Simulating other headphones with AutoEq

Discussion in 'Computer Audiophile: Software, Configs, Tools' started by zottel, Feb 17, 2022.

  1. zottel

    zottel Friend

    Pyrate Contributor
    Joined:
    Jan 22, 2022
    Likes Received:
    1,229
    Trophy Points:
    93
    Location:
    Franconia, Germany
    A thing I've been playing around with recently:

    AutoEq (the script by Jaako Pasanen at https://github.com/jaakkopasanen/AutoEq ) can not only be used with the ready-made Harman results, but you can create EQ settings/convolution files for any target curve you desire.

    That means that instead of EQing headphones to a desired result, you can also EQ them to the result another headphone produced. I.e. instead of telling AutoEq to use one of the targets AutoEq provides, you can just use the measurements of another headphone as the target.

    This has limitations, of course, in what the headphone you use can reproduce, and in the quality of the measurements. And obviously, stuff like the technicalities of a headphone can never be simulated. But maybe it could help with getting an idea of the tuning of a headphone before trying to find a possibility to actually listen to it.

    Maybe.

    I tested it with my Ananda and HD650. With oratory1990's measurements, that worked remarkably well (tuning the Ananda to sound like the HD650). At least for some short A/Bing, the Ananda with EQ sounded really very close to the HD650. (Not so much when I used the rtings measurements. That sounded similar, too, but not as close as with oratory1990's measurements. I'd like to test with crinacle's measurements, too, but his raw data is not included in the AutoEq git as it is only available to his Patreons.)

    Then I tried with an actual headphone I don't have, where I want to know if the sound might suit me or not, the Focal Clear Mg. (I had to manually import newer oratory1990 measurements as oratory1990 does have the Clear Mg, but it's not included in AutoEq yet.) And the result is so dark that I wonder if that can be real.

    I mean, with some brain burn in, pop and metal and hip hop etc. can sound really great, there's a really nice blackground because so much treble is muted, and it all has lots of punch. But to my ears, the sound is so extremely dark that I'm not sure if a headphone called "Clear" can really sound like that. It's totally cave-like. It adds some rather nice feeling of a large church to classical recordings, but to my taste it's off, I don't like it for classical music.

    Have you ever tried simulating other headphones? Does that even make sense or do you think that's complete BS?

    Would you care to test it with headphones you own?

    As there's a bit of a learning curve with using AutoEq, I'd offer to create EQ settings for specific headphone pairs that you could test with your setup, to see if the results are any close to reality. Just send me a PM if you want to try it out, and if your equalizer is limited to a certain number of bands, please state that, too. Note that I can only produce EQ settings for headphones for which there are measurements available in AutoEq (minus crinacle), and it only makes sense, of course, if the measurements come from the same source, i.e. not using an oratory1990 measurement as input and an rtings measurement as output.

    Thanks, zottel
     
    Last edited: Feb 17, 2022
  2. zottel

    zottel Friend

    Pyrate Contributor
    Joined:
    Jan 22, 2022
    Likes Received:
    1,229
    Trophy Points:
    93
    Location:
    Franconia, Germany
    Just to add that, after some more brain burn-in, I like the Clear Mg simulation for classical recordings, too.

    While the not-EQ'd Ananda provide more detail in the treble, they also sounds more artificial, like a recording, while the simulated Clear Mg sounds more like a real concert to me.

    Switching from Clear Mg simulation to native Ananda, in positive words, sounds like entering the room where the musicians play, or in negative words, like switching from playing via speakers to playing via smartphone. To me, it's more like the latter in most cases. (Not as bad by far, of course, but you get the idea.)

    Now, the question is if the real Clear Mg really sounds roughly like this. Unfortunately, I currently don't have the money at hand to actually order one and keep it if I like it.

    P.S.: I wrote in my introduction post that I was looking for a headphone with a sound signature similar to my Dynaudio Music 5. The simulated Clear Mg is close.
     
    Last edited: Mar 4, 2022
  3. zottel

    zottel Friend

    Pyrate Contributor
    Joined:
    Jan 22, 2022
    Likes Received:
    1,229
    Trophy Points:
    93
    Location:
    Franconia, Germany
    Now that I have the actual Clear Mgs on hand: Did it work?

    I didn't do actual A/Bing between my Anandas pretending to be Clear Mgs and the real Clear Mgs yet, but I mostly listened to the EQd Ananda during the recent months, so I know well how that sounds.

    So did it work? The answer is yes and no.

    Yes, because the Clear Mg was easily recognizable as the headphone I had simulated, i.e. I immediately felt that this was more or less the same sound signature as I had heard before.

    No, because even the sound signature by itself (not taking stuff like technicalitites into account) wasn't completely correct, the simulation had a little more bass, enough that I directly heard it with songs I know well.

    I daresay that it is still useful to do such a thing to get an impression of a headphone's sound signature if it's not possible to actually listen to it. It will never be perfect, but it gets into the right ballpark.

    That said, the real Clear Mg is a completely different headphone, of course. The only thing that is similar to the simulation is the frequency response. There's much more detail, much more slam, different behavior with certain sounds like piano, e.g.. But that was clear from the beginning. And, having read the opinions in this forum, I probably wouldn't ever have ordered a Clear Mg if I hadn't heard and liked the FR simulation.
     
  4. spoony

    spoony Spooky

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    651
    Trophy Points:
    93
    Yep, overall tone and balance can be approximated with EQ, but stuff like transient response, ringing, distortion, dispersion (the way the transducer propagates the wave) and timing subtleties due to depth and other elements of the ear cavity will remain.

    Glad it worked for you.
     
  5. Woland

    Woland Friend

    Pyrate
    Joined:
    Jan 13, 2021
    Likes Received:
    1,311
    Trophy Points:
    93
    Location:
    a friendly land
    The frequency response is an aspect you can change after purchase - you're not stuck with it. For example, I apply considerable EQ to my Clears to tame the treble. So arguably frequency response is one of the less important aspects to know before purchase.

    Check out Vtory's impressions on the HD8xx for some insights on how much the owner experience & value proposition can change before and after EQ.
     
    Last edited: Apr 27, 2022
  6. E_Schaaf

    E_Schaaf MOT: E.T.A Headphones

    Pyrate BWC
    Joined:
    Mar 28, 2017
    Likes Received:
    9,589
    Trophy Points:
    113
    Location:
    Staten Island, NY
    Home Page:
    +1

    EQ - an electronic solution to what is often a side effect of mechanical problem. A broad stroke patch but never really a total fix.
     
  7. Woland

    Woland Friend

    Pyrate
    Joined:
    Jan 13, 2021
    Likes Received:
    1,311
    Trophy Points:
    93
    Location:
    a friendly land
    Ed, i have great respect to you but strongly disagree. The two-step sequence of physical tuning then EQ to mop up afterwards is likely what you're thinking about.. and sure, it is commonplace and sloppy.

    But it's also possible to use EQ well, having it in mind from the outset. Powered earbuds and noise canceling headphones are segments where we see that design process atm.
     
    Last edited: Apr 27, 2022
  8. E_Schaaf

    E_Schaaf MOT: E.T.A Headphones

    Pyrate BWC
    Joined:
    Mar 28, 2017
    Likes Received:
    9,589
    Trophy Points:
    113
    Location:
    Staten Island, NY
    Home Page:
    I assumed we were talking about wired passive headphones since that's all that's been discussed in this thread so far - and from a consumer perspective, not as a design consideration. I have no problem with active speakers and headphones that utilize extensive DSP when the DSP itself is an integral part of how the other elements are designed. But that's hardly the same thing as trying to use DEQ to make one passive wired headphone sound like another like OP was asking.
     

Share This Page