How to pick a headphone as an EQ base.

Discussion in 'Modifications and Tweaks' started by JewBear, Nov 24, 2017.

  1. JewBear

    JewBear Almost "Made"

    Joined:
    Sep 28, 2015
    Likes Received:
    270
    Trophy Points:
    63
    Hi guys,

    So I'm considering my next head phone purchase. One thing I've realised is that my musical tastes are constantly evolving, and thus my preferred signature tends to change with it. I used to prefer quite bass heavy headphones but for classical or vocals music that doesn't really suit. I'm now on the hunt for a new closed back but instead of worrying too much about frequency response my main criteria is it's ability to handle EQ. I figure if I can get something that handles EQ well I can tweak the signature to my taste. I intend to use the new Minidsp HADSP and the ears measurement systsem once it comes out. This should allow me to use FIR filtering which will address phase issues to a certain extent. The major issue will probably be internal reflections (reverb?).

    So I'm wondering, what makes a headphone EQ well?

    What measurements are a good indicator of ability to handle EQ?

    Does anyone have any examples of headphones that take EQ exceptionally well?
     
    Last edited: Nov 24, 2017
  2. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    Look at THD curves and resonant peaks. I've found that Audeze LCD-2's handle EQ pretty well mostly due to good driver headroom. I could say the same for newer dynamic cans like Focal Elear.
     
  3. Deep Funk

    Deep Funk Deep thoughts - Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    9,024
    Trophy Points:
    113
    Location:
    Amsterdam
    Home Page:
    So you need a headphone with a big driver, neutral-ish sound, closed but easy to adjust/modify (EQ, dampening & pad-rolling)?

    What is the budget?
     
  4. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    Closed? Can't go wrong with Oppo PM-3.
     
  5. Ice-man

    Ice-man Friend

    Pyrate Banned
    Joined:
    Oct 12, 2015
    Likes Received:
    3,455
    Trophy Points:
    113
    I've been wanting to try the pm-3 with the Sonarworks EQ profile for a while now. And I'm certainly tempted given your endorsement. From my own personal experience I'll suggest the Aeon Flow Closed. Damn fine headphone and EQ's like a boss.

    @Hrodulf I'd like to see some of the MrSpeakers on the Ref 4 plugin list soon, please.
     
  6. spoony

    spoony Spooky

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    651
    Trophy Points:
    93
    Second this, any headphone with a smooth frequency response and lowish distortion will handle EQ with ease, you can't fully 'EQ-out' ringing.
     
  7. k1arg

    k1arg Facebook Friend

    Contributor
    Joined:
    Oct 12, 2015
    Likes Received:
    153
    Trophy Points:
    43
    Location:
    London, UK
    I've definitely had more luck in cases where the distortion curves stay low. But it also depends to some degree on your target curve. That is, for each headphone it'll work better if you don't bump up the same frequency ranges where it has high distortion. So if you try to EQ an HD650 to have an LCD2 bass curve, expect plentiful distortion-induced muddiness. And it seems to be easier to fill a suck-out than to chop a (resonant) peak.
     
  8. k1arg

    k1arg Facebook Friend

    Contributor
    Joined:
    Oct 12, 2015
    Likes Received:
    153
    Trophy Points:
    43
    Location:
    London, UK
    I think that it's also important to note that even if a headphone takes EQ well, each still has a character. Or at least fundamental types remain distinct. For example, EQ an LCD2 and an Elear to the same curve and they don't sound the same. The Elear still has an ability to render dynamics that, to me, the LCD doesn't match, while for the "texture" of bass notes, I'd probably go with the LCD. (Don't hold me to it, it's been some time and this wasn't a direct comparison I did.)
     
  9. Deep Funk

    Deep Funk Deep thoughts - Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    9,024
    Trophy Points:
    113
    Location:
    Amsterdam
    Home Page:
    If you have to test your EQ first just get a Brainwavz HM5 or Shure SRH840 to play around with. Closed, comfortable and each headphone has a good driver to work with.

    Bonus: both of these headphones allow pad rolling...
     
  10. JewBear

    JewBear Almost "Made"

    Joined:
    Sep 28, 2015
    Likes Received:
    270
    Trophy Points:
    63
    Well luckily, I have a pair of PM3 I can retrieve from my brother use for as a base. What about Shure 1540s? I'm thinking these might have a bit too much distortion to EQ effectively. I don't really want to spend the money on the Aeon closed as their distortion profile seems a bit worse than the PM3.
     
  11. JewBear

    JewBear Almost "Made"

    Joined:
    Sep 28, 2015
    Likes Received:
    270
    Trophy Points:
    63
    Just a question, how does Sonarworks handle phase change due to EQ? I guess you have a standard correction based on the frequencies changed per headphone model? Without being able to actually measure the headphone post eq, is phase correction even possible?
     
  12. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    True-Fi filters are all minimum phase with all the phase shifts related to FR changes. Reference 4 software lets you choose between minimum/mixed/linear phase filtering.

    Phase correction is a funny thing. For speakers it usually can mean many things, like slanted or multi-level baffles to space-align driver radiation points, phase linear crossovers... usually time or phase coherence goes out of the window once real listening environments come into play.

    Headphones are a bit simpler in a way that there aren't that many variables for the listening "room". We have attempted actively engaging into phase correction, but the audible effects were minimum at best and not necessarily for the best. Usually bass boosting will increase overall group delay for most headphones, however most people will agree that group delay increase is less audible than the gain in bass response.
     
  13. Ice-man

    Ice-man Friend

    Pyrate Banned
    Joined:
    Oct 12, 2015
    Likes Received:
    3,455
    Trophy Points:
    113
    @Hrodulf what is the recommended phase filter setting and why?
     
  14. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    As the others have mentioned I'd go for something with a smooth response and low distortion. PM-3 sounds like a decent option.

    You would need FIR filters for the reflections caused by your head (or the EARS), but not for the headphone response, as long as the headphone only has one driver per ear.

    Technically you can. The question is if the ringing you EQ is actually the ringing you hear.
    Isn't that exactly what minimum phase EQ does? Correct both the frequency and phase response of the headphone? Of course this only works as long as the measured response is accurate

    But doesn't extending the bass decrease the group delay? I've played with this myself by using one very broad EQ filter at say 10Hz to get flatline bass extension from the HD800. Of course that's a lot to ask from an open dynamic headphone, so I can see why you'd cut off the sub-bass, but IMO the phase shift is audible. In the end I don't use EQ (except for my RE-262 at the moment) so I don't care too much, but regular minimum phase EQ is what I'd use.
     
  15. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,276
    Trophy Points:
    113
    Up to you, man. The advantages and disadvantages are in phase coherence, delay and impulse response. In my opinion they're more interesting in theory and not that audible in praxis, if the filters are well designed. More clear cut preferences in filtering come into play for crossing over drivers or using EQ on different tracks, which need to be phase-controlled. Here's a good video on the topic -

    Additive EQ usually adds group delay. It's one of the arguments against building closed subs and then boosting the lows with EQ. Usually you get the same group delay as with an open box.

    P.S. Are you sure you were hearing group delay? Full spectrum EQ will push the low end THD of most dynamic open backs in double digit range and extra excursion might incur Doppler effect as well.
     
  16. 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 meant lower group delay with EQ, but also more distortion. I do think it's a phase thing - the seamless and effortless quality headphones have that go to 10Hz flat or so, something that you can't achieve with 20 or 30Hz extension, but with steep rolloff.

    I'm not sure I follow when you say additive EQ usually adds group delay, but I do think you're saying essentially the same thing that I meant with the closed vs ported sub. As long as the magnitude response is the same the phase response will also be the same as it's still a minimum phase system.

    I attached two graphs to explain what I meant. One is the HD800 without EQ on my coupler and the other is the HD800 with a 15db (!) Q0.5 filter at 10Hz in JRiver. I used JRiver to play the REW sweep for both measurements.
    Excess phase is the same (within margin of error) and that's really my point. As long as two headphones have the same frequency response, the phase response/IR/square waves, etc. will be the same too. EQ a T1 to the same FR as an LCD-2 and the CSDs will look the same.
    Magnitude and phase response is much better with the EQed one and the group delay is better down to about 25Hz, too. Of course you could try to extend it even lower, but in the end every transducer is a high pass filter so there will always be some phase shift in the lowest frequencies. Below 25Hz the group delay is worse and it hits almost 30ms at 10Hz, instead of 20ms without EQ.
    I'm not sure how accurate the distortion measurements are, but at the 95db I measured about 3.5% instead of 1% at 20Hz. Could be even lower though.
    (This is without mic compensation so in reality bass extension would be slightly better, but it doesn't matter here.)

    Of course you could try to get close to linear phase and FR down to 20Hz with a FIR filter or something. I wonder how that would sound.
     

    Attached Files:

    Last edited: Nov 25, 2017
  17. gixxerwimp

    gixxerwimp Professional tricycle rider

    Pyrate
    Joined:
    Sep 30, 2015
    Likes Received:
    5,773
    Trophy Points:
    113
    Location:
    small island claimed by China
    I'm definitely no expert, but this goes against everything I've learned from discussions on measurements here. An example would be a notch in the FR that turns out to be a peak in the CSD (or the opposite when a peak dies off quickly so it doesn't sound that bad). So I'm not convinced that EQing a T1 to the FR of an LCD-2 will result in the same CSD.
     
  18. TomHP

    TomHP Facebook Friend

    Joined:
    Nov 24, 2016
    Likes Received:
    153
    Trophy Points:
    43
    Location:
    Taiwan
    I think that should be

    "As long as two headphones have the same frequency magnitude and phase response, the IR/square waves, etc. will be the same too"

    OR

    "As long as two headphones have the same complex frequency response, the IR/square waves, etc. will be the same too"

    All of this is assuming headphones behave linearly, which is a gross simplification.
     
  19. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Well, that's obviously true, but it's also kind of pointless.
    What I meant is that two headphones with only the same magnitude response will also have the same phase response. As long as excess phase is closed to 0 across the band it should work, which according to my measurements was the case with every headphone I've measured. Measurement artifacts like dips turning into ringing not included (which is sort of what I meant by saying 'as long as the measured response is accurate')

    Two headphones with the same magnitude response will pretty much have the same phase response. If that's not your understanding then point me to headphones that strongly deviate from minimum phase behavior. I've not seen it in my data.

    Of course headphones don't behave linearly in terms of distortion, but as far as phase gets they're as close as it gets.
     
    Last edited: Nov 27, 2017
  20. TomHP

    TomHP Facebook Friend

    Joined:
    Nov 24, 2016
    Likes Received:
    153
    Trophy Points:
    43
    Location:
    Taiwan
    As an approximation you're assumption is close enough. But it's a simplication to say if the frequency response is the same, the phase response will be too. E.g. different headphone drivers will have different Le (inductance) resulting in different phase shift in higher frequencies.
     

Share This Page