11 ms·
It's a detailed, well-written review, but like most mainstream headphone reviews it makes me cringe a little. Though masked somewhat by self-deprecating remarks
by spekcular 6y ago
It's a detailed, well-written review, but like most mainstream headphone reviews it makes me cringe a little. Though masked somewhat by self-deprecating remarks, there's audiophile "woo" peaking out from behind the corner (e.g. the comment about lossless compression). More importantly, it's missing the most important part of a headphone review: a measurement of the frequency response curve!
It's fairly well known that above the ~$100 price level, headphone sound quality differences mostly disappear once equalized properly [0]. The real test is not stock AirPods versus H9i (as done in the review), it's AirPods versus H9i once they're both equalized to a common frequency target. Otherwise you're testing the manufacturer's tuning as much as the actual quality of the hardware.
Certainly, there's something to be said for the manufacturer getting it right in the first place so you don't have to bother with EQ when listening on a mobile device, etc. But I think the following two points are likely true:
1) People will score the AirPods highly in subjective listening tests;
2) The performance in those listening tests has more to do with the frequency tuning of the AirPods than anything about its hardware, and you could get a virtually identical listening experience with a $100 to $200 pair of Sennheisers and an EQ app.
[0] See e.g. http://seanolive.blogspot.com/2017/02/twirt-337-predicting-headphone-sound_17.html http://seanolive.blogspot.com/2017/02/twirt-337-predicting-h.... Sean Olive has a bunch of papers on this.
- andybak 6y agoMy first thought was "Is this audiophile as in 'believes in magic stones and digital cables that improve the warmth'"? The term "audiophile" has been permanently tarnished in my mind and sits in a corner of my brain next to homeopathy and EM sensitivity.
- tgv 6y agoOr how a felt tip pen applied to the edge of a CD can stop the laser from spilling out, and make the sound much more focused. Here's a recent review I encountered while looking for something tangentially related: "These QED audio cables promote a wide-open soundstage, both vertically and horizontally. They help vocals sound full-bodied and weighty, but with lots of breathing space above them, too. Put simply, if you covet space and detail with sure but nimble footwork and heaps of insight, consider your search for an RCA audio cable complete."
- VRay 6y agoThat sounds great, but I'm happy with my hand-bent coat hanger
- Nition 6y agoI happen to have a link to that classic post handy: https://forums.audioholics.com/forums/threads/speakers-when-is-good-enough-enough.2512/page-2#post-15412 https://forums.audioholics.com/forums/threads/speakers-when-...
- mdoms 6y agoI was certain you had to have made that comment up, but I googled it[0] and indeed it comes from a serious audiophile publication. https://www.whathifi.com/au/best-buys/accessories/best-audio-cables https://www.whathifi.com/au/best-buys/accessories/best-audio...
- jrimbault 6y agohttps://www.headphones.com/blogs/news/apple-airpods-max-review https://www.headphones.com/blogs/news/apple-airpods-max-revi... look for the paragraph : "Frequency Response & Tonality", the issue is these headphones change their tuning on the fly.
- lostgame 6y agoI had made a similar comment before reading yours. I am really glad I am not alone, here. I was absolutely taken aback by the lack of Frequency Response curve graph in a so-called audiophile review, and as a pro audio engineer on the side I am consistently driven nuts by this. There is so, so much subjectivity in headphone reviews, and so little science these days - if I had any interest in doing so whatsoever I would certainly start up my own review site, refuse to review headphones that were given to me by the companies themselves to assure a lack of bias, and - you know - actually do some science. I, for one - would be interested in doing my own sets of frequency response data and comparing them to what these companies should be providing to start, and so often do not. Ugh. Hybrid consumer/pro audio gear is the worst. >.<
- GloriousKoji 6y agoThe problem is all these review sites are pseudo experts writing copy of how it fits into a hipster/techster/trending life with just substance to pass off as an educated conclusion. But don't worry, there is a place for FFT graphs. rtings.com is excellent for reviews on audio and visual gear. Here's their review on the airpod max: https://www.rtings.com/headphones/reviews/apple/airpods-max-wireless https://www.rtings.com/headphones/reviews/apple/airpods-max-...
- passivate 6y agoThis is a headphone for listening to music, not for audio mixing. I wish you success with your review site idea, but it would not be one I would personally seek out (saying this as an audio nut myself). There are a few websites that measure more than just the sound signature though - e.g. https://www.rtings.com/headphones/reviews/sony/wh-1000xm4-wireless#page-test-results https://www.rtings.com/headphones/reviews/sony/wh-1000xm4-wi... As an aside, applying scientific principles doesn't automatically make the review any better. Science works by proposing a hypothesis/model and collecting data to see if the data matches the model. If your hypothesis/model is just average, and not world-class, then the most carefully collected data doesn't hold much weight/value. For e.g. Minor changes in the sound signature/frequency response are only audible when you listen to the same music across multiple headphones, and that too when you're carefully listening. Also the sound signature can vary when the fit of the headphone changes as per the individual head shape/ear shape/etc. All this is assuming there is enough manufacturing tolerance to produce headphones with identical sound signatures, etc, etc. Ultimately you'll have to come up with a hypothesis/model that is better than the current one. Which is not to say it can't be done, but its a tall order, but I hope you give it a shot at-least. If nothing else you'll have a cool blog post :)
- kalleboo 6y agoThese headphones have a DSP that continually reconfigures the response, I'm not sure that any synthetically generated frequency response curve would actually mean anything.
- jablan 6y agoSo it's intentionally distorting sound? Even so, it could be measured and compared, with more complex sound patterns (as opposed to sine waves or whatever is normally used for frequency response).
- pfortuny 6y agoThat is what the OP says: they sound great but you will not be able to reproduce that kind of sound with other headphones because it is dynamically modified (it is what OP calls "fake"). Not criticizing, just stating OPs argument.
- rasz 6y ago> they sound great ... and you cant measure it is at the heart of all audiophoolery. Either it reproduces source material as close to 1:1, or it distorts it tricking your psychoacoustic system to perceive a positive difference, for example by making the sound louder (wins every single blind test).
- pfortuny 6y agoExactly. That is what OP says is “fake”.
- bigtones 6y agoThis is true, and so do a lot of high end headphones that are not studio monitors. Pretty much any of them over $300 have a DSP in them configured to a curve.
- ihuman 6y ago> Otherwise you're testing the manufacturer's tuning as much as the actual quality of the hardware. Shouldn't you test the how Apple tuned the Airpods max, since you can't change it's EQ?
- spekcular 6y agoYou can modify EQ of the source. If you're playing from a computer or mp3 player, there are many apps that do this.
- astrange 6y agoYou can actually measure your hearing and change the EQ to fit it for all AirPods inside iOS's accessibility settings. The Music app also has an EQ but it's very primitive and hasn't changed since early iTunes.
- blantonl 6y agoYes, when I saw "I threw my most demanding tracks at them" I stifled a yawn but could not help but roll my eyes.
- Dylan16807 6y agoI don't see the problem with that? A combination of specific sounds and practice picking them out will make certain tracks far more demanding than others. For example, loud clear bass sounds of various types are a great test that many speakers will fail.
- passivate 6y ago> it's missing the most important part of a headphone review: a measurement of the frequency response curve! Why is it the most important part? As you said this is a main-stream headphone review. From my experience, most consumers don't enjoy listening to music w/ a flat curve - if that is where you're going with this. I've sunk several thousand dollars into high-end gear, and even I don't want a flat curve. A flat curve is most useful if you are an audio engineer working in a studio. I'd much rather use different headphones for different kinds of music, than compromise and use a flat-curve which sounds incredibly boring (to me).
- threeseed 6y agoIt says in the title that this is an Audiophile review. I would expect a frequency response graph and sound signature e.g. U, V, Harman etc.
- passivate 6y agoThe comment I was responding to accepted that this was a mainstream review, but then complained about the lack of a frequency response graph - the importance of which I personally think is way way overblown even as an audio nut myself.
- Green_man 6y agoA frequency response curve should be included because it provides useful information to (some) consumers, not because a "flat" curve is necessarily desirable. While the entire experiance cannot be discerned from a curve alone, and users can eq the signal, it provides a good idea of what the headphones will actually sound like, and a set of priorities and tradeoffs made in designing the headphones.
- passivate 6y ago> it provides a good idea of what the headphones will actually sound like, and a set of priorities and tradeoffs made in designing the headphones. OK, I'll bite. Two headphones with the same response curve can still sound very different (as you also alluded to), so what useful information are you hoping to convey to a consumer? I'm all for objective measurements provided they actually tell the buyer something useful that can influence their decision. If we have to fall back to subjective impressions of imaging, detail, sound-stage, etc, etc, then it becomes moot, and we're back to square one.
- daxelrod 6y agoThis is fascinating, thank you. As a counterpoint to your #2, what proportion people using $100-$200 headphones equalize them properly? This strikes me as similar to displays. Most people don't do any calibration, and reviewers tend to review both the monitor calibrated from factory and after their own custom calibration.
- k0stas 6y ago> It's fairly well known that above the ~$100 price level, headphone sound quality differences mostly disappear once equalized properly [0]. This depends on your definition of "mostly disappear". If you mean, ignoring, soundstage, imaging, distortion, and other parameters that affect sound, then yes, but even a non-audiophile could easily discern the difference between an open-back vs. a closed-back headphone with identical magnitude response (by your definition, the differences between these two headphones should have mostly disappeared). From linear signal analysis, magnitude adjustments do not provide enough degrees of freedom (and thus are not sufficient) to transform any response into another response. And this is true for practical sound signals not just theoretical concoctions. Take a look at any headphone review at rtings.com where they try to capture other parameters in addition to magnitude response. They test for phase response (related to imaging), PRTF (pinna-related transfer function --- related to soundstage), and harmonic distortion. These would not automatically match between headphones that were modified by gain adjustments to have identical magnitude response.
- Dylan16807 6y ago> the difference between an open-back vs. a closed-back headphone with identical magnitude response (by your definition, the differences between these two headphones should have mostly disappeared). What an uncharitable reading! They said the quality differences mostly disappear.
- k0stas 6y agoImaging and distortion directly affect quality.
- Dylan16807 6y agoYes, they do. I don't think that conflicts with anything they or I said? Their claim is that most differences are either quality or eq. Imaging and distortion are more or less just subsets of quality. Open vs. closed is one of those differences that largely falls outside those two buckets. Their suggested comparison of H9i vs. AirPods Max is two models of over-the-ear closed-back headphones.
- pier25 6y agoFascinating. Do you have links to Sean's papers? There doesn't seem to be anything on his blog.
- spekcular 6y agoUnfortunately I think they're locked behind the paywalls of various journals, or internal Harman research. You can piece a lot of together from his blog posts, but it's rather labor-intensive. Google scholar will pick them up you type in "Sean Olive." For example: https://www.aes.org/e-lib/browse.cfm?elib=16486 https://www.aes.org/e-lib/browse.cfm?elib=16486 If anyone reading this finds a free way to access the original papers, please comment.
- Tarq0n 6y ago"The correlation between price and sound quality is close to zero and, slightly negative: r = -.16" This seems absurd on its face and should really cause the researchers to re-evaluate their model. Just because their research shows people prefer a certain EQ doesn't mean there aren't any other benefits to better designed products, like soundstage, detail, clarity etc.