What Type of Headphones Sound the Best?

[Kouioui]

What Type of Headphones Sound the Best?

Please join us with Dr. Sean Olive from Harman as he discusses Harman’s pioneering research on headphones along with loudspeakers that spans over decades to define the metrics responsible for great sound. This discussion includes Over-Ear, and In Ear Monitors (IEM's) and the response curve shape responsible for achieving the highest subjective listening preferences in blind test studies. Can you simply EQ your current headphones to get better sound based on this research? We discuss this possibility with Dr. Sean Olive.

 

[ADU]

Will definitely give this a listen. Thank you for the link, Kouioui.

 

[tusing]

One interesting aspect I found is that the Harman target was preferred regardless of cultural background or musical/listening experience - even very experienced listeners almost universally preferred the Harman target.

Another interesting aspect was that experienced listeners were almost completely unable to differentiate between Headphone A equalized to sound like Headphone B, versus just Headphone B (with the rare exception of when either headphone might exhibit a significant amount of distortion.) This implies that cup reflections and design really might not play as big of a role as we like to think, and audiophiles might not be as good at distinguishing non-FR aspects as they like to think.

As such, the argument that soundstage, detail retrieval, and other technicalities exist independently of FR might simply be false. Perhaps FR really is all there is to it, and audiophiles are desperately trying to justify spending thousands of dollars on gear they don't need.

 

[Yorkshire Mouth]

Wow! Required viewing.

What a great video. I loved the quote "You've objectified subjective listening tests."

Of note to me:

• Distortion doesn't appear to be a big issue, unless it's really bad (I'm paraphrasing).
• The idea with the 64%-15%-21% groups that they're all pretty much Harmen, but with a little more or less bass.
• Hands up who wants Amir to review the JBL Tune 710BTs? Seriously, the score appeared to be the highest on that chart, and they only cost £59.99.
• I'm very much looking forward to the upcoming products to which @Sean Olive hinted.

We're privileged to be around at a time where there's so much research and information, which is bringing us superb sound quality and saving us an absolute fortune.

Thanks to all involved.

 

[MayaTlab]

Hands up who wants Amir to review the JBL Tune 710BTs? Seriously, the score appeared to be the highest on that chart, and they only cost £59.99.

I got them a week ago. I'm in but I'm not certain what we'd extract out of a potential review other than "yep it adheres rather well to Harman's target when measured on a compliant rig" and I guess "THD isn't too bad and shouldn't be a significant problem".
The Harman products I bought lately (N700NCM2, Y600NC, JBL Tour One, K371, some JBL live which name I don't recall, 710BT) would all probably correspond to that description and yet all of them sound quite clearly different to me and I have some fairly divergent opinions about the actual SQ experience, whether it's because of the ANC electronics' behaviour (audible noise(s) from the system, or the stability of the feedback mechanism), sample variation, or the on-head inidividual headphones transfer function (actual on-head FR for me).

 

[Yorkshire Mouth]

My understanding is that JBL are making their headphones "Harman + more bass".

 

[MayaTlab]

My understanding is that JBL are making their headphones "Harman + more bass".

I am skeptical that Harman has reached a point where their intentions in terms of target response at lower frequencies for passive closed backs can be correctly translated to what people will actually experience anyway.
https://www.rtings.com/headphones/1-4/graph#1671/4007

 

[Helicopter]

Great video. I wonder what Harman has cooking.

 

[ADU]

I am skeptical that Harman has reached a point where their intentions in terms of target response at lower frequencies for passive closed backs can be correctly translated to what people will actually experience anyway.
https://www.rtings.com/headphones/1-4/graph#1671/4007

Maybe that's by design, Maya. Since most audiophiles don't seem to like their target, and think it's too bassy anyway!

(^That was just a little joke btw. Intended in fun.)

The design of some of their headphones has been a little poor though in this respect imo. So I do tend to agree with what you're saying on this.

- They need a better yoke, headband and hinge/swivel design that will allow the cups to do a better job of aligning themselves to the size and shape of your head.

- They need to use stiffer and more durable foam inserts that will last longer and hold their shape better than the awful memory foam inserts they've been using up to now on their earpads.

- They need cups that are deep enough, and large enough in diameter and shape to completely surround most people's ears. But no bigger.

- And they need better quality control in their manufacturing so there's less variation in all of this from unit to unit.

If they can do all that, and also provide a consistently neutral sound that is fairly well-extended in the bass (which is admittedly asking alot of a passive design), then I'm happy to get back on board the Harman wagon.

I wish they would experiment with some different driver designs as well.

 

[Yorkshire Mouth]

- They need cups that are deep enough, and large enough in diameter and shape to completely surround an average ear. But no bigger.

That would make them too small for 50% of the population.

 

[ADU]

That would make them too small for 50% of the population.

Agreed. See my correction above.

 

[Yorkshire Mouth]

Agreed. See my correction above.

Spot on, that man. Forgive my pedantry.

 

[ADU]

Wow! Required viewing.

We're privileged to be around at a time where there's so much research and information

Agree!

It's fun, isn't it!... Even if we still disagree a bit on how to interpret some of this stuff.

 

[ADU]

- They need to use stiffer and more durable foam inserts that will last longer and hold their shape better than the awful memory foam inserts they've been using up to now.

I understand why they've continued to use those lousy memory foam inserts in their pads btw. It's a crutch that they're using in an attempt to make up for the poor hinge/swivel/cup designs on some of their headphones.

That is not an acceptable solution though for a product that's intended for pro use.

They also need to go back to using double-sided cables. Because the designs with single-sided cables never sound correctly balanced on the left and right sides, imo.

 

[ADU]

Settin aside some of the design-related issues on the headphones, this was a pretty excellent presentation overall by Dr. Olive that covered quite a bit of ground. Some of which was new to me. There are a couple things that I think should be commented on though.

This has been mentioned before, but in their IEM tests they concluded that people do not perceive them as having as much bass. And that they like the bass level elevated about 4 dB higher than the over-ears. The moderator in the video asks if this preference for more bass could be related to the lack of tactile response in the IEMs. And Dr. Olive cites another study where they were simulating the tactile effects of car vibrations to suggest that this might be a possibility.

This is all highly conjectural though, and not really supported by any hard science imo. And what they are failing to take into account is the possibility that the rig they were using for the measurements could've been responsible for the difference in the bass measurements between the preferred IEMs and over ear headphones, rather than it being related in some way to the listeners' preferences.

This actually makes much more sense. And a much more likely explanation for the difference is that the 711 rig being used for the measurements was not doing an adequate job of simulating the actual resonant characteristics of a human ear canal. And was artificially boosting the bass measurements on the preferred IEMs versus the over-ears, compared to how they would respond in an actual human ear...

 

[ADU]

Also, one explanation that is sometimes forwarded by those who don't like the Harman curve or think it's too bassy is that the preferences in the subjective tests were somehow influenced by, or in some way reflective of tastes at lower listening levels. And effected in some way by the Fletcher-Munson effect / equal-loudness curves.

This is a question that the moderator put to Dr. Olive later on in the video, and here was his response...

The tests were performed at around 80-82 dB, which is comparable to the levels often used or recommended for mastering audio content in a studio (which is generally in the 80-85 dB range).

As mentioned previously, the level of the bass in the Harman over-ear target appears to be pretty consistent with the estimated in-room response of very well extended neutral loudspeakers...

https://www.audiosciencereview.com/...-in-eq-of-headphones.25698/page-5#post-898102

So I'm not sure why we need another explanation for this (like lack of visceral/tactile impact, Fletcher-Munson, and what have you), beyond the fact that some people simply don't like it.

Hearing loss, either hidden or actual, seems a plausible explanation for why at least some people might prefer less bass and maybe something a little closer to the direct sound of a speaker in a pair of headphones...

https://www.audioholics.com/room-ac...ons-human-adaptation/what-do-listeners-prefer

Another possible explanation is that alot of music is mastered on inferior or improperly tuned/configured gear or nearfield monitors, without a sub-woofer or enough extension in the sub-bass. And has had its bass levels overly boosted to compensate for that.

 

[ADU]

Some other possible explanations for a lower bass preference than the Harman over-ear target...

- Well-extended neutral loudspeakers have less actual bass than their estimated in-room spinorama measurements suggest.

- The Harman curve is not being applied correctly to headphone measurements. Or the measurements for the headphones are not being done correctly. And the headphones actually have more bass than their Harman compensated measurements indicate.

- Some people just don't like bass.

 

[Yorkshire Mouth]

Some other possible explanations for a lower bass preference than the Harman over-ear target...

- Well-extended neutral loudspeakers have less actual bass than their estimated in-room spinorama measurements suggest.

- The Harman curve is not being applied correctly to headphone measurements. Or the measurements for the headphones are not being done correctly. And the headphones actually have more bass than their Harman compensated measurements indicate.

- Some people just don't like bass.

I agree. I’d add it MAY be that it’s to do with bass coming from other sources (sound pressure, etc.).

The one thing I disagree with is that we presume it’s all to do with the latter, shrug our shoulders, and carry on as if it were unambiguous fact, when everyone conceded that we don’t know for sure.

 

[ADU]

I personally think that differences in hearing loss, and poor sound mastering (aka the circle of confusion) are probaby the best or at least adequate explanations for the different bass preferences. But would not completely rule out some of the others above. (And maybe a few others that I haven't thought of.)

I prefer to use sound power as my general guide though for a neutral headphone response with diffuse field compensation, rather than the Harman curve. And think that will probably end up working pretty well with the measurements made on the new HBK 5128 rig. Time will tell though, because there still isn't really enough measurement data available for this rig to draw any definitive conclusions on this yet.

Sound power is not quite as good a model for measurements made on some of the GRAS rigs. Mainly because the DF compensation curves used on some of those rigs are just not accurate enough in the upper frequencies, imho. I think it's probably still the best model for the lower frequencies up to roughly 1k though. And a somewhat ok model (albeit with some significant deviations at certain frequencies) for the general overall slope or trajectory of the higher frequencies. I still put most of my faith in the actual measured responses of some of the better and more neutral-sounding headphones as my model for a neutral response in the higher frequencies on the GRAS though. Because that seems to work the best and most reliably.