Hard Data Regarding Concert Sound Levels
Click here to read about the data presented in the following article.
Talk about pushing a hot button! In the time since my first Ear Sense
column appeared in the November issue of Live Sound, I have received more
e-mail from readers than I get spam. Given the volume of spam being sent,
that’s an achievement. Clearly many of you are interested in this
topic, and more importantly, many of you are aware of the damage being
done and the problems we face. Congratulations to the informed!
Since it seems that hard, irrefutable evidence is necessary to create
understanding of exactly what is happening during a typical live show,
evidence is what we will look at in this installment and the next.
The unique data I am using in this series has been processed only to allow
it to be presented in a compact graphical format. No special filtering
or “post-processing” has been done to change anything. The
only filtering used was applied to each full-range dataset to allow us
to also show a narrowed spectrum from 315 Hz to 6,300 Hz, the area of
the ear’s maximum sensitivity, also the area in which the majority
of hearing related damage occurs and is most injurious. It’s the
bandwidth that most closely matches the A-weighting scale used (unfortunately)
in SPL measurement systems.
MEASURING LEVEL
The A-weighted scale parameters shown in Table 1 (roll-off rather substantially
as a matter of fact) filter the frequency response of the measurement
system to narrow the bandwidth to essentially a useable range of 300 Hz
to about 8 kHz.
Note the filtering applied as shown in the table and the summary chart,
and that the only place there is no adjustment is at 1 kHz. For reference
the “C” weighting filter is also shown. (Data compiled from
ANSI and IEC references.)
Table 1: Above, A-weighted scale parameters. Below,
summary chart. For reference, "C" weighting filer also shown.
(Courtesy J. Wolfe UNSW)
Table 2 shows what various regulatory and health agencies believe is a
safe level of exposure per 24-hour period. Within the audiology, medical
and health and safety agency communities, it is generally accepted that
Occupational Safety & Hazard Administration (OSHA) levels are too
high and that recommendations from the American National Standards Institute
(ANSI), National Institutes for Occupational Safety and Health (NIOSH)
and Environmental Protection Agency (EPA) are far more realistic and much
safer.
Table 2: Criteria from various regulatory and health agencies regarding safe level of exposure per 24-hour period. (Table courtesy www.nonoise.org)
Therefore, I am going to use the ANSI/NIOSH/EPA exposure limits as my
reference. (Highlighted in orange.) The time indicated is for exposure
without any hearing protection. I think it is fair to assume that the
majority of the audiences you serve are NOT wearing EPA, ANSI, NIOSH or
even OSHA approved hearing protection. NIOSH recommends that hearing loss
be defined as a significant threshold shift that is specified as increase
of 15 dB in the hearing threshold level (HTL) at 500 Hz, 1,000 Hz, 2,000
Hz, 3,000 Hz, 4,000 Hz or 6,000 Hz in either ear, as determined by two
consecutive audiometric tests.
Our brothers to the north (in Canada) have sort of taken an OSHAlike position
by saying that at least for industrial situations, Canadian standards
restrict noise levels in the workplace to 85 dB for an eight-hour daily
exposure. For each increase of 3 dB, the duration of the exposure to noise
must be reduced by half.
Most hearing conservation groups acidly note that the 85 dB(A) level of
sound for the exposure time of eight hours is one that industry and government
could agree upon (read negotiate to allow) for industrial noise exposure.
It should in no way be construed as a “safe” level for hearing,
although it is chronically misinterpreted as such.
It is crucial to recognize that all the data above is aimed at INDUSTRIAL
exposure to noise, not (yet) at exposure to performance audio technology
or systems. So far no one has made the hard formalized linkage, legally
speaking (read – costs you money), between industrial noise and
our industry – at least in North America. There have been scattered
instances of lawsuits being brought and compensation claims being made
but these are two random to call it a trend.
But realistically it’s only a matter of when – not if –
this is going to happen. Whether the first case is brought by a ticket
holder or an employee or whomever won’t matter much, because once
Pandora’s Box has been opened, the tsunami of litigation and regulation
will commence.
In conversations with several industrial audiologists and workers compensation
specialists both in the United States and Canada, all expressed surprise
that no one had as yet really tried to nail this issue and go for the
serious cash. One specialist pointed out that if someone whose job required
undamaged hearing brought a valid claim for hearing loss, the award could
be “quite substantial” (think seven figures) because the person
could be classified as “permanently disabled” from the perspective
of work-related injury laws and compensation authorities/agencies.
This type of award/judgment could be increased dramatically if the employer/facility/ticket
supplier (and etc.) did not provide information on potential hazards and/or
did not provide access to the appropriate protective equipment. In my
admittedly random survey of workplace injury professionals, almost all
universally stated that the hospitality and entertainment industries do
not offer (as a general practice) either warnings or protection, thus
making them potentially liable under the current laws for HUGE punitive
penalties.
Across the pond in the United Kingdom (UK), Noise at Work
Regulations have been in place for some time (originally implemented in
1989) and were recently updated. These rules apply to “employees”
at such places as nightclubs and discos, but so far, not the patrons of
such establishments. I will cover the UK viewpoint in detail in the next
installment.
So how loud are shows? Are levels really creating exposure that would
violate existing regulations? In a word, YES! Let’s go to the data.
With this article, we present three charts offering data collected at
three different live performances. The performances measured varied in
length, ranging from approximately 60 – 90 minutes, including encores
and intermissions.
   |
Figures 1 – 3: ‘80s rock concert.
Data collected at a concert by an ‘80s rock act. Levels were above
100 dB at least 50 percent of the time and above 95 dB for 90 percent
of the duration.
The first chart in each set is a full range composite of all the data,
acquired every five seconds through- out each show. The left-side scale
is in dB, from 70 – 110, while the bottom scale is in Hertz, from
20 Hz – 20 kHz. The right-side scale shows the percentage of time
during the whole show that levels exceeded the appropriate dB level. The
red band represents 5 percent of the time, yellow is 10 percent, green
is 50 percent, white is 90 percent and black is 99 percent.
The second chart in each set shows the same data, but measurements were
band limited to emphasize the range from 315 Hz – 6.3 kHz. The third
chart in each set depicts the same level data, but it is plotted as bar
charts without the spectral breakdown. The black bar represents linear
SPL data (un-weighted), the red bar is A-weighted SPL data.
Note that A-weighted data is what “sound police” at concerts
will measure with their SPL meters, because it supposedly provides the
best match to human hearing sensitivity. However there are serious questions
within the audiological community and among other scientists and measurement
professionals as to the accuracy of this assumption.
Above 90 dB SPL, the curve was based originally on measuring a level window
up to a reference level of 55 dB. B weighting (almost never seen on reasonably
priced SPL meters) was supposed to cover from 55 – 85 dB and C weighting
was to be used above 85 dB. Somewhere in the regulatory and bureaucratic
morass, the A scale was implemented, but I can’t seem to find out
precisely why or how. As one audiologist succinctly put it, “It
just was.”
For reference, the majority of current regulatory practices substantially
limits exposure above 85 dBA. Each chart has a reference line added at
90 dB to help you calibrate your mind to the levels being shown.
   |
Figures 4 – 6: ‘70s rock revival
act. Levels exceed 90 dB for 90 percent of the show and are at 95 dB for
at least 50 percent of the 65-minute performance.
Figures 1 – 3 show data collected at a concert by an ‘80s
rock act. Both the full spectral data and the filtered (to approximate
A weighting) clearly demonstrate that the audience was exposed to SPL
of 110 dB at about 160 Hz, as well as levels well above 100 dB broadband.
The bar chart and spectral data also show the most disturbing aspect of
this concert: levels were above 100 dB at least 50 percent of the time
and above 95 dB for 90 percent of the duration. Therefore, the audience
was exposed to hearing damage inducing levels for almost the entire 90
minutes of the event.
Figures 4 – 6 offer data collected from a ‘70s rock revival
main act performing at the same venue a few nights later. While the presentation
spectrum is a little different, this act, like the previous one, had levels
exceeding 90 dB for 90 percent of the show and at 95 dB for at least 50
percent of the 65-minute performance.
In both cases, note the big vocal range peak between approximately 800
Hz and 2 kHz, the huge low-frequency (LF) peaks and the high levels seen
even at 5 kHz and 7 kHz. This likely means the PA system was being pushed
pretty hard.
Now let’s examine Figures 7 – 9, with data measured at a heavy
metal main act. It’s obvious that the LF content was substantial
and that the peak has moved down to around 100 Hz, with tons of high SPL
up to around 2 kHz and then a steep roll-off above the vocals. Ready for
this? The audience was exposed to levels at or above 95 dB an almost unbelievable
99 percent of the time! This is a rate considered harmful for 50 percent
more time than recommended by EPA or NIOSH. (And I’ll bet many left
with their ears ringing – any takers?)
Figures 7 – 9: heavy metal band. Levels at or
above 95 dB an almost unbelievable 99 percent of the time!
Finally, Figure 10 offers a plot showing the comparison SPL levels measured
for all three shows. Look at the 90 dB reference line and see how much
of the spectrum was well above it, as well as how much was above 100 dB!
Figure 10: Comparison SPL levels for all three shows.
Look at the 90dB reference line and see how much of the spectrum was well
above it. (click on image for full view)
I believe this data is highly representative of current concert levels
practice and shows that we are producing far too much level for way too
long. If the regulatory folks start to produce data like this –
look out! It would be interesting to have the THD figures from these shows
to see how much distortion was present on top of the very high SPLs.
Remember, these measurements were entirely taken at the house mix position,
not in the front row. If these audience members went to just a few shows
during the summer when the measurements were done, they most likely have
some measurable hearing damage, using the NIOSH criteria of threshold
shift. Multiply that exposure by a few years, and you would see audiometric
charts that look like those from construction workers using jackhammers
or metal fabrication plant employees.
I’m not sure I want to look at the hearing tests for the crews on
these shows - it likely would be too disturbing. Maybe you were one of
the crews where our measurements shown here were taken – have you
had your hearing checked? If so, you too will likely be disturbed by what
you see.
Next installment: More data from more shows, and the UK situation.
Fred Ampel has been involved in the A/V industry for 33-plus years.
His career has included work in sound reinforcement, broadcast, studio,
A/V system design, installations and equipment development. He was founding
editor/editorial director of S & VC. Ampel heads up Technology Visions
and can be reached at www.technologyvisions.com.
Figure 10. Comparison SPL levels for all three shows. Look at the 90 dB
reference line and see how much of the spectrum was well above it.