• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Theory: If I EQ headphones to Harman AE/OE curve, what other attributes should I look for?

JJB70

Major Contributor
Forum Donor
Joined
Aug 17, 2018
Messages
2,905
Likes
6,148
Location
Singapore
Comfort is essential for a headphone. Unlike other audio gear you wear a headphone and if you don't enjoy wearing it then you won't enjoy it regardless of how great it sounds. That's why I understand why some hate Etymotic IEMs, as their deep fit isn't for everyone and I do understand why some just don't like them. Clamping force, stability, contact point feel, earpads, cup shape etc all make a difference, and especially weight. I just can't get on with the Audeze LCD family, it makes me sad as some of them sound very good but I just find them heavy and uncomfortable. Conversely, I find most Beyerdynamic headphones I've tried to be wonderfully comfortable even with that Beyer treble thing.

Easily replaceable pads, I also think replaceable cables are important, preferable with a standard jack type connection. I must admit I like nice quality materials and premium feel, but on the other hand I do also acknowledge that the Sennheiser HD580/HD6xx family are probably amongst the most rugged and durable headphones out there based on my experience with the 580's despite the low rent feel.

For sound signature, that one can vary. I like a flat and neutral sound signature (I love my ER4 SR IEMs) but I also like more coloured tuning when I'm in the right mood. The Sony MDR Z7 is a bit like a "Beats lite" headphone, but I do like my MDR Z7's. And I love my ATH MSR7's despite them being slightly bright.
 

Fluffy

Addicted to Fun and Learning
Joined
Sep 14, 2019
Messages
856
Likes
1,423
I just can't get on with the Audeze LCD family, it makes me sad as some of them sound very good but I just find them heavy and uncomfortable.
I sympathize. I really disliked Audeze for a long time, not only because of the heaviness issue, but also because I just didn't like their sound signature. But then I tried the LCD2C at Canjam. They're relatively light and have none of that dense wood they use for the other ones. And by adding four Dekoni nuggets to the heads tarp, it became really comfortable to put on even for long listening sessions. I like the sound, but it's overall pretty warm so I don't know if it's for everybody. Maybe worth a try though.
 

mitchco

Addicted to Fun and Learning
Audio Company
Joined
May 24, 2016
Messages
640
Likes
2,397
I don't know how the LCD2C sounds relative to the LCD-4z, but the latter has abysmal high frequency coloration that, to my ears, make them un-listenable, even using Audeze Reveal eq. I posted a link to the review that shows the bumpy ride frequency response. One can also download binaural recordings of the headphones comparing them to the original source track and another set of headphones.
 

Fluffy

Addicted to Fun and Learning
Joined
Sep 14, 2019
Messages
856
Likes
1,423
@mitchco That's what I hated in Audeze, those weird sibilances that makes everything piercing or just sound odd. I suspect those headphones are just made for old people that need a big boost in the high end to hear it :p. The LCD2C have none of that nonsense, it's smooth sailing all the way to the top.
 

MZKM

Major Contributor
Forum Donor
Joined
Dec 1, 2018
Messages
4,244
Likes
11,479
Location
Land O’ Lakes, FL
Yeah, good shout. Here is their opinionated weighting of different factors that contribute to their "Sound" score, fwiw:

Score components:
I wonder if Sean Olive has done any research on soundstage. I know nothing about it. But that weighting surprises me given the research into the strong correlation between FR and user preference, which as I understand it has a 95% correlation with user preference based on their latest study.
And for their Critical Listening score, the breakdown is thus:


• Bass: 16.8%
• Mids: 16%
• Treble: 15.2%
• FR consistency: 4%
• Imaging: 7.2%
• Soundstage: 16%
• THD: 4.8%
• Build quality: 10%
• Comfort: 7%
• Breathability: 3%

I think their weights are OK (would put comfort as more important than build quality), but their scoring could use some work, like how much deviation constitutes what score.
 
OP
N

nhunt

Member
Joined
Oct 6, 2019
Messages
53
Likes
51
A tangent but somewhat related: cruising around various headphone forums, it’s amazing to me how far people will go to mechanically modify headphones or adjust them with an amp/wires/whatever that has a non-neutral FR, before ever firing up a parametric EQ. It’s like watching people at a miniature golf course trying to putt a golf ball in a hole instead of just picking up the ball and walking over to place it in there!
 
OP
N

nhunt

Member
Joined
Oct 6, 2019
Messages
53
Likes
51
Sorry for the double post, but one more “shower thought”: maybe another big contributor to the circle of confusion for headphones is the significantly different loudness levels people prefer and the likely lack of loudness equalization.

Anecdotally, I’ve notice that sometimes when I have different people try my headphones they will either crank them way up or turn them way down, much more so than with my loudspeakers
 

pwjazz

Addicted to Fun and Learning
Forum Donor
Joined
Sep 21, 2018
Messages
507
Likes
746
Sorry for the double post, but one more “shower thought”: maybe another big contributor to the circle of confusion for headphones is the significantly different loudness levels people prefer and the likely lack of loudness equalization.

Anecdotally, I’ve notice that sometimes when I have different people try my headphones they will either crank them way up or turn them way down, much more so than with my loudspeakers

Yeah, as someone who owns some fairly different headphones I can attest to the fact that I tend to listen to each at different volumes.
 
OP
N

nhunt

Member
Joined
Oct 6, 2019
Messages
53
Likes
51
Some more factors that I thought of outside of FR curve at reference (some already stated here):
- Quality crossfeed, especially for older mixes
- Equalization that matches loudness contours
- Personal preference for FR in 10Khz+ that isn't part of Harman target
- PEQ that can be applied to wireless headphones connected to portables
- ANC
- Closed vs Open headphones
- Subbass/haptics(?). Have noticed that no matter how much I EQ I can't get a similar senstation from say closed front volume headphones and open headphones
- Soundstage. Although I think this is related to FR but I don't particularly understand soundstage very well, I need Floyd Toole to add a couple chapters on headphones to his book =)
- Comfort
- Industrial design
- Headphones that don't change their FR too much depending on how I place them on my head each time
- Pads that don't compress/deform and alter the FR. Or are easily replaced.
 
OP
N

nhunt

Member
Joined
Oct 6, 2019
Messages
53
Likes
51
Ok, testing time. Just for amusement, I applied the curves from that list to the Audeze LCD2C and Focal Clear, which I own. And OMG NO. What the hell is this? I don't know if the measurements are wrong or this Harman curve is even less ideal than I suspected, but those curves sound awful. They are light years apart from the curves I made on my own for this headphones, even completely the opposite in some cases.

In the LCD2C they decided to put a whopping 10 DB increase at 4060hz! Of course, this is unlistenable and sounds insanely harsh at any volume other than very very low. I actually bought the lcd2c because of their very relaxed and warm sound, and at most I put a general +0.5db upward tilt to bring back some presence. This intense boost in mid treble is totally unjustifiable. The Bass rise is actually similar to what I use (they put +5 db low shelf at 110hz, I use +4 db at 130hz), and the overall curve lowers the very irritating mids that I find to be too boxy, so I also apply a reduction there. But the overall sound signature with their curve is hollow and extremely harsh.

As for the Clear, it's tamer, but still wrong for my ears. They for some reason put a -4db reduction at 200hz, a region I usually boost up because those headphones just lack in the bass guitar region. It gives a really anemic presentation. They also seem to think this headphone doesn't have enough treble energy because they put some boosts there, which makes it even more piercing than usual. I do agree on the general reduction in energy around 3500hz that I also apply myself, but they boost it back up at 4300hz which defeats the purpose. Again, this sounds to me a bit hollow and quite harsh.

Finally, I tried putting on pink noise and listening through the suggested curve of each headphone, to see if it actually results in a similar tonal balance. It's not, and I could even dare to say they sound more dissimilar with the curves then without them.

So my conclusion is that based on this very rough experiment, the Harman curve is just probably not for me. I could see how it would be appealing for short term listening because of its emphasis on presence and impactful lower bass, but for long term it seems to me extremely fatiguing and lacking in warmth.

On the other hand, this is the same sound signature I heard in numerous high-end headphones (maybe with tamer bass though), that I generally despise. So maybe there is a large percentage of the population that do appreciate this sound. Well good for them, I say. But I still don't think every manufacturer should aim for the same response. Otherwise where is the diversity and opportunity for surprise?

Digging this up. I got to play with (2!) Focal Clears this weekend again and after EQing them to oratory1990’s settings, there is this laser like treble peak at ~11khz. I messed around with sweeps, tone generator, and Hotel California to get a sense of where it was at (with a FR plug-in) and added a 0.1 BW -7db peak filter at 11khz. It sounds much more balanced to me now. I’m sure I could dial it in more precisely; this is just a really hasty first take at it but it sounds quite good.

Sonarworks TrueFi also has the same issue and I can’t manually add a PEQ setting there without some external dsp or something.

My theory is that in both cases they cut a bunch of the rest of the frequencies to get the curve to fit, but since they don’t adjust beyond 10k or so, the peak at 11khz becomes ark of the covenant levels, relatively speaking.

My 2 cents.
 

Fluffy

Addicted to Fun and Learning
Joined
Sep 14, 2019
Messages
856
Likes
1,423
@nhunt I had a very similar experience with the Clear. The Innerfidelity raw measurements point to two very sharp peaks, one at around 9khz and one at around 11khz. Because each driver is slightly different, I had to find were exactly do those peaks land in my particular set. After playing with sweeps and music like you, I ended up adding two filters, both with Q of 18 (extremely narrow): -4db at 9,300hz, and -5db at 10,800hz. Turning it on and off during listening gives a subtle effect that is not always noticeable. It really makes a difference only on tracks with accentuated cymbals (particularly hi-hats and rides), and there without the filters each cymbal hit can be piercing and sibilant. Some recordings sound better with the filters turned off, giving more sparkle to an otherwise anemic treble.

Interestingly, the 6khz peak doesn't bother me at all.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
csd-clear-l.png


The 6kHz 'peak' is at the same level as 1kHz so does not stand out... (plot is from the Clear)
The higher frequency peak is higher in amplitude so more bothersome.
These are taken without a Pinna so any resonances that get amplified or subdued are not in play here.
Above 20kHz we see some typical light metalic dome resonances. Way above audible limits but who knows what they might do with 44.1 poorly filtered aliasing.
 
Last edited:
Top Bottom