Monday, February 25, 2008

Turn it up!

The latest issue of the Journal of Speech and Language Hearing Research has an article on the loudness of on-stage loudspeakers (monitors) used by musicians to hear themselves. Jeremy Federman and Todd Ricketts had professional singers adjust two types of monitors to two loudness levels. The two types of monitors were floor speakers and in-ear buds, the two levels were minimum acceptable listening level (MALL) and preferred listening levels (PLL). You can imagine this, having your stereo at the softest possible level that you can still hear, but won't disturb your spouse, that is a MALL. But when you crank it up so you can really hear it, that is a PLL. PLLs and MALLs were set by each musician listening to their own voice. Other musical sources and crowd noise simulations were set at different levels by the scientists.

The graph at top shows that in-ear buds were set consistently at lower levels than floor monitors, about .6 dB, except the left-most condition The different sets of bars show different ratios of Overall Music Level (OML) and Crowd Noise (CN), so the outlier is when the overall music level was lowest (92 dB) and there was no crowd noise. But ANOVAs revealed a significant difference between monitor types overall, with no significant effect of the overall loudness or trial condition.

The authors have a good caution though:
Although a statistically significant result was observed between the two monitor types for PLL, the actual mean difference of ~0.6 dB (see mean data in Figure 2) could be regarded as nonsignificant in a real-world consideration of noise exposure recommendations. For this study, a 3 dB or greater difference in listening levels between monitor types (i.e., NIOSH time-weighted average) would have been considered functionally significant (as opposed to statistically significant) because such a result would potentially double allowable real-world exposure time based on NIOSH recommendations. The statistical significance of the findings is related to the large number of samples involved in statistical analysis.

It is refreshing to see a study that admits the different between statistical significance and actionable significance. A serious result from this study is the average dB of MALLs: 109.0 dB for floor monitors and 103.2 dB for in-ear buds. OSHA safety regulations would allow exposure to 105 dB for 1 hour each day before serious damage might occur to hearing. Most amplified shows last considerably longer than one hour, especially when pre-show soundchecks are included. So to you amplified musicians out there, invest in a good pair of sound-attenuating in-ear monitors and work with a hearing expert to get them fitted well. That can double or quadruple the amount of time you can safely listen to yourself play!

Update: Daniel Wolf shows that classical musicians are also in danger. I had a trumpet teacher who always carried sets of ear plugs in his case, which he would hand out to the violists or 2nd violinists sitting in front of him.

J. Federman and T. Ricketts, "Preferred and Minimum Acceptable Listening Levels for Musicians While Using Floor and In-Ear Monitors," Journal of Speech, Language, and Hearing Research, Vol 51 (Feb. 2008), 147-159.


Harlan said...

For vocalists and brass musicians, any idea if anyone's ever studied the "effective sound level" of one's own instrument while wearing earplugs? I've had gigs where headphones and earplugs were both necessary at times (playing horn for the musical "Tommy"), and I become much louder in my own mix when wearing the latter.

In fact, I'd expect that in order to achieve a useful sound level, you'd have to turn others up much higher in this situation.

Scott said...

I know that bone conduction is a major element of hearing one's own voice or brass instrument. So when you plug your ears from any ambient air pressure conduction, the recorded mix will be added to your bone conducted performance, making you appear louder than anyone else. I don't recall how many dB bone conduction can contribute, though.

Harlan said...

Yeah, completely understood; I realize that's how the brass player would be hotter in the resultant mix. Wonder how you'd measure, though, how many dB this'd be? By comparison with external reference source?

Wonder if anyone's ever tried to do the noise-cancellation thing for something like this, generate the opposite sounds and what-not to cancel something like this out.

wtfisjohn said...

Prof Speigelberg, i work in live sound reinforcement for a living out here in the east, mainly concentrating on being a monitor engineer. Recently, i did a concert with The Trampps (of disco inferno fame) and, man, did they need in-ears. We try and convince the older groups the most that if they're going to be playing and touring, by some in-ears for pete's sake! i had the onstage monitors cranked so hot that i was constantly buzzing around keeping the dreaded ring from going out of control (at one point i had 6 different frequencies ringing in 6 different monitors in a sounded cool, really cool, but, um, yeah, not supposed to happen!). The keyboard player's channel was actually opened up all the way, i couldn't put another decibel through it, and after the first set he came over and asked for me.

In ears are a magical thing. If you're a group using sound reinforcement in any large setting they're a must in my book. and monitor engineers beware those stacks. i suffered minor high end hearing loss for 3 days after the tramps concert because i had to sit next to the stacks. i had "sound canceling ear buds" in for my mixing and it didn't save me.

-John Chittum-