ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: dB SPL - A Weighted vs C Weighted, Fast vs Slow  (Read 560 times)

Douglas Cyr

  • Newbie
  • *
  • Offline Offline
  • Posts: 26
dB SPL - A Weighted vs C Weighted, Fast vs Slow
« on: May 11, 2021, 05:39:14 pm »

Hello, I just got a RadioShack SPL meter and am curious about the most accurate way to use it, as well as the general differences between the different weightings. I know they're meant to reflect our ear's changing frequency responses at different pressure levels, and I've read that if measuring above 90 dB SPL I should use C weighting, below 90, A weighting.

Any other information I should consider?

Thanks,
Doug
Logged

Caleb Dueck

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1387
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #1 on: May 11, 2021, 08:10:05 pm »

Hello, I just got a RadioShack SPL meter and am curious about the most accurate way to use it, as well as the general differences between the different weightings. I know they're meant to reflect our ear's changing frequency responses at different pressure levels, and I've read that if measuring above 90 dB SPL I should use C weighting, below 90, A weighting.

Any other information I should consider?

Thanks,
Doug

Generally, A-weighted Slow mimics human ear perception the best. 

Slow doesn't 'catch' the transient spikes, so will give a lower number.  The human ear gathers info over time, so super short duration spikes don't seem as loud.

A weighted doesn't take the bass into account (Google A vs C vs Z weighting curves).
Logged
Experience is something you get right after you need it.

Scott Bolt

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1523
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #2 on: May 11, 2021, 09:38:28 pm »

For PA speaker measurements, I generally advocate for C weighting which (as Caleb pointed out) actually takes into account the lower frequencies.

Pretty much any cheap crap HF horn can put out TONS of HF noise (even if it is REALLY bad noise) to make an A weighting show impressive SPL numbers IMO.

The original purpose of the A weighting was to assess SPL levels that would damage the human ear.  As a result, little to no LF is considered.

Now, the real problem is that SPL is just a poor way to assess a speaker.  Let me explain....

SPL is a weighted average of a bunch of frequencies.  A speaker with a very unbalanced frequency response .... with lots of annoying HF, and lots of holes in its frequency response can show very impressive numbers.

Many people (especially the many highly knowledgeable people here) can use either speaker processors, or even a multi-band eq on a mixer to "tame" most speakers to make them decent sounding; however, what you find once you have done this is that while the speaker now sounds much better, it can't put out much SPL at that quality of mix.

There really needs to be another measurement that ISN'T so subjective.  140db SPL of crap noise is STILL CRAP ;).

If there was a way to force a speaker to put out a balanced sound and then to measure that with a flat scale, it might make a nice measurement.  Seems like this is a great dream to have, but for now, we are left doing real world ear testing and requesting opinions of others that have heard many different speakers ;).

In general, more expensive (especially powered) speakers are capable of putting out higher output levels while sounding good doing it.  In fact, many of the better powered speakers remain pleasant even into full clip.
Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1233
  • Edmonton, AB
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #3 on: May 11, 2021, 09:46:44 pm »

{...} Any other information I should consider?

Like any measurement tool, the first question you need to answer is "what am I trying to measure?"

dB SPL (weighted or otherwise) is used in many different contexts. Are you concerned about hearing loss? Regulatory compliance? Sound system potential? Night-to-night consistency?

Without knowing what you're trying to measure it's impossible to make suggestions about how to measure it.

-Russ
Logged

Luke Geis

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2234
    • Owner of Endever Music Production's
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #4 on: May 11, 2021, 11:25:59 pm »

The only time I pull out my dB meter is when I have the volume police in my area. Where I live we have an 80db A-weighted limit at any adjacent property line. A very hard limit to heed to!

I prefer C weighting because it does show you a little more of a picture of what is coming out of your PA. As mentioned, it is really only useful for measuring above about 90db. I peek at the A-weighting to get a sense of how bad I could be hurting people. If you see 100+db of A-weighted media, it is pretty darn loud. I try and keep my shows at or around the 105db C-weighted mark. It is a pretty visceral experience when you have 110db+ of media spanking you, but you can't bludgeon people like that for long.

I have come to find that the more linear the PA, the more muted and tame the volume appears to be even though the meters say it is set for slay. If your PA seems quiet even though you know you are pushing hot levels, be sure to check your SPL meter for safety. Well tuned PA's seem to sound quieter and more tame, giving a flase impression of the actual SPL.
Logged
I don't understand how you can't hear yourself

Douglas Cyr

  • Newbie
  • *
  • Offline Offline
  • Posts: 26
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #5 on: May 12, 2021, 12:16:02 am »

Like any measurement tool, the first question you need to answer is "what am I trying to measure?"

dB SPL (weighted or otherwise) is used in many different contexts. Are you concerned about hearing loss? Regulatory compliance? Sound system potential? Night-to-night consistency?

Without knowing what you're trying to measure it's impossible to make suggestions about how to measure it.

-Russ

Good point. I initially got it to use with an RTA mic to calibrate my studio monitors, and have since been using it to measure different environments to try to get more of a concrete sense in my head for different loudness levels.

https://library.municode.com/md/frederick/codes/code_of_ordinances?nodeId=PTIITHCO_CH15OFIS_S15-21.2GEREOIPR

The Noise Ordinance in my city is incredibly low! Am I misinterpreting that, or is it saying that residential noise levels can't exceed 65 dBA?!
Logged

Matthias McCready

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 344
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #6 on: May 12, 2021, 10:53:18 am »

Good point. I initially got it to use with an RTA mic to calibrate my studio monitors, and have since been using it to measure different environments to try to get more of a concrete sense in my head for different loudness levels.

https://library.municode.com/md/frederick/codes/code_of_ordinances?nodeId=PTIITHCO_CH15OFIS_S15-21.2GEREOIPR

The Noise Ordinance in my city is incredibly low! Am I misinterpreting that, or is it saying that residential noise levels can't exceed 65 dBA?!

My favorite line from the ordinance

"It is unlawful for any person or persons to play, use, operate or permit to be played, used or operated, any radio, tape recorder, cassette player or other machine or device for reproducing sound, if the sound generated is audible at a distance of fifty (50) feet from the device producing the sound..."

 :o :'( :-\

Yes your city ordinances are rubbish, as are most. The people who write them often have very little understanding of physics as relates to sound propagation. At least they specify an SPL weight (dBA) as many municipalities don't specify. However they do fail to specify if the SPL dBA is fast, slow, 1 minute, etc.

Logged
Measure twice, and cut once; this is especially important if you are a mohel.

Michael Lawrence

  • Full Member
  • ***
  • Offline Offline
  • Posts: 129
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #7 on: May 12, 2021, 12:39:53 pm »

I would just like to submit for consideration that the idea of choosing between A and C weighting curves based on the measured level to "make the measurement look more like it sounds" is rather outdated and doesn't actually hold much technical water, even though we have all been taught something similar.  That was certainly the original intent of the weighting curves when they were standardized in the late 1930's based on the original equal loudness contour research, which differs quite a lot from the currently accepted values.

I have hopefully managed to attach a graphic that shows how the 100 phon equal loudness contour compares to A and C weighting. Neither are a reasonable approximation of the tonal response of human hearing at concert level. B weighting would be a much better fit, and the fact that it has been removed from most of the applicable standards suggests that perhaps the intent of an SPL measurement is not to mimic human loudness perception (In fact, loudness meters use a modified B weighting curve because it does attempt to characterize human loudness perception), although in my opinion, creating a one number metric that can accurately incorporate all of the numerous factors that play into human loudness perception is a tall order indeed. An AES panel at 8AM running at 85 dBA would feel way too loud but a rock concert running at 95 dBA might feel not loud enough. There's so much context to be considered in how our brains process this stuff, but that's quite a different discussion.)

We have had the benefit of almost a century of audiological and psychoacoustic research since the weighting curves were first put in place, so it's not unreasonable to expect that there may be a more nuanced approach since then.

I realize this rocks quite a few boats but once one comes to peace with the fact that SPL measurements don't well characterize human loudness perception, one might reasonably ask "then what good is it?": an objective measurement that is used to answer objective questions, such as:

  • am I delivering a consistent product from night to night?
  • am I putting people at risk / is this sound level safe?
  • am I breaking a law or violating a rule?

A lot of the work I have been doing over the past year or so is working with FOH engineers to help them exploit the difference between perceived loudness vs measured level to end up with mixes that feel loud, full, dynamic, and impactful, but don't hurt people and don't get your band fined. Rather than try to smooth over the gaps, learn to stick your fingers in there and take advantage of the disconnect.

 
As has been pointed out already, weighting curve (and time domain metric) are best chosen based on what we want to measure. C weighting when we want to include the LF, A weighted when we don't.

C weighting is effectively broadband (typically differs from unweighted / Z weighted measurement by less than a dB for broadband material) and plays an important role in characterizing both nuisance noise and characterizing broadband / LF sound exposure hazards. As noted above, it's very common for jurisdictions in the US to have A-weighted nuisance noise limits at the property line, which is an unfortunate example of taking a measurement that fails to characterize the problem, and there are many cases of events being completely compliant from a legal perspective and also causing legitimate nuisance noise complaints. From an application standpoint this is often addressed by octave-banded SPL measurements, with 63 Hz octave band being where the majority of the "noise pollution" action is.

A weighting, somewhat by accident, ends up as a decent approximation of the transfer function of the ear such that it allows us to reasonably characterize the energy range that tends to correspond to noise-induced hearing loss, which is why NIOSH, OSHA, and WHO noise exposure metrics specify A Weighted measurements. (See the Appendix 1 here.) That is not to say that LF can't damage hearing, it certainly can, but that's more recent knowledge and there are still a lot of questions to be answered.

For anyone interested in digging into this at a deeper level, I would recommend this AES report on understanding and managing sound exposure and noise pollution at outdoor events, which has a bunch of great resources collected throughout. For a quick crash course on what the SPL metrics mean and which you should choose, here's a good start: https://www.rationalacoustics.com/spl-guide/
« Last Edit: May 12, 2021, 12:43:26 pm by Michael Lawrence »
Logged

Dave Garoutte

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2528
  • San Rafael, CA
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #8 on: May 12, 2021, 02:53:20 pm »

I did a sound survey for the local 'shopping center', where I took readings all around the property line.
I found that in general the ambient traffic noise was louder (sometimes way louder!) than the music.
However you could still hear the music UNDER the traffic noise.
The human ear/brain is amazingly efficient at distinguishing different sounds at the same time.
If the city was trying to 'save' the neighbors from hearing the music, I don't think any noise limit would accomplish that.
« Last Edit: May 13, 2021, 12:05:19 pm by Dave Garoutte »
Logged
Nothing can be made idiot-proof; only idiot resistant.

Events.  Stage, PA, Lighting and Backline rentals.
Chauvet dealer.  Home of the Angler.
Inventor.  And now, Streaming Video!

Michael Lawrence

  • Full Member
  • ***
  • Offline Offline
  • Posts: 129
Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #9 on: May 12, 2021, 03:11:38 pm »

I did a sound survey for the local 'shopping center', where I took readings all around the property line.
I found that in general the ambient traffic noise was louder (sometimes way louder!) than the music.
However you could still hear the music UNDER the traffic noise.
The human ear is amazingly efficient at distinguishing different sounds at the same time.
If the city was trying to 'save' the neighbors from hearing the music, I don't think any noise limit would accomplish that.

Exactly, that's a great example of why nuisance noise is such a tricky thing to characterize. My neighbors love to crank up the bass, and it's plainly audible at my desk even though it barely shows up in an SPL or Spectrograph measurement. There's a lot of work still to be done there and it's very interesting but will likely always be very difficult to legislate.
Logged

ProSoundWeb Community

Re: dB SPL - A Weighted vs C Weighted, Fast vs Slow
« Reply #9 on: May 12, 2021, 03:11:38 pm »


Pages: [1]   Go Up
 



Page created in 0.029 seconds with 19 queries.