• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

EQ-ing the 250-ohm Beyerdynamic DT 770 Pro

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
This has been sort of an on-going project for me. So thought I'd start a thread here for sharing and discussing different ideas and approaches on this. Anyone is welcome to participate, and give their own opinions, ideas, suggestions, etc. on the subject.

Since I'm currently using the 250-ohm version, most of my ideas/suggestions here will be for that particular model. Some of the basic principles discussed here should also be applicable to other headphones though, including possibly some of the other versions of the DT-770. The 32 and 80-ohm versions of the DT-770 have a slightly different frequency response than the 250-ohm DT-770 though. So for the best results, you'd probably want to find some more accurate and reliable FR data for EQ-ing those particular headphones than simply using the 250-ohm data I'll mostly be using and discussing in my examples below. (One possible resource for that might be Crinacle's graphing tool, which I believe has FR data for both the 80 and 250-ohm versions. I think all the diffuse field compensated versions of his graphs may be behind a paywall though.)
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
What I'll generally start with for my EQ-ing is the diffuse field (rather than raw) frequency response of the DT-770, which looks like this...

Diffuse Field Frequency Response of 250-ohm DT-770 with Fresh & Worn Earpads:

DT770NEWWORN.jpg


As noted above, I believe these plots are for the 250-ohm version. And the red and blue curves don't represent the right and left channels in this case. Instead they represent the DT-770's response with new (red) and worn (blue) earpads.

Since Oratory has plots for both, I also compute and use a frequency response curve that is based on an average of the new and worn curves, which looks somethin like this...

DT770AVG.jpg


And I'll also use the same curve above with a bit of additional smoothing applied to the bass region. Because I think it's possible that Oratory's plots may slightly exaggerate the audible fluctuations in that area a bit, based on comparisons to DT-770 plots on some other sites. And I don't want to over-correct the response in that range too much.

DT770AVGSMOOTHED.jpg


I'll also use Ora's original plots with new and worn pads as well though, as sort of the extremes of the DT-770's potential response range, which can also be handy. So I'll use basically all of the above. And have about 4 or 5 different references just for the headphone's starting FR.

EQ-ing the DT-770 isn't the easiest, so the more info there is to start with, the better in my opinion.

The original diffuse field plot of the DT-770's fresh & worn response curves shown above is from Oratory1990's graphing tool btw.
 
Last edited:

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
10,509
Likes
23,281
Location
The Neitherlands
Best to EQ the lows to taste and circumstances (seal) anyway.
Another question is how 'worn' the pads were.
In the DT990 case the pads were so worn that the drivers rest against the ear.
One would have replaced the pads by then for comfort reasons alone.
Add to that not all DT770 are created equal and there is unit to unit variation depending on how old the DT770 is.

The message here is: There is no 'correct' or 'average' EQ. Just tune the bass to your needs and leave >200Hz EQ as it is.
 
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
Welcome aboard solderdude. And thank you for the above insights.

I have to disagree with you a bit about there being no correct or average response, because in my experience, there can be response curves which are more (or less) neutral or transparent sounding. So that's sort of what I strive for. At least as a starting point in my EQ-ing. And what I'll actually use as my reference or guide for this are the response curves of other headphones with a more neutral sound.

The DT-770 is probably one of the more challenging headphones to EQ well though, because its sound can change so much depending on the condition and wear of its pads. So I'd definitely agree with you on that. And may also have a couple ideas on how to handle that.
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
Just tune the bass to your needs and leave >200Hz EQ as it is.

I like your idea of just tuning the bass as well, to balance out some of the brightness in the treble a little better. That could also be an effective way of dealing with the different amounts of wear on the pads, since the effects of that seem to be concentrated to a large extent in the lower frequencies.

This might give the headphones a more U-shaped sound than they already have. Which might be good for listening at lower volume levels.
 
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
One of the advantages to using the response of other headphones as a guide (as opposed to a more generic target response curve, like the Harman curve) is that you can use whatever headphones that you think would probably sound the best. And also experiment with various different sound signatures.

You can even mix and match the sound signatures if you want, and use one set of headphones for the response in the bass or lower frequencies, and another set for the response of the treble or high frequencies. Which is also something I will do.

In the graphs above, I've normalized all of the FR curves to a point in the center of the frequency range, at approximately 633 Hz. This is something you can do fairly easily with Ora's graphing tool. And it's why the gridline for 0 dB passes through all of the curves at that frequency on the graphs.

NORMALIZATION.jpg


I do this for a couple different reasons. But one of those reasons is so I can use a different target or set of headphones as a reference for the response of the lower frequencies below 633 Hz, and another set for the response in the higher frequencies above 633 Hz. And experiment with some different sound signatures for both halves of the frequency range. And have an easy cross-over reference point for aligning all the curves.

I'm gettin a little bit ahead of myself on some of this. But there are some distinct advantages in breaking down a headphone's response in this way, for the kind of approach that I use. Because there are relatively few headphones which have an ideal response (imo) in both the higher and lower frequencies, that could be used effectively as a reference for the whole response range. So dividing up the duties more along the above lines can potentially make things a little easier.
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
I use a variety of criteria to select the headphones in Ora's database to use as my EQ references, including the sound power of neutral loudspeakers, opinions of other users/reviewers, a slope between -1 to -1.5 dB/ocatve, the Harman curve with diffuse field compensation, and also my own listening tests and experiences with a fair number of different HPs. And have pretty much narrowed down the list of headphones to use as a starting point for the bass response to some of the following:

BESTBASS.jpg


AKG K371
Apple APM
Audeze Sine (with Cipher cable)
Beyerdynamic Custom One Pro Plus (with velour pads & bass setting 3)
Neumann NDH-20
Onkyo A800
Philips Fidelio X2HR
PSB M4U 8 (with ANC on)
Sennheiser HD598CS
Sony MDR-7506


I think the shape I'm sort of looking for looks kind of like this though. With sort of a broad hump in the bass between about 20 and 200 Hz. Ignore the stringers on these...

BASS.jpg


Since I'm only looking at the bass levels, I've greyed out the area above 633 Hz for the time being.

The above graph with my slightly smoothed and averaged diffuse field plot of the DT-770 overlaid for comparison...

DT770BASS.jpg


Hopefully you can begin to see a little better where a few of the potential weak points are in the DT-770's lower frequencies on this graph The 250-ohm DT-770 is pretty well extended in the sub-bass frequencies. But it lacks a little warmth and fullness in some parts of the upper and mid-bass (which isn't easily correctable with just a bass shelf), that would probably help to better balance out some of its brightness in the upper frequencies.
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
index.php


The headphone with rather bright peaks in the low treble on the above graph is the Sony MDR-7506 btw. And the headphone with the fairly pronounced dip in the upper-mids is the Neumann NDH-20.

Both of these headphones have a pretty good FR in the lower frequencies though. Which is really all that matters in this case. (The Sony also has a slight dip though at around 225 Hz in the upper bass, not unlike my DT-770.)

On DF plots like the one above, headphones with bright spots or peaks in the low treble that excurse appreciably above the level in the midrange between 500-1000 Hz will tend to be noticeably on the brighter side.
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
The same 10 headphones above, compared to the average estimated sound power response of three well-extended neutral-ish loudspeakers (the Dutch & Dutch 8C, Infinity Intermezzo 4.1T and KEF Reference 5)....

Diffuse Field Response of 10 Headphones with Good Bass Extension Compared to Average Sound Power of 3 Well-Extended Loudspeakers (shown in black):

3SPEAKERAVERAGE.jpg


Sound power is a pretty good model (imho) for the diffuse field response of neutral headphones, particularly in the bass and midrange.

It's only useful as a very rough guide for the shape in the upper frequencies on Oratory's plots though, because there are some pinna-related effects that also have be taken into account there as well. And the only way to really do that is by looking at the actual measured responses of some neutral headphones in that area (which is something I'm still working on).

The three speakers which are represented by the black curve above probably also have better than average cross-over designs. So unlike many less-expensive speakers, there is not much of a dip or directivity error in their off-axis/sound power/in-room response at around 2 to 2.5 kHz in the upper mids, where the tweeter and midrange drivers cross over. With larger samplings of speakers, the directivity errors (and attending dip in sound power response) will probably start to become more visible around there.

The pinnae that Oratory uses for his in-ear measurements also seem to cause a small dip in approximately that same 2 to 2.5 kHz range as well btw. So most neutral headphones should have some type of a dip there, whether they're attempting to approximate the cross-over errors on speakers, or not.

I can also assure you that the DT-770, and also other headphones would sound very wrong! with Ora's diffuse field response curves smoothed out to the same degree in the treble as the speaker curve shown above. So I am posting this primarily for its relevance to the response in lower the frequencies.
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
DF Response of 10 Headphones with Good Bass Extension Compared to Average Sound Power of 5 Well-Extended Speakers:

5SPEAKERAVERAGE.jpg


Speakers included in the black average sound power curve above:

Dutch & Dutch 8C
Infinity Intermezzo 4.1T
KEF Reference 5
Infinity Prelude MTS
Mesanovic RTM10

The Infinity Prelude MTS and Mesanovic RTM10 are also pretty well-extended in the bass. But not as well as the 3 speakers in the first comparison above. So the average response is a bit lower and less extended in the sub-bass with these also included.

Without a separate sub-woofer, there are relatively few speakers that extend as well into the sub-bass frequencies as the first three speakers above. So the more speakers that are included in a sampling, the less extension there is likely to be in the sub-bass.

I may want to expand the range of bass responses a little on my list of headphones to better reflect this as well. For the DT-770 though, I'm probably going to want to stay on the higher side in the bass, because it's already well-extended there to begin with. And because it should help to balance out some of the brightness in the treble.
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
BESTBASSANDTREBLE.jpg


I'm a bit further along in the process of trying to define a treble response. But this is approximately where I was at yesterday, when I was looking at the treble responses of about 20 or so different headphones. I spliced together the 20 headphones for the treble with the 10 best headphones for the bass on the above graph just to get sort of a general picture of how things might look.

Some particular features that I look for in the treble response are a continuation of the general downward slope or trend of the curve from the bass. A dip in the 2 to 2.5 kHz range. And sort of a broad hump between 2 and 20 kHz. The broad hump (or series of smaller humps) should rise above the curve's overall downward slope or trajectory by at least a few dBs in the low treble. But probably should not exceed the amplitude levels in midrange around 500 to 1000 Hz. (Or only by a small amount.) And there should also be some fairly pronounced notches in the treble at around 9-10 kHz, 15 kHz, and 20-21 kHz.

Although the treble is still rather poorly defined, I think you can see the beginnings of some of these features on the above plot. I'll probably also err a bit on the higher side in the treble, since that is the natural sound signature of the DT-770's. And I want the final response of the headphones to be maybe a bit U-shaped.

The above plot with my average DT-770 FR curve (without any EQ correction) overlaid...

BESTBASSANDTREBLE2.jpg
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
TREBLETOP12.jpg


I think these would probably be my top 12 for the response in the treble.

1More Triple Driver (perfect seal)
Beyerdynamic Custom One Pro Plus (velour pads, bass setting 3)
Focal Radiance
HFM HE400i (2020 pads)
HFM HE4XX
HFM Sundara (2020 pads)
Monoprice M1570 (velour pads)
Onkyo A800
PSB M4U 8 (ANC on)
Sennheiser HD800S
Sennheiser HE-1
Verum 1 (leather pads)


The overall shape is pretty similar to the average response of the 20 headphones in the last post above. So not sure I've really narrowed things down that much more.

The Sennheiser HE-1 and HFM Sundara with 2020 pads are probably two of the best on the above list imo. Probably followed by the Onkyo A800 and HFM HE4XX. The headphone that dips down too much at ~7k in the treble is the Verum 1 with leather pads btw.
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
One quick and dirty solution I use for EQ-ing the DT-770 (while I continue to work on the above) is just a simple slope correction.

The slope can be adjusted in Equalizer APO's Configuration Editor, using a simple 2-point Variable Graphic EQ filter. And moving the point at 20000 Hz in the treble up or down in 1 dB steps, as described here, using the up/down arrow keys on my keyboard. The screengrabs below show the effects of some different slopes on the raw (rather than DF) response of my average DT-770 curve...

I generally prefer the sound somewhere around the middle of the range below. But it depends alot on the content. The sound of the DT-770 with no slope is too bright imo though. And the sound with a -10 dB slope is too bassy.

NO SLOPE:

NOSLOPE.jpg


-2 dB SLOPE:

MINUS2DB.jpg


-4 dB SLOPE:

MINUS4DB.jpg


-6 dB SLOPE:

MINUS6DB.jpg


-8 dB SLOPE:

MINUS8DB.jpg


-10 dB SLOPE:

MINUS10DB.jpg
 
Last edited:

temps

Active Member
Joined
Mar 9, 2020
Messages
177
Likes
234
imo the Sonarworks average calibration for the DT 770 250 ohm sounds excellent. I run that at about 60% wet signal and it's just right. Turn it up to 100% EQed and it starts to sound very bizarre, flat and distorted in a bad way.

On the hand, the Sienna profile for the DT 770 250 ohm sounds terrible.
 
  • Like
Reactions: ADU
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
imo the Sonarworks average calibration for the DT 770 250 ohm sounds excellent. I run that at about 60% wet signal and it's just right. Turn it up to 100% EQed and it starts to sound very bizarre, flat and distorted in a bad way.

Interesting. Haven't tried either of the above, so thanks for giving some opinions on this, temps.
 
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
I am still in the process of doing/completing a rather long and arduous cross-country move. So that's why I haven't been as active on the headphone forums here lately. And haven't posted any new plots or graphs in some time. I hope to be able to get back to doing some more on that pretty soon though.

I sold my AudioTechnica M50x btw. So the DT-770 is my only HP right now.
 
Last edited:

temps

Active Member
Joined
Mar 9, 2020
Messages
177
Likes
234
Interesting. Haven't tried either of the above, so thanks for giving some opinions on this, temps.
They're both interesting programs, but PC only. Sonarworks is mostly just to get headphones to a flat, reference sound - not sure what curve they use but when you turn it all the way up, it's definitely not enjoyable to listen to. You can run either "average" profiles based on every measurement they've done of a headphone so far, or you can get an individual calibration done. I got one of those by buying a pair of DT 880s from them.

The individual calibration is quite a bit different than the average calibration. Green line is the EQ being applied, blue line is the original response.
Average for 880s
1642110891869.png


My individual calibration for 880s
1642110909837.png


For reference this is the Sonarworks 770 measurement. I wonder though, how useful this is... for a headphone as popular and long-running as the 770, there must be so many different versions it may be very difficult to EQ without doing a good number of measurements on your own.
1642110929678.png
 
  • Like
Reactions: ADU
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
Very interesting, temps! Thank you for posting this.

The EQ curves on the above plots appear to be essentially the inverse of the FR plots on the above graphs, except in the very low frequencies in the sub-bass, where the EQ curve rolls off more. What this suggests to me is that the FR plots shown above have already been corrected and compensated with Sonarworks' proprietary target headphone response curve... whatever that might be.

Without some better info about the graphing system they are using and their target compensation curve, it's somewhat difficult to make heads or tails of their FR plots. Probably the best way to utilize those, is to compare them to the Sonarworks FR plots of other HPs.

I can see from the three graphs that you've posted above, for example, that the DT-770 has a somewhat similar FR to the DT 880 in the midrange and to some extent also in the higher frequencies. Where the two headphones seem to differ in their response is mainly in the lower frequencies. The DT-770 is more withdrawn than the DT-880 in the 200-250 Hz range. But better extended in the sub-bass (which makes perfect sense, since the DT-770 is a closed dynamic headphone, and the DT-880 is a partially open dynamic headphone).

I have a few thoughts on Sonarworks' EQ curves as well. But will save most those for another post. In a nutshell though, there are some similarities (and also differences) to a few of Jaakko's AutoEQ curves for the above HPs, which were based on the Harman headphone target with a slightly reduced bass response.
 
Last edited:
OP
A

ADU

Addicted to Fun and Learning
Joined
Jul 29, 2021
Messages
626
Likes
539
I see, btw, from looking at the ASR headphone and measurements index that there are still no official ASR reviews or graphs for any version of the DT-770 yet, which is sort of too bad.

I have also not had a chance to try or listen to any of Beyer's newest headphones, such as the DT 700 Pro X, and DT 900 Pro X. And it would be interesting to see some more plots of those as well, by ASR or others.
 
Top Bottom