Plug N’ Chug, Part One
September 19th, 2008(This is the first post in a two-part series on earplugs, sound, decibels, and the noise reduction rating, which is the number used to “grade” earplugs. In part one, I’ll lay the groundwork and discuss a couple of simpler problems with the noise reduction rating. In part two, I’ll discuss more complicated issues with the rating, loudness levels, and the different earplugs I tried, as well as my conclusions on how good various earplugs are and why.)
After last week’s brief delay, I’m back for a full post. Brief aside: My Intrade bet on McCain has performed handsomely since I took it: it’s up 7-10% on average. Now that Congress is considering allowing online gambling again, I may even be able to revisit my Intrade posts (here and here) and use real money this time. We’ll see.
Enough about that. Every once in awhile I like to do a post that requires buying a bunch of stuff and trying it out, like fortune cookies. It’s been awhile since I’ve done one, and I was feeling the itch to do it again.
This time I chose earplugs. Why? Well, let me give you a little background here. I’ve always had what I perceive to be very sensitive hearing. (My (anecdotal) experience has been that this is one main cause of introversion, by the way. I believe introverts have so much external stimulus that they’re somewhat overwhelmed most of the time. )
Back in high school I read a psychology book (for fun - hey, quit laughing) and I saw the term “hyper-sensitive”, which had next to it a picture of a dog barking and a man cowering, almost in pain. “That’s me,” I thought as soon as I saw it. I guess it might be a mild version of “hyperacusis”, or excessive sound sensitivity (but it’s quite rare, so who knows). I also have infrequent tinnitus and a history of ear infections and other nonsense in my family. My twin brother (and editor) even went deaf for a week in one ear mysteriously, for the same reasons. (He’s OK now.) Plus, I attended many loud rock concerts at a young age, which may have caused the tinnitus.
So it’s possible that I may have a mild form of hyperacusis. All I know is that, when someone slams a door or honks a horn near me, I want to explode. It makes me feel like a real-life version of Daredevil sometimes (except for that whole blindness thing, of course).
The flip side of this is I often seem to hear things other people can’t. Fellow nerds may recognize one version of this, which is the ability to “hear” TVs and monitors that are on even when no speakers are playing anything. (Not much of a “super power” compared to Daredevil, but what can you do.) I can tell this by their really high pitched whir, even from several rooms over. Usually these sounds are produced by flyback transformers, which often emit sounds near the upper limit of what humans can hear (NTSC uses 15,734 Hz). My dad thought I was lying about this until we tried a lengthy experiment. Data-driven analysis runs deep in my family, I suppose.
Anyway, I can also hear many other unusual things. I can hear people clearly even when they whisper 20 feet away. I have a great ear for lyrics in songs, no matter how muddled. And I can always, without fail, anticipate people coming to my front door long before they get there. Sometimes I even hear faint knocking in the middle of a conversation in another room. It’s not all roses, though - for example, I often speak too quietly for other people to hear, since it sounds quite loud to me. Seems as though sensitive hearing encourages introversion in multiple ways.
It’s natural, then, that I would end up loving both headphones and earplugs. 75% of movies are total hell for me without earplugs, in fact. Sleeping is also rough when you can hear even very quiet noises throughout the night. I used to wear earplugs infrequently. Now I wear them every night, and often while reading as well.
When I first started wearing earplugs, though, I didn’t know a lot about them. There’s a lot of kinds of earplugs out there, including wax, foam, silicone, and rubber. Each type has its pros and cons and is worn a different way.
Plus, you have no idea how much sound they all block out. In my case, I naturally wanted the earplugs that reduced noise the most. You would think there’d be an easy way to determine this. Supposedly there is: all earplugs sold in the U.S. are labeled with a “Noise Reduction Rating” (NRR), a single number that tells you how much sound, in decibels, is “attenuated”, or blocked out by the earplugs. Pick the highest number and you’re done, right?
Alas, that would be too easy. In my shopping I saw earplugs that ranged in NRR from 12 to 33 decibels (dB). However, when I bought the earplugs, I frequently found out that how much noise they blocked out was not always proportional to the NRR. If you want to read how the NRR is calculated, here’s an in-depth link on the topic. Feel free to get back to me if you ever finish it. It might take awhile.
To spare you such mind-numbing reading, I’ll discuss earplugs and the NRR more practically. There’s a lot of issues to cover, so please bear with me. (It’ll take the majority of two posts to do so.)
The first problem is that many people don’t know how to wear earplugs properly, particularly the wax and foam ones. Wax earplugs aren’t supposed to go in your inner ear at all, for example. The package on my wax earplugs warns that if you do so, you will require “removal by an Ear, Nose, and Throat Doctor ONLY.” Not good.
Foam earplugs have the opposite problem. People often only get them in as far as the outer ear, never making it inside the ear canal; as it turns out, I did this for months when I wore them. Since the plugs only absorb sound at that point and don’t stop the ear canal from resonating, they were only about half as effective as they otherwise would be.
Eventually I learned to roll the plugs tight enough to go in my (thin) ear canals, and I liked them more after that. Anyway, the NRR now attempts to take this into account, too. When earplugs are “NRR-certified” by the company or an independent lab, they are only allowed to give test subjects written instructions. As I mentioned, this inevitably means some people use them wrong, usually not getting a good seal or not inserting them far enough. Thus, when they measure the attenuation for each person, some people get bad scores. The NRR averages this effect. That’s not great, because for most people the NRR will either be too high (they will use the earplugs improperly) or too low (for those that know how to use them).
And that’s not all. The NRR is rated in a unit familiar to most people: decibels. Seems straightforward, right? But what IS a decibel, exactly? I only had a vague idea myself, until I recently did some research. I’m going to share that research here with a lengthy aside on the decibel.
(***Feel free to skip down if you already know or don’t care about what decibels are***)
Strictly speaking, a decibel is a tenth of a Bel. You might reply, “Well, that doesn’t tell me much.” Which is true, since it’s a relative unit - that is, a ratio of like units. That means the decibel is “unitless”, since the units on the top and the bottom cancel each other out.
The unitless decibel, therefore, can extend to many different fields, such as electricity and sound. The most common context people know, though, is dB (SPL), short for “Sound Pressure Level”. Thus, the kind of decibels we’re talking about here have the Pascal, a unit of sound pressure, as their actual base unit (which is eliminated by the dB calculation). In case you weren’t sure, pressure is the ratio of force to area. The more force you have in a smaller area, the greater the pressure.
Earplugs, then, reduce the pressure (at your ears) by the amount of dB stated in the NRR. Most earplug packages, as required by law, say that the NRR typically ranges from “approximately 0 to 30″. (There’s something disturbing about the concept of a 0 dB earplug, but I digress.) In actual products, the NRR is more like 12 db to 33 dB, as I said.
Fine and good, you might say, but what do decibels actually correspond to? There is a great quick guide here for you to get the answer to just that question. (There’s also a good decibel range chart at that Wikipedia link on decibels.)
To give you a practical example, 0 dB is the point where, at the “auditory threshold”, you can’t hear sounds anymore. The “auditory threshold” in question is at 2,000 Hz, which is about midway (logarithmically speaking) through the 20 to 20,000 Hz range most people can hear. That’s because people hear sounds at different frequencies differently. You can check out 2,000 Hz tones here and here, for reference. (Also, here’s a refresher on frequency, if you need it.)
Basically, you can’t hear 0 dB at most frequencies. However, if you had a slightly higher frequency 0 dB tone (3,000-5,000 Hz, say), you might be able to hear 0 dB or even (confusingly) negative dB values. Speaking of negative dB, they have rooms so quiet that they actually have -9 dB and less as the ambient noise level. These rooms practically “eat” sound. Good luck hearing anything in there.
As for higher dB levels, here’s a reposting of one of the dB charts I linked -
* Near total silence - 0 dB
* A whisper - 15 dB
* Normal conversation - 60 dB
* A lawnmower - 90 dB
* A car horn - 110 dB
* A rock concert or a jet engine - 120 dB
* A gunshot or firecracker - 140 dB
Also from that page: “Any sound above 85 dB can cause hearing loss, and the loss is related both to the power of the sound as well as the length of exposure. You know that you are listening to an 85-dB sound if you have to raise your voice to be heard by somebody else. Eight hours of 90-dB sound can cause damage to your ears; any exposure to 140-dB sound causes immediate damage (and causes actual pain).”
If you didn’t have a working understanding of decibels before, you should now. But even then, like me, you might still be confused with the actual numbers. The decibel scale is based on not one, but two separate logarithms. The first has to do with pressure: an increase in 10 dB means a 10-fold increase in pressure level. So, even though 30 dB is only 10 more than 20 dB, it’s 10 times more pressure, which is not so intuitive.
To further confuse things, the other scale has to do with perceived loudness: an increase in 10 dB also means, on average, a 2-fold increase (or doubling) of perceived loudness. To take the same example, 30 dB sounds twice as loud as 20 dB. In summary, an increase from 30 dB from 20 dB means:
- 50% increase in decibels ((30-20)/20)
- 10 times as much pressure
- Twice the loudness
Confused yet? I know I was. I read in several places that the decibel is “one of the most misunderstood units around”, and for good reason. It’s measuring three different things with one unit, all of which are important and different.
(***If you skipped my decibel explanation, start reading again here***)
The problem is that the ear is incredibly sensitive. Since you can hear things from 0 dB all the way up past 180 dB, you need to be able to scale your hearing up and down rapidly, as necessary. Therefore, when you hit your ear with 10 times the power, it only seems twice as loud to you. That way you can keep hearing much louder sounds without it driving you crazy with the noise. If your perceived loudness scaled linearly with the power of the sound, normal conversations or lawnmowers would drive you crazy. They’d sound like an oncoming train. Conversely, if your ears were less sensitive, you might not be able to pick out differences in sound levels as well, and you would probably misjudge things more at the extremes.
So the dB scale is probably one of the better uses of a logarithmic scale, despite its tendency to confuse. I still wish that decibels would scale linearly, though, since research shows over and over again that people want and do best with linear scales. However, decibels capture both sound pressure and perceived loudness in two elegant logarithms (log base 10 and log base 2), so I guess this is the best we can do.
The really important logarithm here, for our purposes, is log base 2, which grows much faster and more like a straight line than log base 10. (Check out the graph here to see this visually.) That means it’s easier for people to understand, at least. It also makes the mental math simpler: doubling is easy for people to calculate and comprehend.
So not only are earplugs often worn incorrectly, it’s tough to even know what their loudness ratings even mean, since they’re measured in those oh-so-confusing decibels. (I am hoping my long-winded explanation gave you that impression, at least.) Still, there are more problems to address. I’ll get into other issues with the NRR, what people find loud and why, and the myriad of earplugs I purchased and tried next week. See you then!
| | | del.icio.us |
September 20th, 2008 at 11:22 am
I enjoyed your piece (and look foward to your follow-up), but one part was problematic:
The really important logarithm here, for our purposes, is log base 2, which grows much faster and more like a straight line than log base 10. […] That means it’s easier for people to understand, at least. It also makes the mental math simpler: doubling is easy for people to calculate and comprehend.
In fact log base 2 is just a multiple of log base 10, and can’t be said to be any more like a straight line. In a way the mental math is easier with log base 10 because you can just count the zeros: log10(100)=2, log10(1000)=3, etc.
Nick, you’re right about both points. What I really meant to say (and failed in over-simplification) was that, for very small values, log base 2 is more closely approximated by the line y=x than log base 10 is. When you’re talking about the mental math here, people are often using it for smaller differences and hence dB values, so maybe it seems a bit more linear in that case. However, this was an ancillary point at best.
What I also meant is that mental comprehension is greater with a log base 2 scale than with log base 10. You’re right that the actual math is easier. But actually imagining and understanding the increase is easier with doubling than multiplying by 10, I think. Hence the use of “doubling rates” in finance. I stand corrected, but hopefully at least these clarifications make sense.
As as side note, how funny is it that your blog happens to have “Log Base 2″ in the name. Maybe you found my post searching for the phrase?
It seems like some good stuff here, I’ll have to add you to my RSS reader. Thanks for your comment!
- Dave
September 25th, 2008 at 10:48 pm
I hear what you are saying about the dB scale. At work yesterday this actually came up as there is what is called a PSD for random vibrations. If you want to bring the PSD down to half levels you have to test to -6dB. This initially threw me off because power drop in circuits is one half at -3 dB. Not to mention that in communications there is a dBm unit too!
dBs are also hard, not just mathematically, but because you can not really conceptualize them real well. You can point out how long a meter is, and most people could guess how long 2 meters is based on that. For sound, even with a linear scale it would be confusing. I am pretty sure most people can not estimate very well how something with half of the volume would sound. We can determine a greater than less than relationship well, but any kind of objective measurment gets difficult.
October 4th, 2008 at 10:13 pm
Very interesting post. I was intrigued by the link to Orfield Labs, the “quietest place on Earth”.
Hearing your heart beat, your blood pulse, the sound of your own ear buzzing and your body functioning like you’ve never heard before has a tendency to be a bit unnerving.
Sounds real trippy…