January 5, 2012 at 2:50 pm

Kids Might Actually Care About Sound Quality

Conventional wisdom dictates that people, especially the younger ones, have ruined their appreciation for the deep, rich, crisp sound that previous generations of music fans saw as the holy grail. However, conventional wisdom isn’t always wise. A new study indicates that the kids’ ears are alright, in that they prefer the sound of lossless music to MP3s, and flat-response speakers to those that distort sound in favor of bass or other frequencies.

Apathy about sound quality is widespread to the point of ubiquity, to hear some tell it — or perhaps it’s just a self-fulfilling prophecy that we’ve all aided by taking it for granted. One rapper I spoke with backstage at CES a couple of years ago (Akon) told me he mixes his albums for a single, tiny, mono cellphone speaker, a perfect demonstration of this cycle of sonic misery. People listen on shoddy headphones and speakers, the thinking goes, so producers might as well mix for them — the same way they used to take car stereos into account. The result then gets further degraded by lossy compression algorithms such as MP3 and nobody cares.

Not so fast.

First, a couple of caveats. This study (.pdf) was conducted by Harman, which, as a manufacturer of speakers, definitely has a horse in this race. Also, it only polled 18 Los Angeleno high school students: 13 boys and 5 girls. Still, the study appears to have been conducted with a fair degree of rigor — double-blind, randomized testing consisting of twelve 30-minute trials for the MP3 vs. lossless part of the study, and four trials for the speaker test.

Its findings are clear: These kids can not only distinguish between good and bad sound, but they prefer the good stuff — and when they do, they do so with more conviction than when they prefer the degraded stuff.

These results stand in stark contradiction to an eight-year study by Stanford University’s Jonathan Berger cited by the New York Times and others, which found that younger listeners not only don’t care about good sound quality — some actually prefer that it be degraded.

As the Harman study itself points out, music compressed at high bit rates using modern compression technology like AAC is often indistinguishable from the lossless CD-quality version. So its test compared 128 Kbps MP3s with the CD-quality version. These don’t have to be physical CDs, of course; lossless digital files are also CD-quality, and 256 Kbps AACs sound like CDs to most people.

In other words, the point here isn’t about CDs vs. digital files; it’s about good sound vs. bad sound.

Across the board, these subjects appear to have preferred uncompressed sound and speakers with an even frequency response. However, there were “significant” differences in the ability of each individual to discern those differences, indicating that some people simply can’t hear the difference, whereas it’s much clearer for others (maybe the ones who take good care of their ears).

The test also looked at whether kids can tell the difference between speakers, using four trials. We’ll leave out the details, especially because Harman Kardon found that people preferred its own Infinity speakers. Like all studies, this one seems to conform to the desires of the people who paid for it. But what’s interesting is that all listeners preferred “the loudspeaker with the widest, flattest and smoothest frequency response curves,” in addition to preferring uncompressed audio.

This idea that nobody cares about sound quality anymore looks like a case of false conventional wisdom, as far as this study is concerned.

Of course, the test didn’t ask listeners whether they would be willing to pay real money or alter their behavior to access better sound quality, which is the test faced by high-end audio products in the real world. (We think it’s worth spending over $100 on headphones, for instance… but do you?)

What about the idea that today’s listeners prefer the “sizzling” sound of MP3s, which lack low and high frequencies and are missing other information, often resulting in shimmery-sounding cymbals? Or the notion that they prefer speakers with uneven frequency response, like the ones in my newish car that seem tilted unfairly towards the bass end?

This study found “no evidence” to support either of these bits of conventional wisdom.

(via productionadvice.co.uk)

  • http://twitter.com/mymusicthing David D’Agostino

    I’m not convinced.  I’d like to see the results for “listening-habits” from the pre-survey, and would bet that the majority of subjects listen to most of their music with earbuds, not speakers.  So this test proves that they prefer CDs over MP3s when played over a high-quality system with good speakers –  not that they necessarily “care” about sound quality.

    My 16-year old daughter can easily hear the difference between CDs and MP3s when played over our stereo system, and she prefers the CDs.  But she listens to most of her music through iPod earbuds or on YouTube with computer speakers.

    She received a digital piano for Christmas, and I offered her Sony and Sennheiser headphones for practice. She tossed them aside and plugged in her white earbuds because they “sound better”.  To her, what I consider a more natural and full-range sound is “muffled”. 

    And even lossless files played through the iPod DAC and headphone jack are compromised, regardless of whether the sound comes through earbuds, quality headphones, or a good stereo system.

    I know there are exceptions, but for most kids, I think the battle has already been lost.  If someone wants to raise the standards for audio-quality among younger listeners, they should probably start with digital music players, high-quality files, and headphones, not speakers…

  • Anonymous

    Nonsense.  It took me weeks to find headphones that didn’t have a trailer-trash high bass boost at a nonaudiophile price.  And the ones I did find weren’t as hi-fi as I am accustomed to, but at least they weren’t noxiously booming.

  • Sean E Olive

    Thanks for reporting on my “preliminary” study. An AES preprint describing the details of the study can be found here: http://www.aes.org/e-lib/browse.cfm?elib=16108
    The study is in its preliminary phase and the sample size (as pointed out) is too small to generalize to a general population of teenagers. I’ve since tested 189  college students from US and Japan, which I will be reporting the results in a followup AES paper presented in April at the AES Convention in Budapest.The focus on this research was to see if I could replicate Berger’s results (not an easy task since he didn’t publish the details of his research),  hence my choice of using sound files encoded at MP3 128 kbps over a higher bit-rate.  Not all listeners were equal in terms of discrimination and reliability of their preference choices/ratings. However, there was no evidence that any of the listeners preferred the low-bit rate MP3 over the CD-quality format and/or the loudspeakers with the non-flat, smooth frequency response.CheersSean OliveDirector of Acoustic Research

  • Sean E Olive

    The pre-survey revealed that the students listened over a variety of playback devices: headphones, computer speakers, iPod stations, and automotive audio stations.
    But you are correct to assume that there is a difference between preferring accurate sound reproduction in a test, and actually purchasing products that fulfill your tastes.One of the problems is that the audio industry doesn’t make it easy to identify the good products from the duds. The performance specifications we provide are less useful than the specifications on the side of a tire, and the opportunity to  A/B two loudspeakers in a retail store are a thing of the past. Standard headphone specifications are also pretty misleading and not very good indicators of how they sound.So the consumer, must work very hard to find the right information to make intelligent purchased decisions based on sound quality.The science largely exists to develop meaningful performance specifications that indicate how good the loudspeaker/headphone sounds, but the industry doesn’t seem to be able to get its act together and develop standards. Finally, there is evidence in the scientific literature that MP3 artifacts (particularly ones of a temporal nature) are more audible over headphones than loudspeakers. So, you might expect even greater preference for CD to MP3 if the study were conducted using headphones.

  • Edward L Wainwright

    I hope that you’re wrong and that our young uses head phones and ear plugs are only for convenience and so not to get us old folks upset over their listening levels. Side note, I prefer my daughter listening to speakers (real or computer) due to the possibility of hearing injury. That being, said I know for a sad fact that my daughter cannot hear the difference, with my music, whether its source is from 96 or 128 MP3 or CD with quality speakers

  • Kensays

    I can ‘t speak “for” Akon but I’d be more inclined to “assume” the rational behind the “…tiny…cellphone speaker” comment is the well-worn Berry Gordy/Motown mastering methodology that if the mix sounded good enough for their high standards, on an inferior system, then it would certainly play well under ideal conditions. Maybe jumping to the conclusion that Akon actively pursues poor sonic quality is accurate. I don’t know I didn’t talk to him, you did.

  • http://twitter.com/mymusicthing David D’Agostino

    I hope I’m wrong too!

  • http://twitter.com/mymusicthing David D’Agostino

    Sean –

    I’m on your side, just not convinced that kids really care that much about sound quality.  And I realize that “care about sound quality” was part of the headline and not the study. Looking forward to the next set of research results.  Hope you can convince me!  Thanks.



  • Sean E Olive

    Since this paper was reported I have some new results reported here:http://seanolive.blogspot.com/2012/05/more-evidence-that-kids-even-japanese.html

    Based on 58 high school and college kids, as a group they preferred CD to MP3 in 70% of the trials. I tested loudspeaker preferences with this same group along with 149 Japanese college students and found all groups on average preferred the most accurate loudspeaker and preferred the least accurate speaker. The evidence suggests that kids are alright.