1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

If you still love Etymotic ER4, this is the thread for you...

904 905 906 907 908 909 910 911 912 913
915 916 917 918 919 920 921 922 923 924
  1. panasonicst60
    No as he said it will not be audible because it's so close. I believe I can tell the difference.

    For example I have speakers with ribbon tweeters and no amount of eq can make a non ribbon tweeter sound similar. Great ribbon tweeters have an airyness sound to them that no other tweeter will match imo. It's maybe an apples to oranges comparison. :thinking:
    Last edited: Jul 10, 2019
  2. redrol
    ER2 has lower distortion. Buy it if distortion numbers are your thing.
  3. JohnYang1997
    Remember what I was trying to say was to let some of you think differently. I'm not saying etymotic is bad or cheap or anything. If you are able to get the etymotic sound without eq, that's really difficult to many manufacturers. Tho some of those who really know what they are doing can actually get close or get even better than etymotic. The existence of digital is like a bug or cheating. There are drawbacks and I also prefer the real thing without eq. But the ability to make it the way it is is incredible. People feel really hard to get their head around that. And here is the file, I spent more time on tweaking. I think it's really close now.
    Last edited: Jul 10, 2019
    chinmie likes this.
  4. csglinux
    He said very, very similar, but not 100% on account of non-linearities. It's pointless to debate percentages, but we all agree it isn't 100.

    Apologies. I think I read your post wrong. I think you're saying you think you can tell the difference between EQ'd headphone A and headphone B. Maybe you can, but that doesn't negate @JohnYang1997's argument.
    Last edited: Jul 10, 2019
  5. Joe Bloggs Contributor
  6. JohnYang1997
    Here it is. Change it to .csv file and import via equalizer apo
  7. Joe Bloggs Contributor
    Still not seeing it :thermometer_face:
  8. JohnYang1997
    Actually if you are talking about really dynamic it's all done in distortion. Just the way it is introduced us different or how the distortion is generated. When distortion is very low, it's reverse provable there is no big dynamic change. And in order to hear it it needs to be at least 0.01% (most headphones surpass that) and in order for most people to hear it or for us to hear it clearly it needs to be over 1%. Sadly there is no way to control real dynamic without introducing distortion.

    However that's real dynamics. If you change the frequency response eg v shaped, you can perceive it as higher dynamic. Maybe everyone's definition of "dynamic" is different (it's subjective because these are not real dynamic).
  9. JohnYang1997
    Seems headfi doesn't like it.
    (copy the following to a text file and save it as .csv file)
    20 -11.4
    25 -11.4
    31.5 -11.4
    40 -11.4
    50 -11.4
    63 -11.7
    80 -11.7
    100 -11.5
    125 -10.5
    160 -9.4
    200 -8.7
    250 -7.7
    315 -7.3
    400 -6.5
    500 -6.2
    630 -5.5
    800 -5.2
    1000 -5.3
    1250 -4.1
    1600 -3.7
    2000 -4.2
    2500 -3.7
    3150 -4.8
    4000 -7.3
    5000 -7
    6300 -6.3
    8000 -8
    10000 -7.9
    12500 -10.5
    16000 -9.8
    20000 -21
    redrol likes this.
  10. csglinux
    About those ribbon tweeters.... That could simply be another example of how difficult perfect EQ compensation is.
    Imagine you have headphone A that has a sharp zingy peak at 15 kHz, and headphone B that falls off a steep cliff at 10 kHz. Now try to EQ B to match A. That's going to be difficult, but that's a hardware limitation - it doesn't disprove the fact that FR is of overwhelming importance.
  11. JohnYang1997
    I never said it's 100% inaudible. But low enough most people can't hear it. And when we like you and me do hear a difference it's minimal.
    On about speakers, there is room that plays a big role. The dispersion/ angle response of driver will considerably affect the sound. If you can listen in anechoic chamber, a well done filter will generally match the sound. Metal dome may have minor peaks and valleys in the frequency response that can be hard to correct. Also high frequency extension can't be matched by a lower bandwidth speaker. The higher frequency may fold back into lower frequencies and affect the sound. But in general if you can listen to speakers in anechoic chamber you can match the sound of tweeter to almost audibility limit.
    However on speakers in general, unless both are single driver sealed box it's almost impossible to match the phase response and the actual produced waveform. There is another thing called linear shift invariant in digital processing also recursive/iir and fir filters. If the speaker it self will reiterate itself like ports does to the sound wave there is no chance to correct that 100%. Also the crossover is as bad. If the two drivers are not aligned in a way that's perfect there will be non correctable phase issue. Like I said in the very beginning, single driver headphones earphones tend to be approximately minimum phase system, which are easily correctable by minimum phase filters. Ported headphones earphones don't matter much because the port won't go back to your ear canal. So no crucial phase issue there.
    Also if there is huge peak in the frequency response ringing in the csd it's also hard to correct or impossible to correct. Because it breaks the minimum phase approximation as well.
    Sony mh1c responds well because it doesn't have big peaks in the high frequency range. Only a bit lf too much distortion at 1khz and very low frequency can be somewhat a problem. Other than that it's almost ideal.
    Last edited: Jul 10, 2019
    panasonicst60 likes this.
  12. Joe Bloggs Contributor
    Here's my attempt--it's from measuring the two earphones (the MH1C, and the ER3SE I own) using an IEM coupler and subtracting and smoothing the difference. A trough near 20kHz has been manually edited out.

    Inside the zip is a convolver file which you can use with more programs than just EqualizerAPO, e.g. the optional convolver component for foobar2000--which is also inside the zip.


    Here's the curve itself for reference:
    mh1c2er3se curve.PNG
    Last edited: Jul 20, 2019
  13. Joe Bloggs Contributor
    Why is every value negative? Looks like you can dial up every point at least +3.7dB

    In other news, in my measured version 20Hz is down 20dB from 1kHz. Lol
  14. JohnYang1997
    Because I was done in almost matched volume. If you want you can normalize it.
  15. JohnYang1997
    Not all mh1 are the same. There is definitely variants. Surely yours look too much bass. According to mine, to mro's and many others it's roughly around 10db. I allow of variation of 3db in the bass. The key is in the mids. And the highs.
904 905 906 907 908 909 910 911 912 913
915 916 917 918 919 920 921 922 923 924

Share This Page