A-Weighted for Avoiding Hearing Damage?

[hduncan]

I have actually had sort of a hard time figuring this out, but I finally found out that OSHA set the level of 90dBA (A for A Weighted, I believe) as the max dB level for someone to listen to for 8hours without hearing damage.

I think it actually has been dropped to 80-85dBA but, thats beside the point.

Should I be setting my SPL Meter to A-Weight when measuring the loudness of a source to protect my ears?

When set to C weight, the measured level in my car jumps 20dB!

[steveschizoid]

No, because I would always err on the the side of caution. A-weighting, in my understanding, devalues low frequencies as well as higher frequencies - 5 KHz and above in accordance with hearing sensitivity at lower levels. However, as the SPL rises, the equal loudness contour/Fletcher Munson curve flattens. According to the below linked article, A-weighting becomes invalid above 55 dB. The idea to measure the OSHA standards using A-weighting probably came from industry lobbyists.

A friend of mine who is an audiologist says that he has seen many, many audio related clients who have something audiologists refer to as "boilermaker's ear" - basically normal hearing with the exception of a 5 dB drop in the 4-7 KHz range.

Here is a good article:
WATCH YOUR SOUND DOSAGE

[MarkRB]

Originally Posted by steveschizoid

basically normal hearing with the exception of a 5 dB drop in the 4-7 KHz range.

That'll be the guys who like to mix on NS-10's then

[ScottTunes]

Originally Posted by steveschizoid

"boilermaker's ear" - basically normal hearing with the exception of a 5 dB drop in the 4-7 KHz range.

Originally Posted by MarkRB

That'll be the guys who like to mix on NS-10's then

LOL!

[djejnyc]

Thanks for the article link, I'm interested in this subject, having just past my 20 years (!) working in the studio....

[hduncan]

So.....

I should use C weighting?

Damn, that means everything needs to be MEGA quiet..

Really?!?! Damn I was so excited today when I found out it was A-weighted.

Can anyone else confirm this?

[R3k]

Damn Yamaha ;-)

QUOTE from the bottom of WATCH YOUR SOUND DOSAGE="Subsonic filters should be inserted in the instrument outputs of certain instruments. For example, Yamaha instruments are known to output DC transients."

[Johnny the Fox]

The short answer is: Yes, A-weighted.

The A in dB(A) means it's A-weighted. C-weighted would be dB(C).

The level limit you refer to (80-85dB(A)) is a daily dose if you work for 8 hours. To measure your dose you would have to measure the mean A-weighted sound pressure level LAeq at your hearing position for 8 hours.

You can go higher than 85 if you have longer breaks with lower levels. For instance 2 hours with 90dB(A) and 6 hours with 80dB(A) will lead to a mean 8 hours-level of 85dB(A). If you work longer than 8 hours the levels have to be lower.

If your sound pressure meter doesn't show LAeq you can use the SLOW time weighting for orientation.

Please be aware that the purpose of these limits is to prevent the need for hearing aid before you are 65 years old and not to maintain your ability to mix hit records. It's a good idea to give the ears longer breaks of 70 dB(A) on a regular base after working with the big speakers on.

Best regards

[hduncan]

The website provided seems to have a lot of info, but I dont know if I want to just accept it.

Then again, better safe than sorry...




Does perception have anything to do with this though?

It seems to me that the damage to your ears doesnt conform to the fletcher-munson curve.... but instead, that a certain amplitude of a certain frequency will kill the hair cells that detect it, in which case, why would that be different from any one cell to another, or by how loud it is perceived (which is entirely mental, because the hairs dont have exponential reaction, they have a fixed scale reaction...)

Am I making any sense here. It seems like nobody really knows what's going on, hahaha..

[Johnny the Fox]

As Andy wrote, the area around 4kHz often suffers first. Low frequencies < 500 Hz at typical studio levels and most industrial work places are less critical. That's why of the standard frequency weightings (A-C) the A-weighting correlates best and is used world wide.

Best regards

[Flying_Dutchman]

Fletcher Munson curves represent curves of equal loudness.
If you hear a 1khz sine wave at 90dbspl, the phone value is 90 phon.
Ok, if you want to hear a 3,2khz tone as loud as the 1khz tone, 90 phon, the curve indicates a value of around 82dbspl. That just means that a 3,2 khz sine wave at 82dbspl will be as loud as a 1khz tone at 90dbspl, Off course, this is an average, many humans are measured. The reason why you hear a 3,2 khz sinus at lower dbspl levels is that the human ear is a quater wavelenght resonator. You will hear frequencies that fit with a quater wavelength into the "tunnel" of your ear better.

Ok, whats going on with a,b,c,d weightings. The weightings are applied to measure a spl with characteristics of a human ear/brain. The 80-85 mixing area should use the c-weighting by definition. It´s used for 80-90 phon. But often the a-curve is used. But to me it doesn´t make sense, because it´s just a different low-end roll-off (expect of d). Problematic to the human ear are high frequencies. The human ear got functions to prevent high spl pressure. Eardrum is less tightend and the mechanics of the ear get softer. But this natural protection circuit got a reaction time.
A deep frequency will be easier recognized because it is slower than the high frequency that will kill your little hairs inside the ear easier, because its too fast to be protected.
So if we measure a low filtered full frequency range signal by "a" filter, the meter will show less energy, but does it mean that the critical frequencies got less power, too? No...

[Johnny the Fox]

Flyin Dutchman, you are right,

but:

90dB @ 4kHz = 91dB(A) and 89dB(C)

90dB @ 250Hz = 81dB(A) and 90dB(C)

4kHz @ 90dB is more dangerous than 250Hz @ 90dB

The A-weighted level is much higher at 4kHz than at 250Hz.

That's why the A-weighted level gives you more information about the potential damage and is used for this purpose.

Best regards

[steveschizoid]

Originally Posted by Johnny the Fox

Flyin Dutchman, you are right,

but:

90dB @ 4kHz = 91dB(A) and 89dB(C)

90dB @ 250Hz = 81dB(A) and 90dB(C)

4kHz @ 90dB is more dangerous than 250Hz @ 90dB

The A-weighted level is much higher at 4kHz than at 250Hz.

That's why the A-weighted level gives you more information about the potential damage and is used for this purpose.

Best regards

I didn't know where you found these number, so I tested myself with a sine wav and an spl meter. I was surprised, my results were similar! However, since real world noise is usually closer to pink noise than a sine wav, I still think C weighting may still be safer. For example, I measured a song from Death Magnetic at 90 dB C-weighted, but it only measured 87 dB A-weighted. Then I roughly filtered out everything above and below 4 KHz and the result measured 69 dB. I know that 85 dB is a common benchmark for mixing and mastering; does anyone remember what Bob Katz had to say about this? And isn't there also the added factor that the more compressed your mix is the louder your ears experience it, even though the db meter will register the same level?

[Johnny the Fox]

Andy, if you knew me, you wouldn't be surprised .

You are right, for music C-weighted is on the save side. But with a lot of bass you sometimes can end up at quite low loudness levels.

Now, Bob Katz's suggestion is a different story. If I remember it right his SPL values are in dB(C). So in that case you have to use the C-weighting for your measurements. I do it like this myself.

But these levels are mainly chosen in order to get a compatible frequency perception and better mixes, not to avoid hearing damage (although they will help for that as well).

Best regards

[Flying_Dutchman]

Originally Posted by Johnny the Fox

Flyin Dutchman, you are right,

but:

90dB @ 4kHz = 91dB(A) and 89dB(C)

90dB @ 250Hz = 81dB(A) and 90dB(C)

4kHz @ 90dB is more dangerous than 250Hz @ 90dB

The A-weighted level is much higher at 4kHz than at 250Hz.

That's why the A-weighted level gives you more information about the potential damage and is used for this purpose.

Best regards

Right,

ok, i got a false in thinking.
Many dbspl values appear without beeing related to frequency, but they are.
It refers to 1khz, they don´t take an average....
If they would take an average, the weightening would lower the dpspl, because bass energy is filterd out. The average dbspl is lowered.This could lead to a higher energy level in the problematic frequency area.
But they don´t take an average... it´s at 1khz in the manuals, or should be.
At some manuals there is no refernce printed, but it shall always be measured in 1 meter at 1 khz, right?
Anyway, a,b,c,d measurement is dbspl over frequency and there we got a fletcher muson kind of implemented behaviour.

[hduncan]

Ok, you guys went way over my head here. Can you tell me if I should be measuring via C-weight or A-weight to prevent hearing damage?

It is my experience that C-weight shows a higher dB measurement than A-weight for the same apparent loudness of source material. I assume this is because Cweight takes in more sound/bass whereas the A rolls off the bass and only accounts for the more sensitive parts of the hearing..

That said, If I attempt to not listen to anything over 90dBC, I will definitely be saving my hearing for sure, because if its music, there is bass in there that takes away how loud the mids can be. Right? There is less headroom for the mids if the bass is in there, thus the mids arent making up 100% of that 90dBC...

But dBA disregards the bass because it takes a lot more bass volume to hurt the ears than the HF counterparts...


So the fletcher munson curve really does show you how susceptible the ear is, not only is it a graph of loudness, but a graph of vulnerability?

AKA, the dip betweek 1k and 10k shows that those frequencies are easier to damage?

[Arksun]

I just got an spl meter myself today to get a rough idea of how loud is too loud in my studio and was surprised at the difference in A and C db levels until I saw the frequency plot difference.

I'm also a little confused as to which to go for as reference point, perhaps taking both db measurements and using the average?. Going with C would be the safest bet as thats always the higher value listening to music.

According to my meter (which is supposedly +/- 1.5db accuracy), set to show max value when I'm generally listening to music its around 70db(A) / 80db(c) and when its piece of music I'm really getting into its around 74db(A) / 84.5 db(C). And that didn't seem too loud, eek!. I know when I've worked on stuff I've played it fair bit higher. Probably a good 90-95db(C) , but that would still put it within 85db when A-weighted. Hmm...

Time to get that time machine and tell my younger self to a) Start wearing earplugs earlier when clubbing, b) avoid that time I went to the paddocks @ Formula 3000 race and subjected them to ear splitting screaming engine noise at very close range and c) Avoid the time I completely overloaded my monitors and blasted a good 125db+ for about 10 seconds

[johndykstra]

Originally Posted by hduncan

AKA, the dip betweek 1k and 10k shows that those frequencies are easier to damage?

Doesn't really work that way. Noise induced hearing loss will generally present itself in the 4-6k range regardless of the frequency of the noise itself.

The interesting thing about the inner ear, is that there are only hypothesis' regarding how exactly the nerve hairs "hear"... being embedded in the skull of a live human, only so much can be done to find the truth without damaging the hearing, so we are really left with guesses.

The most popular and widely excepted theory is that there are sections of hairs that are responsible for triggering particular frequencies. Shorter, finer hairs devoted to treble... down to longer ones dedicated to detecting bass. All of these hairs live in and share the same space... which is filled with liquid. When the bones of the middle ear vibrate against the wall of the inner ear, this liquid is sent into motion. The hairs, moving much like seaweed through waves in the ocean detect the vibration and send information to the brain. While every hair "feels" every vibration, they only trigger an electrical impulse when their frequency is present. A large loud wave of sound will send the inner ear liquid into a tidal wave... so to speak. The hairs responsible for bass detection are more robust, but the poor little high freq hairs don't stand a chance. The can be ripped out completely, knocked down against the side of the inner ear, and often times, become tangled up with one another (tinnitus). What is still baffling, is why the hearing loss generally presents around 4-6k. A common theory is that it is due to their location in the "snail shell" curves of the cochlea... sort of a high risk zone. It seems higher freq hairs are further in, and thus better protected from the intensity waves.

Generally speaking, age related hearing loss starts at the highest frequencies, and gradually works it's way down the range. As we age, the liquid of the inner ear begins to thicken, making the detection of higher (smaller) vibrations more difficult. Image vibrating two bowls of liquid. One is water, the other is pea soup (this is an exaggeration clearly). Both of these mediums could transduce a 20hz vibration much the same. Now a 20khz vibration on the other hand...