Why People Often Disagree About the Sound of Headphones

of 05

Scientific Reasons Why People Often Disagree About the Sound of Headphones

Man working at desk with headphones
Thomas Barwick/Stone/Getty Images

Of all the types of consumer audio products I've tested, none has been as perplexing as headphones. In the many panel tests I conducted for Sound & Vision, and the ones I now participate in for The Wirecutter, there are often large differences in the ways the listeners perceive and describe the sound of a particular headphone. We see more differences when we read the reader comments. Even after we weed out the trolls, it's obvious some people are hearing things a bit differently.

of 05

No Two Ears Are the Same

Industrial Research Products

Reason #1: Ear Canals Vary Radically.

Jacob Soendergaard, sales engineer for G.R.A.S. Sound and Vibration (the company that makes my headphone measurement gear) told me about this phenomenon, and was kind enough to direct me to a very interesting PDF that describes the development process for the ear/cheek simulators and head-and-torso simulators we use today.

As S.C. Dalsgaard of Odense University, one of the scientists involved in the program noted above, so wisely and wittily put it, "Man is manufactured within a very wide tolerance."

Soendergaard elaborated: "Each minute variation in geometry (ear canal shape, amount of folds and creases in the canal, aspect ratio of canal, location of double bends, size of tympanic membrane [eardrum], etc.) will affect the hearing perception -- particularly at the high frequencies with very short wavelengths."

You can see this in the chart above, which is an abbreviated version of a chart that appears in the PDF I linked to. This chart compares measurements taken inside the ear canals of 11 test subjects with the response of a coupler designed for hearing aid measurements. For each test frequency, you can see the coupler response (the solid line), the average response of the 11 test subjects (the circle) and the range of the responses (the thing that looks like a fat, sideways H).

As you can see, the response of the ear canals doesn't vary much below 1 kHz, but above 2 kHz the response differences become large, and by 10 kHz they're huge, about +/-4 dB. To put this in perspective, a response difference of +/-2 dB -- say, reducing bass by -2 dB and boosting treble by +2 dB -- is enough to effect a large change in the tonal balance of a headphone.

Soendergaard and I were discussing measurement in this case, but our discussion pertains to subjective listening, too, because your eardrum is effectively your measurement device, occupying roughly the same physical plane as the microphone inside the ear simulator. As Soendergaard said, referring to frequencies between 10 and 20 kHz (the upper range of human hearing), "If your measurement device is offset by a millimeter between each fitting, you’ll see vastly different results on the same person."

Thus, the differences in ear canal shape -- and the inevitable differences in the way that headphones, and especially in-ear headphones, interface with different shapes of ears and ear canals -- can cause headphones to respond very differently to different ear shapes at high frequencies. Just 1mm difference in the fit can make a headphone with flat response sound too bright or too dull.

I saw this principle in action a few years ago, when an audio writer (who'll remain anonymous) told me he really liked a particular in-ear headphone. This was a headphone that most reviewers agreed sounded extremely dull, and that my measurements showed had a huge roll-off above 3 kHz. I've collaborated with this writer in the past, and while he and I generally agree in our assessments of speakers and even over-ear and on-ear headphones, his assessments of in-ear headphones were completely different from mine. (Later, an audiologist told him that his ear canal shape was extremely unusual.)

of 05

Everybody Has a Different Sense of Space -- With Headphones, at Least

Office.com Clip Art / Brent Butterworth

Reason #2: HRTFs Vary Radically.

Head-Related Transfer Function (HRTF) is what your brain uses to locate sound in three dimensions. It involves differences in the time that a sound arrives at each of your ears; differences in sound levels at each ear; and differences in frequency response caused by the acoustical effects of your head, shoulders, and pinnae when sounds arrive from different directions. Your brain processes and interprets all these cues to tell you where a sound is coming from.

Headphones bypass the acoustical effects of your body, and alter the timing and level cues you would normally get when listening to a live performance or a set of speakers. Unfortunately, your brain doesn't have an "HRTF Bypass" button. When you're wearing headphones, your brain still listens for those HRTF cues, doesn't hear many and thus gives you the feeling that most of the sound is coming from inside your head.

As I learned when I visited a company named Virtual Listening Systems in early 1997, everyone has a different HRTF. To created what became the Sennheiser Lucas headphone processor, VLS measured the HRTF of hundreds of test subjects. They did this using tiny microphones placed inside the subjects' ear canals. Each test subject sat in a small anechoic chamber. A small speaker on a robotic arm emitted MLS noise bursts. The robotic arm moved the speaker through more than 100 different positions, at various horizontal and vertical angles, each time emitting the test bursts so the microphones in the subjects' ears could "hear" the effects their bodies and ears had on the sound.

(Headphone enthusiasts may note this is similar in some ways to the measurement procedure Smyth Research uses in its A8 Realiser processor.)

I got to go through VLS's test myself. The company's scientists then took my results and ran them through a processor that would alter an audio signal to precisely mimic my personal HRTF. The result was amazing, like nothing I've heard from any other headphone processor. I heard a precise, perfectly centered image of a vocalist directly in front of me -- something that technologies such as Dolby Headphone could never achieve for me.

VLS took the results from hundreds of test subjects to create the Lucas processor's 16 different presets, each one tuned to simulate a different HRTF. Clicking through all the presets, it proved difficult to settle on one. I remember some were obviously better than others for me, but I had a hard time choosing among the best four or five presets. None worked anywhere near as well as the tuned-just-for-me processing I heard in VLS's lab.

That's probably why most headphone processors have far fewer options. Still, though, they have to shoot for some sort of average HRTF. Maybe you'll get lucky and fall close to that average. Maybe the effect will be too extreme for you. Or maybe it'll be too subtle.

Because everyone's HRTF is different, each of our brains has a different compensation curve -- sort of like an EQ curve -- that it applies to incoming sounds. When that compensation curve is combined with the characteristics of your body, the result is the sound you hear every day. When the characteristics of your body are eliminated through the use of headphones, your brain still applies the same compensation curve. And because each of our compensation curves is a little different, our responses to the same headphone can be different.

of 05

No Seal, No Bass

Brent Butterworth

Reason #3: The Fit Changes the Sound.

Getting good performance from headphones depends to a large extent on the fit. Specifically, this means the fit of the earpads of an over-ear headphone around your ear, the fit of the earpads of an on-ear headphone on your pinna, or the fit of the silicone or foam ear tip of an in-ear headphone inside your ear canal. If there's a good seal, you'll get all the bass the headphone was designed to deliver. If there's a leak anywhere, you'll get less bass -- and you'll perceive the headphone's tonal balance as more trebly.

In part, the physical characteristics of your body determine the fit of the headphone. For example, if none of the tips that come with an in-ear headphone fit you well, that headphone won't sound as good to you. This can be a problem for me because I have unusually large ear canals, and for my colleague Geoff Morrison because he has unusually small ear canals. This is why I always praise manufacturers for including five or more sizes/styles of ear tips with their in-ear headphones. It's also why Comply foam tips are worth checking out if you're dissatisfied with the sound of your in-ear headphones.

Poor fit is also common with on-ear and over-ear headphones. I would speculate it's a bigger problem with the latter, because there are so many potential obstructions to a good seal. These include long and/or thick hair, eyeglasses, and even ear piercings. Push the ear pads out just a tad, even half a millimeter, and you're likely to lose enough bass to have a big effect on the sound of the headphone.

Over- and on-ear headphones can fit some people better than others. Some audiophile-oriented headphones like the Audeze LCD-XC have ear pads so large they can't seal around the ears and cheeks of relatively small people, especially women. By the same token, some supposedly over-ear headphones don't actually have enough space to accommodate large earlobes like mine.

It's worth noting that a bad seal can have a positive effect. With bass-heavy headphones, a little less seal can make their response sound flat -- something we experienced when doing the Best $100 In-Ear Headphones shootout for The Wirecutter. My favorite headphone of this bunch was the Grain Audio IEHP, which to me had a wonderfully flat and natural response. The IEHP sounded so good that I assumed that the largest of the supplied silicone tips was giving me a good seal. For everyone else, though, the IEHP's bass was way overpumped. Apparently I wasn't getting a tight seal, but everyone else was -- and it completely changed my perception of the headphone for the better.

of 05

Reasons That Aren't Exclusive to Headphones

Brent Butterworth

Reason #4: Personal Tastes Differ.

Of course, there are also reasons that people report different perceptions of headphone sound that are just as applicable to other audio products.

The first is the most obvious: Different people have different taste in sound. Some may simply like a little more bass than you do, or a little more treble. Obviously, they'll prefer different headphones than you do.

That's legit to a point. Above and beyond normal variations in taste, some people do have misguided -- or more bluntly put, wrong -- ideas about sound. We've all encountered people whose idea of good sound is little more than ridiculously loud bass. Some audio enthusiasts prefer highly exaggerated treble, which they mistake for detail and accuracy. I went through that phase myself, but the invaluable writings of J. Gordon Holt straightened me out.

Whatever makes these listeners happy is OK, but it doesn't extrapolate to useful judgments about headphone sound except for others who share their extreme tastes, and no competently performed, unbiased evaluation is likely to confirm their assessments.

Reason #5: Hearing Ability Differs With Age, Gender and Lifestyle

While most of us start out life with fairly comparable hearing capabilities, our hearing capabilities change over the course of our lives.

The more you're exposed to loud noises, the more likely it is that you've lost some of your hearing at high frequencies. This is especially a problem for people whose recreational activities (going to loud concerts, driving race cars, hunting, etc.) and/or work (construction, military, manufacturing, etc.) exposes them to loud noises.

The older you are, the more likely it is that you've experienced some high-frequency hearing loss. This is especially an issue with men. According to the paper "Gender differences in a longitudinal study of age-associated hearing loss," from the Journal of the Acoustical Society of America, "...hearing sensitivity declines more than twice as fast in men as in women at most ages and frequencies..." This is in part because men are typically more often involved than women are in activities where they're exposed to loud sound, such as all the ones noted above. And that's because studies show men are more comfortable listening to loud sounds, by a factor of +6 to +10 dB over the levels at which women are comfortable listening.

Obviously, the perceived characteristics of an audio product will change as the listener's hearing changes. For example, high-order distortion harmonics, which occur at frequencies 5 or more times the fundamental frequency of a sound, will obviously be more troubling to a 25-year-old woman than they are to a 60-year-old man. Likewise, a 12 kHz response peak may be barely audible to the 60-year-old man, yet intolerable to the 25-year-old woman.

What Can We Do?

The obvious question is, how can we evaluate headphones in a way that's meaningful and useful to every listener? And for every headphone?

Unfortunately, we probably can't. But we can come close.

In my opinion, the answer is to use multiple listeners with different head shapes, different genders and different ear canal shapes/sizes. This is exactly what Lauren Dragan does in the headphone reviews she organizes for The Wirecutter, and it's what we did at Sound & Vision when I was there.

I link to other reviews of the headphones I review when possible. I also incorporate lab measurements -- here and in my headphone reviews for SoundStage! Xperience -- to give an objective idea of what a headphone's response is.

The "gold standard" would be to incorporate multiple listeners plus lab measurements. I did this in my Sound & Vision days, but I'm not aware of any publication that does it at present.

There's one simple rule we can all take from this: Be careful before you ridicule other people's opinions of headphones.

Special thanks to Jacob Soendergaard of G.R.A.S. Sound and Vibration and Dennis Burger for their help and feedback on this article. If you have any questions or comments, please e-mail me at the address listed in my bio on this site.