Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Jeff Carter]

Somewhat more seriously...

Depending on dynamics, "OSHA-safe" maximum audio levels for a 2-3 hour show (assuming the audience has no other noise exposure) would be at least 100 dBA.

You'd really have to wring out a typical church PA system to get those sorts of levels, and as the volume approaches absolute max the intelligibility and mix quality tend to suffer. As long as the mix quality is decent, I suspect that SPL levels are likely to be within acceptable limits (and as dick rees said above, if the mix is bad, it should be turned down anyway).

Edited to add:
Standards in Europe are far stricter than they are in North America, with mandatory hearing protection above an 8-hour average exposure level of 85 dBA (instead of 90 dBA), and allowable exposure time cut in half for every 3 dB increase (instead of 5 dB).

For a two-hour rock show, this reduces the maximum "OSHA-safe" weighted average SPL from 100 dBA (North America) to 91 dBA in Europe. When in Rome, hand out the earplugs...

[Will Jarrell]

Kevin,

Rider issues notwithstanding.. I think if you were to ask guest engineers to keep the level between 95 and 100 dBA on the high end you will be fine, assuming they are professional. 

Will Jarrell

[Brad Weber]

Rider issues notwithstanding.. I think if you were to ask guest engineers to keep the level between 95 and 100 dBA on the high end you will be fine, assuming they are professional.

Of course A-weighting is technically completely inappropriate for such high SPL levels since the A-weighting curve reflects the human hearing response at levels closer to 40dB(SPL) and at 90 and 100dB(SPL) levels grossly undercompensates for the low and very high frequency contribution.  However, it has become widely accepted and in some cases for the very fact that it allows increased low frequency content.
 
Basically, if you are runing at higher SPL levels and see an increase in the SPL level reported when you switch between A and C weighting on a meter, then you are likely taking advantage of a now widely accepted misapplication of A-weighting.
 
The flip side is that if you have someone with a XdB(SPL) requirement with no specified weighting, then making C-weighted or even better Z-weighted (Zero/flat/linear/unweighted) measurements may result in higher readings and thus being able to meet the specified SPL level at a lower loudness level.
 
 
One of the factors that hasn't really been mentioned is crest factor or dynamics.  Very dynamic music, that with a high crest factor, may have peak levels that exceed the average levels by 20dB or greater.  In comparison, heavily compressed music may have much more limited dynamics and a much smaller crest factor, in some cases less than 6dB.  This is relevant as a) the same peak level for both can relate to quite different average levels and b) the same average levels for both can related to quite different peak levels.  This is one reason why a single number peak or average level criteria may not work well if the content varies significantly.
 
For example, if you apply a single peak level criteria to both contemporary music performances and orchestral performances then due to the dynamics involved the average level of the contemporary music may be up to 20dB or so above the average level of the orchestral music.  So even though they have the same peak levels and while some complaints regarding the contemporary music being 'too loud' may be a subjective reaction to the content, other comp[laints may be a reaction to the greater average level.
 
What it really comes down to is that perceived loudness has multiple factors including the not just the amplitude of the sound pressure level but also the sound's frequency content and dynamics.  Single number dB(SPL) criteria are very limited in how they can account for such factors.

[Jeff Carter]

Of course A-weighting is technically completely inappropriate for such high SPL levels since the A-weighting curve reflects the human hearing response at levels closer to 40dB(SPL) and at 90 and 100dB(SPL) levels grossly undercompensates for the low and very high frequency contribution.  However, it has become widely accepted and in some cases for the very fact that it allows increased low frequency content.

...
 
What it really comes down to is that perceived loudness has multiple factors including the not just the amplitude of the sound pressure level but also the sound's frequency content and dynamics.  Single number dB(SPL) criteria are very limited in how they can account for such factors.

Good stuff, Brad, thanks for posting this.

However... any noise exposure regulations I've seen use A-weighting.

So while I agree with you that A-weighting is far different from human hearing response at high SPL, the use of A-weighting in noise exposure standards would seem to indicate that A-weighting must be a reasonable model as far as noise-related hearing damage is concerned.

If you've seen anything that indicates otherwise, I'd love to know about it.

[Arnold B. Krueger]

Good stuff, Brad, thanks for posting this.

However... any noise exposure regulations I've seen use A-weighting.

So while I agree with you that A-weighting is far different from human hearing response at high SPL, the use of A-weighting in noise exposure standards would seem to indicate that A-weighting must be a reasonable model as far as noise-related hearing damage is concerned.

If you've seen anything that indicates otherwise, I'd love to know about it.

Yes, the A-weighting model correlates with the hearing-damage model. Audibility has nothing to do with it.

Your ears can enjoy and sustain without damage a ton more dB at 30 Hz than 3 KHz.  At 10 Hz the difference is even more extreme.

[Ivan Beaver]

Yes, the A-weighting model correlates with the hearing-damage model. Audibility has nothing to do with it.

Your ears can enjoy and sustain without damage a ton more dB at 30 Hz than 3 KHz.  At 10 Hz the difference is even more extreme.

A weighting is intended for relitively quite sounds-like below 80dBSPL or so.

But that does not keep people from putting a meter on the A scale and "measuring something-no matter how wrong that measurement is.

But they get a reading-so "therefore" it HAS to be true .

It has been said by the experts that above the 80dB point (or somewhere around there), C weighting should be used-as it more closely relates to hearing damage.

But then there are different kinds of "experts" that have different opinions.

[Jeff Carter]

A weighting is intended for relitively quite sounds-like below 80dBSPL or so.

But that does not keep people from putting a meter on the A scale and "measuring something-no matter how wrong that measurement is.

But they get a reading-so "therefore" it HAS to be true .

It has been said by the experts that above the 80dB point (or somewhere around there), C weighting should be used-as it more closely relates to hearing damage.

But then there are different kinds of "experts" that have different opinions.

I'm not up to speed on research in that area so I'll take your word for it. However, if that's true, can you explain to me why every workplace noise exposure standard I've ever seen uses A-weighting rather than C-weighting?

[George Friedman-Jimenez]

... It has been said by the experts that above the 80dB point (or somewhere around there), C weighting should be used-as it more closely relates to hearing damage....

Ivan, do you know of any evidence that supports the assertion that dBC is more closely related to hearing damage than dBA?

[Brad Weber]

I'm not up to speed on research in that area so I'll take your word for it. However, if that's true, can you explain to me why every workplace noise exposure standard I've ever seen uses A-weighting rather than C-weighting?

Cost and simplicity.
 
Understand that A-weighting is intended to reflect human response at I believe a loudness of 40 phons, which is 40dB at 1kHz.  And I think C-weighting represents the 100 phon loudness response.  But those are just the responses at two loudness values, the response does not suddenly make a big jump for the 40 phon response to the 100 phon response, it varies continually between and to either side of them.  That matters because if you wanted to accurately track the effects of human response changing with loudness it would have to be a continually varying weighting, which would be very complex to address.  So any single weighting, be it A, C or Z weighting, will really only be accurate at one loudness level and not representative at others.
 
Also consider that most of the criteria you're addressing was developed during the days of analog meters and is still often measured with analog meters.  High precision, analog metering with multiple weightings was expensive and it was pretty common for meters to have just one weighting.  In fact, I believe that the original criteria were octave band based but while that was fine for acousticians and audiologists, the instrumentation was expensive and the data difficult to interpret for others so to simplify things it was revised to a single number with some compensation for frequency response.
 
Adding to this, while A-weighting significantly under reports the low frequency contribution, it actually maintains or even adds to the levels reported for the frequencies at which the ear is most sensitive and that are most critical speech intelligibility.  While not hearing the bass or high hat is definitely a loss and to be avoided, it is realistically probably not the potential negative or risk that losing the ability to hear a siren or someone yelling at you to get out of the way might be.
 
As Arnold noted, studies have apparently shown that of the common response weightings, A-weighting seems to most closely relate to noise induced hearing loss.  But also remember that that most criteria is addressing relatively continuous noise over longer exposure periods.
 
Finally, keep in mind that you are talking about generally government set criteria with often competing considerations and finding asomething everyone will accept may trump being technically accurate.
 
It should also be noted that while it was historically accepted that low frequency hearing loss was less of an issue, my understanding is that with often increasing exposure to higher low frequency noise levels that perspective is evolving.  And this is probably paticularly applicable to audio and music reproduction, a church audio system of 40 or 50 years ago may have produced close to the same dBA levels as a modern system but chances are the low frequency output levels are much greater in most contemporary systems.

[Marty McCann]

Generally, people do not lose Cillia (hair cell receptors) that respond to low frequencies, so OSHA is not concerned about the content at the lowest of frequencies.

Try this with a three or even four-way system . . . . while listening on axis . . . . turn it up until you think that it is damn loud, but does not hurt . . . . now mute all outputs except for the high's . . . .    

The lows were actually causing compression in your middle ear bones that was attenuating your perception of how loud the highs actually were.