Behringer DEQ2496 Equalizer Review

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by purr1n, Aug 25, 2020.

  1. crenca

    crenca Friend

    Pyrate
    Joined:
    May 26, 2017
    Likes Received:
    3,829
    Trophy Points:
    113
    Location:
    Southern New Mexico
    This is a good idea, assuming it is not already a thread and I have missed it.

    I don't do anything in the bass (maybe I am not a bass head? ;) ), but I do shave 2 db off that 1.2khz peak with a relatively low Q (I have .9 right now) with the Utopia. I must like Focal FR because I don't do anything with the Clear. ZMF Verite Closed I fill in that 3-4 k "BBC dip", but I hear ringing/distortion with anything more than 2 or 3 db. I am also constantly messing with low and high shelfs as well, never being satisfied. Back when I listened to the Elear I used EQ more extensively, and unlike my experience with the ZMF's the Elear "took" it pretty well.

    EQ does not change a HP fundamentally, but used appropriately and with limited expectations it can help get you where you want to go.

    Remember to use the global DB slider in Roon so that you don't get into + gain overall to avoid digital clipping.
     
  2. gixxerwimp

    gixxerwimp Professional tricycle rider

    Pyrate
    Joined:
    Sep 30, 2015
    Likes Received:
    5,773
    Trophy Points:
    113
    Location:
    small island claimed by China
    Yup, been there, done that. If you scroll up a bit from my post that was linked at the top of the OP, you'll see what I was using before I got the Behringer.

    I'm not a Linux guy, so getting it all working properly was hard enough. I looked into it, but it didn't seem worth the pain.

    I usually find an FR measurement of the phone in question (made by someone I have a good reference for) and start playing around with the opposite of peaks and valleys to see if it brings me closer to my preferences. Other people's settings can be all over the place. For example, I tried Tyll's HD800 settings and, though it was a useful and educational process, my personal end result was almost totally different. Something else to remember is that FR troughs sometimes correspond to CSD ridges ...
     
  3. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Ugh yuck. That approach, using overlapping PEQ, and attempting to surgically address each and every peak on a headphone is a bad idea for many reasons.
    • There is no guarantee that your pinna and concha will be shaped like the artificial ears on that GRAS. As a result, you may actually be introducing more error.
    • Related to the above, while overall slope matters, what most important is eliminate broader dips or peaks as localized "trends".
    • As already mentioned, Harmon curve is bad. That stuff is based on Harmon curve.
    • Without CSDs, we don't know if dips may be peaks in disguise. EQ'ing dips may result in a much worse result. Also, our ears are much less sensitive to dips.
    @Carlos CPA: The best way is to use your ears, using the SBAF compensated EARS, FPC, and FAC (for high frequencies) measurements as a guide. You can use this data and visualization tools desribed here: https://www.superbestaudiofriends.o...ithub-headphone-frequency-response-data.9679/

    TBH, despite the examples shown above, I prefer a much lighter approach. The thing is, them headphones are like people and are always going to want to act a certain way.

    Focal Utopia could use these settings. There could be an entire discussion around this.
    1. A small shelf +3db @20Hz starting gently from 80-100Hz rising as we go down to 20Hz. (depends)
    2. A -3db dip @5kHz with a Q of 2
    The basis of #2 is the Free-Air Sim measurement:
    upload_2020-8-28_8-39-8.png

    The basis of #1 is from the FPC measurement:
    upload_2020-8-28_8-41-21.png

    Yellow line above is more in line with the B&K listening position in room target (black line below). The blue line above is more in line with the 1db/octave slope target (bassier version of B&K, red line below)

    [​IMG]

    EQ'ing is an art. It's an art for speakers because the moment we move our head or move to the next seat over, the FR is completely different. The difficulties for EQ'ing headphones are even greater per points above. I could just as easily write a computer program that went through all of my measurements and suggested an EQ for every headphone. Doing so would be irresponsible however.
     
    Last edited: Aug 28, 2020
  4. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    IIRC the AutoEQ project (which these files are from) tries to smooth out HF peaks a lot in order to avoid overcorrection. You might still say it's not smoothed enough, but as always it's a matter of compromises. I haven't tried his EQ targets with different headphones, yet so I can't really say how well it works. I'm sure it's not too bad. A lot of effort was put into getting it right.

    While I also don't fully agree with what oratory1990 did (especially regarding using the Harman target with its bass shelf, which then necessitates an IMHO forward midrange tuning to not make it sound too warm), but I find you really have to appreciate it. He took measurements from many different sources, applied correction curves for each and uploaded correction files with multiple different target curves. I think it's an extremely valuable resource and I'm sure it will serve as a good starting point for EQing. I don't want to imagine how many hours he spent tweaking every aspect of it, it's just crazy. I really admire what he did. Also don't think he's a hardcore objectivist based on that project.

    Also it seems you've been using the Utopia for some more time now. Do you find that you get used to the 1kHz peak? When I had quick listens years ago that was my main gripe with it, tonality-wise. I also find it somewhat bassy tbh, but I know our preferences diverge in that regard.
     
  5. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Sorry, I don't appreciate it; but what can I say? How 'bout you try these corrections out and report back? Maybe you will understand better why I don't care for this Auto-EQ effort. What's the point of "brilliant effort" when the results are complete shit? Anyone proposing the EQ curves below should be figuratively shot, drawn, and quartered.

    HD650.
    upload_2020-8-28_11-16-17.png

    Utopia
    upload_2020-8-28_11-20-16.png

    And yes, the smoothing is hardly sufficient. In the case of Focal cans, the highs are quite different from GRAS, EARS, FPC, FAC, RANDY (I have GRAS measurements too - I've used one). Trying to correct all those wiggles from 2kHz up and more likely to do more harm than good. Seriously, does the HD650 need a small high Q boosts at 4kHz, 6kHz, and 11kHz? Or does the Utopia need strategic high Q dips at 1.7kHz, 6kHz, and 8.2kHz?

    All this tells me is that it's the work of someone who is good at numbers but has absolutely zero experience with sound.

    I've never had an issue with a 1.3kHz peaks, which is found on various headphones. Let's say it is a coloration that I find acceptable. I generally don't find Utopia bassy and don't tend to run it on bassier amps as most people prefer. The 45 Studio I'm running it from has feedback and is maybe less than 2-ohms ZOut.

    --

    If it makes you feel happier, I'll write my own Auto-EQ routine based on a variety of EARS, FPC, and FAC data. I doubt you'd show as much appreciation for my effort.
     
    Last edited: Aug 28, 2020
  6. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Disregarding the bass part, what's so bad about those EQ curves? Making the HD650 somewhat brighter seems fine to me and the Utopia curve doesn't look too bad to me, either. Notch at 1.3kHz, bump at 2kHz - seems fine to me. Notch at 6kHz and lowering the upper treble seem like sensible choices to me, aswell.
    It seems he uploaded many files with a somewhat modified version of my IF target aswell, something which I have commented on before. As mentioned, I find Harman makes the midrange somewhat forward to compensate for the bass, which could make some sense.

    Regarding Utopia bass it may also be a matter of distortion or texture. Idk, I preferred HD800 bass.

    I'd try them out, but I have neither phones here. Would have to revert both my HD600 and HD800 back to stock in order to try those curves, which is not worth it.
     
  7. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    It's not somewhat, it's a f'ton brigher. Stop being obtuse and unnecessarily contrarian. We have not already learned the lesson from the Drop iFi CAN HD650 EQ special edition?
    upload_2020-8-28_11-38-50.png

    It's the high Q. A high Q notch won't do much to address the obvious boost in the presence region of the Utopias.
    And even then, which one is correct?
    upload_2020-8-28_11-43-48.png
    upload_2020-8-28_11-45-57.png
    upload_2020-8-28_11-44-8.png
    upload_2020-8-28_12-10-33.png


    Of course. But is this what we want? Stop making arguments for the sake of making a good point. We want good sound, not good points. And even then, if we have V or U-shaped, I'd rather the high boost to take effect later, much later, past the upper-mids for sure. I really don't want my "beautiful (Asian) female vocal" (or Taylor Swift early records) any more "enhanced".

    You do realize that these suggested tuning could actually kill @rhythmdevils?
     
    Last edited: Aug 28, 2020
  8. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    I didn't mean to be contrarian*. You know I still think your changstar-type V1 measurements are generally the most valuable for the range beyond 5kHz. I've expressed before that I think the V2 creatology foam coupler was a step backwards in many ways. It's why I tried to find a to that combine the properties of both some time ago (which seems to work well for open headphones).

    I'd rather ask you this: Do you think the difference between the HD650 and Utopia targets is accurate to your ears? As mentioned there are different absolute targets which we can apply instead of Harman. But the difference is the important bit here.
    Considering how the treble of the Utopia is somewhat boosted in this target (which I may not even disagree with), it's really no wonder you'd find the HD650 target bright.

    I think using the 200Hz dip as a reference for the HD650 data is cherry-picking here. You know the bass shelf they used at Harman is going to result in a bit of a dip there. Using 1kHz as a reference puts it more at a 4dB treble boost, which honestly seems quite fine to me. Maybe 4dB is still a bit much, not sure.

    Of course it's been a while since I heard them, but what I heard sounded more spotlit than a broad boost. The sharp 6kHz peak in the measurements I uploaded actually seems somewhat accurate to me. But it's been a while since I heard one.


    *I said I didn't like the target and that seems to be your issue. But changing the EQ target is easy, so the main part of his effort is still valid.
     
    Last edited: Aug 28, 2020
  9. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I don't think the difference leads to something which is accurate. I just uploaded Tyll's measurements for Utopia (flipped for EQ). Even if I contoured Tyll and Oratory's compensations to be comparable with my FPC and FAC (V1), we see that the sharp high-Q dips and peaks for each are all over the place (keeping in mind what when it comes to high-Q adjustments, small shifts in frequency are huge!)*

    A good approach may be to glue the V2 (FAC) with V1 (FPC). But even this introduces some problems with V2's lack of damping which may emphasize peaks with depending upon the Qts of the driver.

    * Which goes back my point that surgical EQ could do more harm than good.
     
    Last edited: Aug 28, 2020
  10. Psalmanazar

    Psalmanazar Most improved member; A+

    Pyrate Slaytanic Cliff Clavin
    Joined:
    Sep 27, 2015
    Likes Received:
    5,345
    Trophy Points:
    113
    The bass distortion will be massive on that eq. The HD 650 already has more distortion than a driven Neve 1073 basketball kick. The 650 is a rubber bouncy ball kick drum. Why make it worse?
     
  11. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    This is why I don't boost low bass, avoid bass shelves, or at least do so with caution. With Utopia and Focal cans, one needs to be particularly careful. Those things are seriously open baffle and have a ton of excursion at lower frequencies. Very easy to make those drivers slam against their limits.

    But hey, a lot of people aren't really hearing bass. They are hearing the harmonics. I seem to be the only one in the world who can hear past that fantastic transient response and notice the massively distorted bass from of SR1a at louder volumes.
     
  12. rlow

    rlow A happy woofer

    Pyrate Contributor
    Joined:
    Jul 18, 2017
    Likes Received:
    7,738
    Trophy Points:
    113
    Location:
    Canada
    Just seeing that bass shelf makes me not trust the rest of the EQ adjustments. When I did some messing with EQ on my HD650 (stock) I tried a shelf like that (actually was around 7–8dB up at 20Hz, starting gently around 130Hz) and it was a bloated distorted mess. I did it based on what seemed to be indicated in the Drop ZenCan 6XX curve, but it was complete garbage. I starting lowering it 1 dB at a time and I believe I ended up lowering it to around 2-3dB up at 20Hz, starting around 80-90Hz - seemed to be the best I could do without causing a resonance on certain bass notes and sounding overly bloated and muddy. I don’t know WTF is up with people thinking that putting a bass shelf like that on an HD6XX would sound anywhere acceptable.
     
  13. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Imma just gonna keep posting this over and over.


    Sigh, Sean Olive has caused so much harm in this hobby.
     
  14. Carlos CPA

    Carlos CPA Friend

    Pyrate Contributor
    Joined:
    Aug 29, 2017
    Likes Received:
    572
    Trophy Points:
    93
    Location:
    Baltimore
    @purr1n, thank you VERY much. In my initial zeal, I tried the using EQ settings from the two Utopia compensation curves in Oratory1990's database before reading your subsequent post, quoted above. The EQ settings from Oratory1990 were, shall we say, most disappointing. They RUIN the Utopia. These headphones were not engineered to sound like Beats.

    I have tried your suggested EQ settings, as shown below:

    [​IMG]

    [​IMG]

    What a difference!!! This light touch, so far, has seemed to subtly enhance what the Utopia can do without introducing any new deficiencies. I am indebted to your sage advice.

    I think a new thread with suggested EQ settings for different headphones from some of the Trusted Ears on this site may be helpful. There does seem to be a lot of art and science at work here, and I think some of our senior members may be able to enlighten those among us with less wisdom as to starting points for EQ on certain headphones.
     
  15. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    ^ +100000

    I think an EQ sticky, with maybe a thread where the ears can argue over each other on best EQ - maybe with simple rules: no more than x3 PEQ. Those Utopia settings I gave you came not from measurements, but just me using my ears. The key is light touch and willingly to throw data away that doesn't jive with experience. Headphone measuring science isn't quite there yet. This is the reason why Jude gets a new coupler from sciency German guys every other month and announces better measurement results ... that are still problematic.
     
  16. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    In my opinion the ultimate test of any AFR measurement is having it flipped and via convolution based EQ applied to the measured headphones. The result should be tonally indistinguishable from a reference speaker system. Otherwise the best one can hope for is consistency within a measurement system.

    The problem with Jude that he wants a machine that does content marketing. He isn't willing nor capable to fight for better measurements.
     
  17. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    FWIW, on the Utopia I'd try maybe 4 filters, based on my listening experience a while back:
    1. Broad bass shelf bringing bass down 2dB or so. Not subbass boost as I think the small driver already struggles and extension is good anyway. Say 110Hz, Q0.4, -2 dB
    2. Bring down 1.3kHz: Say 1340Hz, Q1.87, -3.2 dB
    3. Raise 2kHz: Say 2000Hz, Q3, 1.9 dB
    4. Sharp, high Q notch around 5.5 to 6kHz. Use szynalski.com to find the frequency. Q possibly as high as 5 to 7, not sure.
    As a result of bringing down the bass and an important part of the midrange this could sound somewhat lean. I'm really not sure as I haven't tried it. Can always use a low-Q treble filter to bring it back down.
    That's a sample size of one with very different targets, very different HATS and a different scale on the plots aswell. I've linked it before, but IMO this is much more interesting in that regard:
    https://www.reddit.com/r/headphones/comments/g18b86/comparing_headphone_measurement_systems/
    As you've suspected the standard deviation does get somewhat larger past 5kHz, but not necessarily as much as one might think at first. For non-in-ear phones things do however seem to get wild past 10kHz.

    But you're right... I just went through my measurements again and the sharp notch at 6kHz in the EQ target wouldn't work for me. The peak I measure is more at 5.5kHz and having the notch at the wrong frequency could be a disaster in that regard.
    In fact that's what the HD800S sounded to me: The resonator is tuned to 6kHz, but (as mentioned before) with my ears there somehow isn't a peak at 6kHz. This is not limited to the HD800, I hear and measure a strong dip at 6kHz with the HD650 aswell. So by tuning the HD800S resonator to 6kHz the treble got much rougher and actually peakier to my ears. I listened to music and sine sweeps and I actually heard a dip at 6kHz IIRC.
    The SD resonator (which is a much lower Q resonator) works much better for my ears and with the right mods I simply fill in the hole (the hole itself creates somewhat of a peak at 6kHz and dip at 4kHz).

    I think one part where we disagree is that for headphone measurements I mostly religiously look at the 300Hz-4/5kHz range and tend to somewhat ignore the treble and bass part, save for level. I'll get a worse seal with many closed headphones anyway. I find oratory1990's effort quite good in that regard, with the exception that I think the Harman target could be somewhat shouty/forward. Really somewhat U-shaped as a result of the bass emphasis as we're seeing here. Nothing I've not said before. But again, you can always apply a different target.


    To be honest I think the fight against Harman's bass target is not one we are going to win. We just have to hope that headphone designers won't follow it too much. Unfortunately it seems they are, which affects me a lot since I can't handle anything that's just 2dB bassy. Still, I've already started ignoring it and mentally compensating the plots for neutral bass which will look rolled off. Heck, maybe it's just a matter of preference in the end as many people seem to prefer more bass than I do.

    Anyway, I actually meant to work on a different post tonight. Spent too much time in this thread instead and the other one is going to be somewhat shorter as a result.
     
    Last edited: Aug 28, 2020
  18. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I don't think we are in disagreement here. My problem is the auto-EQ results from 5kHz-12kHz. The Oratory stuff, if a Harmon->B&K is apply is good from 300Hz to 4kHz (I won't give him up to 5kHz :D).
     
  19. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I'm looking at past data. The V1 or FAC coupler would be a good candidate for Auto-EQ, provided the following:
    • The occasional nulls (that turn into ringing in CSDs) are ignored.
    • It may be a good idea to not try to correct narrow dips anyway because of frequency shifts depending upon one's ears and head.
    • Likewise same for narrow peaks.
    • We have a way to account for lack of damping, which for high Qms/Qes drivers can have exaggerated peaks.
     
  20. Cynric

    Cynric New

    Joined:
    Dec 29, 2019
    Likes Received:
    18
    Trophy Points:
    3
    Location:
    DE
    Uh, a thread about one of my favorite nerd toys! It's got quite a reputation and there are lots of useful (and useless) mods for it around.
    I've been using one of these for my desktop speaker system (Adam A3x + Eve TS107) for a few years now and I am really happy with the setup. The Behringer is fed through SPDIF and I use it for a very slight Stereo expansion and as a DAC. Oh and for enjoying the RTA display of course.:D
    I did try the AutoEQ with my room but ended up not using it - the spot where the mic sat sounded great, every other position sucked..

    The DEQ2496 analog out is actually pretty good. They use the former totl AKM AK4393 chip along with a decent balanced routing to and on the analog board (except for the filter stage that goes single ended for a bit). Give the PSU and PCBs an overhaul with new Panasonic FR caps, swap the driver opamps for some OPA1656 and the filters for OPA1612 and you're jolly good. Some guy (don't remember who and where, probably on DIYaudio) said the filter ceramic caps were causing distortion, so you could swap them for some PPS types on the way. And then there is the frenchman who builds custom clocks for these things - although I would rather just fiddle in some nice 10ppm MEMS oscillators and remove the quartz driving circuity.
    Did those mods (except for the clocks) with mine and with the Adams it just sounds great. I've always liked AKMs voices..
    If you can pick one of these things up cheaply, it's highly recommended!
     

Share This Page