C &
now Z. The ear responds to frequencies in a non-linear way; some tones
being more easily perceived than
others. Health, some
drugs and pressure levels can also
change the ear's
frequency response. Because of
this, ﬁlters are applied to sound levels
and these modify
the frequency response
and the common
'A-frequency-weighting’ is intended to approximate to the way your ears hear sounds.
The symbol for the A weighted decibel
is dB(A). It should be noted that A weighting does
not show a 1 to 1 relationship to perceived loudness or even the full risk
of damaging human
hearing, but BY LAW, we must use it and
in IEC 61672
it is mandated to be ﬁtted to every sound
'C' weighting is mainly used
for lower frequency sounds usually in connection with
(see below). It is ‘level’
from 31, 5Hz to 8kHz
'Z' or ZERO
weighting is introduced in IEC 61672
and is used
instead of the old LIN or FLAT that
varied from manufacturer to manufacturer. For a Class 1 instruments it is ’level’
from 16Hz to 16kHz, in other words,
f/2 and 2f times the ‘C’ limits..
'B' and 'D' weightings have been 'retired'. ‘B’ was a ‘half-way’ house
between ‘A’ and ‘C’ and ‘D’ was just
for pure ‘non-bypass’ jet engines and these are only military
today now Concorde has been axed.
Time Weighting: F, S and I. When sound
level meters were
ﬁrst developed, 'dial
and pointer' needles
were used to display the level measured. These instruments had a response time depending somewhat
on the inertia
of the needle and were
categorised initially as 'Fast' and
'Slow' responses. Subsequently 'Impulse' response was
added, but in the 1960's,
they were re-named 'F',
'S' and 'I'
to be the same in every language. ‘F’ and ‘S’
are still used
today, to get some
idea of the
speed of change of noise
levels. 'I' time-weighting is not really
used today and indeed the new standard
IEC 61672 does not include
of the standard, but many SLMs
are still capable
of reading data
with this time-weighting, as it is still in German statute law.
SPL: Sound Pressure
Level. This is the pressure
after it has been "logged" into decibels. This is a rapidly changing
ﬁgure that only tells
us what the dB is now, and always
has a time weighting applied.
This does not give us much useful
information. Formally, SPL is the 20 times
the log of the root
mean square of the pressure.
Level: (common abbreviation Leq) When SPL is measured
and recorded a graph may be produced, & Leq can be described as the area
under the curves
expressed as a average line.
If noise levels
are rapidly changing
then Leq can be a useful
tool and for this reason
it is also called the "Equivalent Continuous Level". It is properly known
as time-average sound level
(symbol LAeq,T) but mainly only pedants use this terminology.
Sound Exposure: (symbol E). Formally, this
is the time integral of the squared
A-frequency-weighted sound pressure of a noise event and
is expressed in pressure-squared-time, i.e.
Pa²s or Pa²hr. Sound
Exposure is in fact the most useful
metric available, as being
’real’ pressure based,
two Exposure values
can simply be added, unlike
two decibel levels.
It is not yet universally used as traditional acousticians, being brought
up on decibels do not like change.
To be fair, they also have some practical reasons
as well as just 'tradition'. Sound Exposure being
Pa² time is proportional to the energy
of an event.
2dB + 2dB = 5dB,
using pressure in Pascal 2Pa
+ 2Pa = 4Pa.
Level: (symbol LAE, abbreviation SEL). This is simply the sound exposure
expressed as a logarithm. Effectively Leq is SEL divided by time. Again
like SE, SEL is an amount of noise. Thus
while Leq, the average level
can go up and down, SEL
can only go up, you cannot get
'less sound' once
it has happened. The habit has
grown up of calling SEL the 1 second equivalent Leq.
This is a misunderstanding of reality,
although these have
the same numerical value. SEL is invaluable for measuring the noise of a single
event and indeed
some people even think SEL
stands for 'Single
Event Level'. SE, SEL, Leq and noise dose
are all simply
mathematically related and if time
is known, they
can all be calculated very
most modern sound level
meters only measure
SE; they calculate everything else internally.
Level: (symbol Lmax).
This is quite simply the maximum time and frequency
weighted sound level ie LASmax, the maximum level using A frequency-weighting and S time-weighting.
Level: (symbol Lpeak)
This is not
the same as Maximum Sound
Level. Peak Sound
Level records the peak of the original pressure
wave, not the ‘rectiﬁed’ result.
This is commonly
associated with C or Z frequency weighting, but has no time weighting. If the noise
being measured is impulsive such
as a hammer being used,
then the Peak level may easily be 20dB higher than
the maximum sound
level, this is due to the time
weighting being applied
to the max
sound level. To give
some scale. The time constant for 'I' (Impulse) response is 35millisec. The maximum permitted acquisition time of Peak is 100