Adventures in high-end decibel metering

[Tomg]

So just a quick and dirty update on the decibel/sound metering using Larson & Davis LXT (high end impulse db meter, and yes I can buy 2 Royale's and a scope for what it costs)
I thought maybe some of you astute readers would find some of the info useful.

After much learning from the engineers over at Larson's engineering facility. This is some of the things we've found that applies to us.
So decibels operates on the square root of the distance to the sound source. Meaning if you double the distance you half the SPL readings, or 6db lower, which on the decibel scale is half the SPL. The opposite is also true, halving the distance to the sound source doubles the SPL (6db)

Since our Mic is measuring from 80db to 190db, the noise floor is such that the impulse metering can not be triggered below 104db.
Mil-spec measuring of DB levels on guns is done 1m away from the mic, in our case, most moderated airguns will not trip the 104db threshold at that distance. But the easy remedy, and most practical is to use out knowledge from above. Halving the distance is doubling the sound pressure (SPL) or 6db. Which will trip the minimum level, and therefor give us a good reading. So now you can do your readings at that distance, deduct 6db's from all of them individual, and you get back to the Mil-Spec levels.

For instance a stock FX Royale .25, on high power, shooting 25gr pellets, measured 112db at 0.5m, so to compare that to any other gun decibel metering, you deduct 6db. So apples to apples is about 106db.
Now keep in mind this is peak impulse SPL, not sustained or timed SPL which is how 100% of all decibel meters you can get your hands on without spending a wad of cash, works.
The reason you are interested in peak impulse SPL, is that it is what causes hearing damage, impulse levels of 130/140db causes instant hearing damage.
For the causal reader, a difference between 106 and 120 doesn't seem a lot, however if you the math 10^(14/20)=500% louder (every 6db increase is a doubling of the total SPL)

[Tomg]

I should mention #1. sustained db over 120 will still cause hearing damage as a result of time exposure.
But unlike impulse levels, sustained noise is much more in your face and tells you it's loud.
#2. Dampening the sound in a gun has a practical "stop" on the lower end where it's own systemic sounds cannot be lowered anymore, like hammer strike, trigger, cocking bolts, etc etc. For instance a break barrel can only be silenced until you reach the level of the piston/spring noise, after that you have to figure out a way to reduce that noise to be able to lower the total noise.

[terry mc]

very interesting Tom.
So does the mil spec define the axis of measurement relative to muzzle direction?

I have found in my limited measurements that shallower angles tended to result in
higher measurements than say 90 degree angle.
This is of course with low end instruments, of questionable calibrations.
So results are considered for delta reference only.

[Tomg]

Terry,
Correct, the plane to which you measure is perpindicular to the distance from the mic. I.e. the muzzle is in line with axis forming the 1m distance to mic. Seen from above the bore would be "I" and the direction to the mic would be "-----" the muzzle would sit on that line.

[FuzzyGrub]

The standard for measuring sounds is 1m on axis, but for obvious reasons doesn't work well with guns!    I have wondered if the standard to measure guns was 90 degrees or 180 which would be more appropriate for the shooters ear.  Thx for providing that.  Have you seen anything significant in the polar response from the different axis's?

I tried a modified RS sound meter for quicker response times.  I mounted it 2' in front of muzzle, 6" bellow the pellet path.  I still got inconsistent readings and attributed it to room modes/echoes, ie it was in my basement range.  I never did take it outside which is more anechoic chamber like.   

I'm curious to what kind of pulse width you are seeing on this impulse? 
 

[Tomg]

There is a new "standard" for firearms that are being implemented albeit not widely used yet, and that is indeed straight 90 degrees out from the ear closest to the gun.
I do think its 0.5m rather than 1m. And the goal is to make sure the firearm that can be shot without ear protection is under 120db at this location.
As far as testing at different location at polar angles around the muzzle. I have not tried yet, but that is certainly an avenue that is interresting.

[Tomg]

Response times in normal db meters even when set to fast is unfortunatly galaxies away from being able to capture the peak rise from a gun.
Fast in normal db meters is about 1/100th (0.01) of a sec. Which is pretty fast one would think. However to capture the peak rise times on guns you need something in the 0.00002 range. Think of it as cycles, it takes a reading every so often. As you can see from the numbers above, there is no way for a 0.01/s cycle to intercept the rapid peak rise. So what is meassured is really the average between two cycles. Which is called a timed average (fast for everyday use) but clearly insufficient for impulse sounds.

[FuzzyGrub]

In days of old, had calibrated mic modules feeding a mic amp feeding an A/D converter into a PC, for measuring near-field response of speakers.  I forget what the sampling rate was, but 50kHz would barely be able to measure a 20kHz signal. 

Anyway, now I can't hear high freq well, so don't mess with that anymore.  But have always found it interesting. 

[Tomg]

I should clarify the peak rise, so think of it as such: an impulse SPL rises super quick, almost like the shape of an inverted U, and duration is very short at the the top. If your cycle time is to slow, you can never capture the rise-peak level of the impulse because the structure of the measurement is taken in to long of a period which would include the overall noise profile in that given time span. You might intercept a partial of the impulse here and there, but never the whole thing.
If you think of your eyesight, which have a refresh rate of roughly 60hz which have the upside that showing 60 frames a second to your eye is an seamless experience, you cannot differentiate between the frames regardless of the action it portraits. So if we increased the frame rate to 1000 frames a second and swapping out 1 of those frames with a white frame would barely if at all register to you as a shade, let alone a distinct white frame.
The same idea applies to the measurement of gun impulse noises, the peak rise is like increasing the frame rate to something of that magnitude, which dictate that the measurement device also needs to be of sufficient speed in order to capture it.
In the above example you would need a camera that could capture 1000f/s to film a sequence that had a 1/1000 white blink in it.

The decibel meters you can get for reasonable of money (under $1000) can be likened to your eye in the above example.