And not a bad choice either!
IMO the 660S & S2 are underrated.
hah they have very similar graphs indeed. We need to find someone who has them both.
I’ve tried them both but at totally different venues a few months apart, my notes on them are somewhat similar I suppose:
660S2:
There’s good bass, it doesn’t feel amazingly sharp and tactile for pop (but still pretty good), not much sub bass, but it feels pretty balanced throughout the FR. Treble isn’t that bright. It feels a bit roomy slightly boomy. Sounds good for vocals.
After trying a bunch of other headphones then going back to the 660 S2 again: immediately it sounds quite warm and pleasant, good vocals, sounds like you’re in an echoey room but in a nice way. Clampy headband, bass sounds a bit flat and flabby, got a warm grand feel, soundstage is limiting though, sub bass isn’t there but mid bass is ok. Not really better than 600 EQed. Listening to the Hd660s2 right after the HD650 (no eq): tighter bass is felt immediately, almost impressive for “Force projection” song, sounds a bit too roomy boomy for “Dogs chasing cars”. Not enough sub bass. Good treble. Really good for vocals.
Focal Utopia: It has a very grand sound, slightly boomy bass, not quite enough sub bass, recessed treble, not quite there in terms of ultimate sub bass clarity. Mid range/Vocals feel slightly recessed. Very good sound stage. Good impact and tactility but not the most detailed or clean feeling even though the impacts are there. Ever slightly rounded edges for sharp notes.
I guess the main difference from my notes is the soundstage. I rated the Utopia very good, and the 660S2 limited. And that’s not something you could affect much with EQ.
1 Like
Crin has a good rep. Here are some more 8XX and 800S graphs by another reputable grapher who uses a GRAS system.
Baaed on these, there are a few fairly noticeable differences in tonal balance.
1 Like
Howdy, Spearthrower_Owl.
For more on this, please see some of my thoughts in the 2nd post here…
1 Like
Not familiar with this grapher. That does not mean they are not good graphs though. The plots look reasonably good.
There are two plots of the Utopia though in Oratory’s graphs…
He also has a graph of the HD660S that could be used for comparison…
There’s a list of all Ora’s graphs here btw…
https://www.reddit.com/r/oratory1990/wiki/index/list_of_presets
Resolve has also posted raw graphs of the 2022 Utopia and HD660S measured on his HBK 5128 rig. The graph of the 660S looks like its has more smoothing though. 
Muddy you say. He he he.
Drop released the 6XX as a members-only limited edition of the 650 in midnight blue, but used exactly the same drivers as the 650. Buyers got a functional 650 for a huge discount. People also wanted a budget edition of the 800 / 800 S, but Sennheiser didn’t want to cut its prices. So, the 8XX drivers were literally hole-punched to limit its potential. They sold for “less” than the 800 products, but were intentionally crippled.
Prior to the release of the 8XX, Drop posted the number sold for each product, but they started hiding sales figures with the 8XX. And they allowed returns. I’m not sure why people would spend close to $1K on an 8XX when the 800 S can be had for around $1,400.
1 Like
OK, thanks for the reply. Yeah, older listeners should not be trying to use EQ to compensate for diminished sensitivity at 16kHz! Everything over 12kHz is quite likely just functionally inaudible once you’re over 50 years old, and should just be thought of wistfully, if at all.
Choosing a headphone for a bit more ear gain, or using a little EQ, at a healthy listening volume, to compensate for some mild noise-induced hearing loss, on the other hand – that’s going to be fine. But also probably unnecessary.
So I finally got a bit of time and an eqMac PEQ to play with some of these tips. I started with my Mega5EST IEM and tried to adjust this 1800-2200 and ear gain region to see how impacts the soundstage. I’ve tried different widths and levels of the scoop. I’ve used the “Hotel California (MTV live)” as the test track for this exercise.
The only effect I have noticed is that if I reduce this 2kHz area, the vocals start to sound a bit less present and they feel slightly more distant and quieter. The rest of the soundstage with the guitars, the drums and the crowd just stay where and how they were.
The inverse of it, with a hump at 2 kHz makes the vocals dominate the stage and at the same time they sound nasally and boxy, and the other instruments get muffled and faded away to the background. Nobody would use it for their music this way, of course.
Then, adding more energy to the 6-7.5kHz region just makes “s”, “t” and hi-hat sounds more sibilant. If I add more treble, it feels just like more treble and harshness. And things like 1-2 dB peak filters with a narrow Q (like 3-4) I don’t notice any tangible difference with or without.
Overall, most of the changes I tried to make to my Mega5EST don’t change my perception of sound stage or detail. Maybe these have to be very fine changes to very specific frequencies in very specific combinations, but I can’t tell which ones without looking at a raw measurement done on my head. 
2 Likes
Great tests and thanks for the detailed breakdown of what you tried and heard!
I also haven’t had much success changing soundstage with just eq, iems or headphones. Would love to hear more from others’ experiences.
Detail however is a different story, I bought several cheap headphones simply to test my EQ method and skills, and it’s pretty incredible how much of an improvement could be had for tonality and detail. If you don’t feel much changes in detail that’s a good sign that the original tuning is decent.
That said of course there are physical limits to detail that you could add via EQ. I find that headphones which have some kind of physical damping material in the way (on ear, or iems with small tip holes, or closed back with echoey character in the cup) seem to have lower potential in that regard.
1 Like
I can’t really tell if you’re serious because of the wink, but the whole reason the content team kvetches about non-HRTF-related HpTF variation (or rHpTF—relative HpTF—variation) is because while they might look similar when measured on one head, they’ll look meaningfully less similar when brought to another head, especially over 2 kHz.
Blue is HD 660S2, green is Utopia, red is the HD 660S2 EQed to adhere to Utopia. All measured on GRAS 43AG-7 with KB500x anthropometric pinnae
When you take this to a different head, like the 5128, the changes you made aren’t guaranteed to bring the HD 660S2 closer to the Utopia’s response on this head.
This is true of headphones on human heads as well. EQing one headphone’s measured response to another headphone’s measured response on one head is nearly guaranteed to not produce a congruent sounding result unless the measured frequency responses you are EQing were taken on your own head.
4 Likes
The problem here is that the HD 800S and HD 8XX have different acoustic impedances due to the latter having a significantly more closed-in front volume due to Drop utilizing tape to damp the interior.
Unfortunately this effects both the leak tolerance—where the HD 800S is almost certainly more tolerant of leaks in the bass, meaning it’ll roll off less when exposed to small leaks—as well as the overall positional variation. The effects of this can be seen in the attached GIF.
4 Likes
I think you make some fair points here, listen_r. This is one of the reasons I prefer to use raw measurements made on the 5128, for my own EQ-ing. And other raw measurements made on rigs with a dummy head, rather than a flat plate. (I thought Ora used a head for his GRAS measurements. But maybe that’s not right?)
It would also be interesting to compare the actual EQ/difference curves for the different rigs, like the one I posted above for the 6XX and 800S, to get a better idea of the similarities and/or differences between how the rigs behave.
Instead of targeting a single headphone in my EQs, btw, I usually try to target an average of at least two or more headphones that come close to the response I want. I know you and Resolve have some concerns about this as well.
There are no guarantees with any of this from my perespective. It’s just somethin interesting to try.
Do you have any over-ear headphones, MrSehrKalt?
I have a Focal Clear OG, and a Beyer DT770 but the latter is modded and won’t measure anywhere near the available graphs.
I plan playing with EQing the Clear next to observe the effects. Right now I’m just enjoying it with listen_r’s preset too much 
Which to my ears among other things adds more depth to the room information and makes the treble sound more smooth and “detailed”.
1 Like
Yes, but that variation from our perspective is more or less random, so moving from one head to another, the two responses might look less similar, or they might look more similar, yes? In any case, of course you’re right, to do this properly and in earnest for your own benefit, you’d want to adjust for HRTF.
But yeah, I was winking because while I think this is a very fun and interesting topic, I don’t think the proposed aim can be achieved. Not because I think there is some special magic other than sound waves! I don’t. But I’m skeptical that how we measure, and EQ to measure, fully captures reality.
1 Like
Here are the observations from testing these tweaks with my over-ears (Clear OG and DT770 paper mod).
Adding a dip to ~2kHz area, or even wider band 1800-4500 kHz. What I’ve noticed it does to my perception of the sound stage overall is that it muffles part of the room reflects so that the sound that is is positioned in the center of the mix sounds a bit further away, compared to no-EQ in this region. Disabling this filter makes the center image more present and “in your face”. It doesn’t affect overall stage width at all, at least to my perception. And I tried both subtle and quite drastic level values like -6dB, and Q values from 1 to 4.
The other audible thing the wide band dip in the ear gain area does to my hearing is that it smooths the attack of the instruments overall.
The “detail” tweak for 6-7.5kHz is less obvious. For my Clear adding a bump there doesn’t make a meaningful difference, and surely doesn’t make it sound more “detailed”. And my DT770 is modded to pull down the Mt. Beyer with a bulldozer. So adding an EQ peak at 6750 Hz brings some of that sharpness back, as expected.
I also tried the “punch” tweak with a 100Hz bump on all 3 headphones. Didn’t notice any change in that characteristic. Partially because these are all quite punchy headphones already. Partially because I tend to agree with Resolve that some of the essential components to the feeling of “punch” reside in the midrange and maybe even lower treble.
Did any of these tweaks sound better to me than the original sound signature of these headphones? Not really.
But it’s just me and just these couple of tweaks. I’m generally not very susceptible to the spaciousness effect in the headphones. E.g. when I was auditioning a Clear MG, an Arya, and an HD800S, the Arya and the HD800S sounded wider to me than the Clear, but not by a mile. And I couldn’t get the hype over the HD800S’ crazy wide soundstage, it didn’t sound much wider or more precise than Arya to me.
Regarding the feeling of the “detail”, I’m still quite convinced that the devil is in the details. Adhering to a smoothed target is one thing. But how we perceive the attack of certain instruments, decay, sharpness, smoothness, etc. can be in very narrow frequency bands and their harmonics. And here is where the finer details of the FR and headphone’s design inducing that FR come to play. E.g., even though Mega5EST measures much better in the treble area than the Clear OG, I still prefer the treble of the Clear over the Mega5EST for the sense of “detail” and naturalness. If I EQ them both to even out their treble to fit within the preference bounds, it still doesn’t change my pull towards the Clear’s sound.
Curious to hear how other people perceive alterations in different frequency bands with regards to the subjective sound qualities.
2 Likes
Clear OG is the original Clear, I take it. And not the MG.
Can’t find an OG Clear plot made on the HBK 5128 here. Jude has one in his Clear MG review though.
Like many of the early Focals, the OG Clear has alot of energy in the upper mids around 1.3 kHz, or so. Based on Jude’s graph the OG Clear also lacks some energy in the very high frequencies (mostly above the threshold of my hearing).
The early Focals are also a bit “peaky” in the treble. So I can understand your interest in trying to smooth some of that out. That may be somewhat difficult though. And not sure how much benefit there would be.
When I get a bit further along with some of my EQ-ing, maybe I can post some more specific tweaks to compare with listen_r’s. When comparing this with something like the HD800S though, what immediately leaps out is the difference in the upper mids. The HD800S is much more withdrawn there than the OG Clear. (Neither headphone is really neutral there imo.)
I try not to use the current Headphonesdotcom pref curves btw, because they conflict somewhat with the DF+SP model I use.
I posted a few requests in Listen_r’s thread a while ago in my interest about whether you could EQ one set of headphones to sound like another (e.g. OG Clear to Utopia, Arya Stealth to Susvara). As has been discussed, I understand there are some challenges with those ones, but I’d still be keen to try if anyone’s game to give them a go.
I have a further request that I think might be more achievable: could anyone present an EQ profile that can make my 6xx sound like a HD 600?
I’ve always suspected I might like the 600s more than the 6xx, which I find slightly veiled and too dark compared to my Hifimans. If I recall correctly, Blaine actually mentioned EQing a 6xx instead of buying a set of 600s in the black Friday sales live stream. Since I understand there’s not much difference between them, hopefully this one’s more doable.
There’s hardly enough differences between the two to eq I think. The differences are within the margin of error in manufacturing tolerances, headphone positioning and pad wear.
You could lower the bass by 3db I guess.
1 Like
Perhaps. I think all of MJ’s challenges should be doable though with measurements here and elsewhere.
They will be subject to limitations inherent to EQ though. Imo, it is virtually impossible to EQ the FR of one headphone to match another headphone without any EQ applied, without some loss in the signal or sound quality occurring. How audible that loss would be is debatable though.
1 Like