A Slippery Notion: Why Single-Number SPL Values Are Useless

A Slippery Notion: Why Single-Number SPL Values Are Useless

Q: I was thinking about going to a gig but am worried about the loudness and damage to hearing… I know these dB values are quite dependent on the PA, engineer, venue, etc, but was wondering what would be the kind of average dB someone going and seeing a live electronic music gig in a medium sized indoor venue would be exposed to.

The gig is techno and i’m not really confident the people putting on the gig are concerned about volume control for the audience.

I’m going to answer a different question, and then you’ll see why your question is very hard to answer. This is one of those questions that seems simple at first and ends up getting very complicated very quickly when you dig into it. Buckle your seatbelt. The truth of the matter is that a single-number SPL value is virtually useless.

For example, here is a screenshot of Smaart’s SPL History timeline:

There are thirteen different SPL metrics being tracked by the software, and all are different – and equally valid – ways of answering the question “how loud is it?”

For a dB value to be meaningful, we need to indicate both time-domain information (over how long of a window was it measured / averaged / integrated) and frequency-domain information (what weighting curve is in use). Without both of those pieces of information, we can’t really draw any useful conclusions about the number. Again, all those 13 traces on that chart represent the same actual event measured at the same time from the same spot – they’re just different combinations of integration times and weighting curves, so we need to know which we’re talking about.

So when venue management says “keep it under 100 dB” that’s a technically meaningless statement without much more specific information.

To address the specifics of your question, when we’re talking about hearing damage we’re usually talking about noise exposure over a period of time. For that, the “Slow/Fast/Peak” metrics used by most handheld meters are pretty much useless because they show us a very short period of time, and we’re far more concerned with the equivalent exposure to the ear over a longer period. (We say “equivalent” because shows have loud bits and soft bits and so we need to integrate over time to really get an accurate idea of how much energy we’re exposing our ears to.)

The National Institute of Occupational Health and Safety says that for a two-hour exposure (about the average concert), you will minimize (not eliminate!) your risk of exposure-based hearing damage at levels below 91 dB SPL (A weighted). Measuring as an A-weighted equivalent sound level over a window of 10 or 15 minutes (shorthand: LAeq10 or LAeq15) gives us a better idea of what we’re actually dealing with in terms of exposure.

The trick is that equivalent sound level (Leq) metrics are not measured by most meters. Usually those only have SPL Fast or Slow, with your choice of A or C weighted. (Fast is a 0.125s time constant, and Slow is 1s.) Those give us very little information about the type of exposure that damages the ear.

Most of the laws and regulations that deal with this stuff are very poorly written.

Case in point: here is an SPL timeline from an event I mixed last week, logged with a calibrated, class-compliant microphone:

The gold-brown line is LAeq10. This is the metric that tells us whether or not we’re doing damage. Here it hangs right around 94 the entire time, which is the NIOSH recommended limit for an hour, and this is a one-hour show, so we’re good.

The green trace is Peak C. (If you’re curious about the name, SPL is by definition an RMS value so Peak C is by definition not an SPL measurement but we still log it because it’s in the standard.) You can actually see each individual song, and that it’s hanging around 112 – 114 or so throughout the show pretty consistently. If a person was metering this, they would probably (improperly) cite you for violation. Same for the pink trace – that’s SPL C Slow, and here it’s sitting around 100. Depending on the wording of the statute and/or knowledge of the person enforcing it, it might be concluded that you’re operating at an unsafe level, when you’re not. A shocking amount of legislation just says “98 dB” (not even “SPL”) and doesn’t give a weighting curve or an integration time, and without that you can’t really approach compliance in a serious way. And more importantly – it has no bearing on whether or not the show is actually at an unsafe level. So there is a lot of education that needs to happen around this issue for us to move forward.

Best answer: I don’t leave the house without earplugs.

One Reply to “A Slippery Notion: Why Single-Number SPL Values Are Useless”

Leave a Reply

Your email address will not be published. Required fields are marked *