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

Etymotic ER2SE - A Reference Headphone for your Ears and Your Couplers

First
 
Back
1 2 3
5 6 7
Next
 
Last
  1. hakuzen
    No problem here. Yes, any iem can be used for this purpose, even black MH755. I'll do full measurements, including THD, IR, sensitivity, and tip rolling (some etys bi & tri-flanges and foam, spiral dots, spinfits, kz starlines normal and flipped, final e, sednas, chinese foams..). Different output impedance measurements is not needed, because they both are dynamic drivers.

    It's great you got a demo GRAS coupler!! Looking forward your measurements
     
  2. castleofargh Contributor
    to those not familiar with that, I suggest that you abstain if you're not clear on the SPL output of the IEM with the way you amp is set. or at least make sure that the level isn't very loud for your ears, and leave a good deal of time between steps. there is a warning in REW, it's mostly there for speakers, but if someone is ever going to ever ruin an IEM with a measurement, that's one of the likely ways.


    otherwise I'll also measure anything I can think of, but right now I'm almost troubled because my old fanless crap computer only has an serial output for the screen, and I stupidly got rid of the small old monitor I had for it after forgetting why I was keeping it(what can i say, I'm a genius). so I'm struggling to find the absolute cheapest solution as that stuff serves no other purpose beside trying to measure maybe 3 IEMs a year for a super specific question in the quietest level of noise possible in the middle of the night but not on a weekend. I HAVE REAL PROBLEMS IN MY LIFE!!!!! :stuck_out_tongue_closed_eyes:
     
    Slater likes this.
  3. SilverEars
    Roger that!

    I recently got a fake MH750 on Ebay and reported it. Fakes are just too obvious if you hear it. It sounds totally off. Also, I thought it was suspicious that they covered the Sony logo with black tape.

    Actually, I can include the fake MH750. I can also include a set of AKG Samsung iem that comes with Galaxy 8 series phones. It has a pretty good response, until you get to the main treble area. Once it makes it to @yuriv , I'm totally fine with him doing any mod he wishes to get the treble response smoother.
     
  4. yuriv
    That sequence works for me. The international shipping cost won't be a problem.

    Other notes:
    If I understand it correctly, the ER2SE came with some measurements from Etymotic and they have an SPL for a 200 mV, 1 kHz sine with the ER2SE at the reference plane. Or is that your calibration? If I may make a suggestion: if the IEMs are going back to you, maybe make another measurement, but this time at 500 Hz.

    The ER2SE has a rising response at 1k. Also, I've noticed some variation in measurements made with these IEC60318-4 compliant couplers around that region. The coupler's first internal volume (the one with three slits) has a Helmholtz resonance around 1.19 kHz if I'm not mistaken in my calculations. So any variation on the size of the slits or the volume of the cavity will have an effect on the coupler's response around there. I've noticed it with the two couplers that I have. Measurements of Etymotic IEMs online show slightly different bumps around there too. Is the IEM really doing that, or is it the coupler?

    A measurement at 500 Hz is far enough from that resonance. Moreover, it's also around where the ER2SE's and MH755's responses are flattest. A reference signal at 500 Hz might be a better point for calibration if we all are to agree on an SPL.

    Another possible advantage of the 500 Hz calibration is that IEC 60268-7: Sound System Equipment — Part 7: Headphones and Earphones specifies a test level of 94 dB SPL at 500 Hz, or 1 mW at the headphone’s rated impedance. I don’t have a copy of the standard, but an online search at least reveals the standard test level. See here for one example: https://www.audioxpress.com/article/Headphone-Testing-Part-1-The-Basics



    Adding distortion measurements to the test suite is good. Just let us know the exact conditions so we can try to replicate them. If we’re worried about frying the ER2SE or the MH755, maybe we should toss in a throwaway IEM that we don’t care too much about. Here’s an example:

    Moondrop Spaceship right channel no grille.png
    Moondrop Spaceship, right channel, grille removed - stepped sine test
    12 points per octave, 131,072 FFT length, 2 averages​

    I made the foolish mistake of bringing a brand-new pair of Moondrop Spaceship to the gym when I forgot to pack my EarPods and Jabras. Don’t ask how it happened, but the left earpiece got crushed under a barbell. So now the right channel becomes the throw-away IEM for stepped-sine testing. I shouldn’t be too worried though, because the signal source is limited to 1 Vrms.

    The above graph shows a few things that can go wrong with the measurement. The recording interface can pick up power line noise (60 Hz in the U.S.). This is with a Startech USB adapter on a Lenovo Thinkpad Carbon X1. The Apple USB-C audio adapter seems to be cleaner. I’ll have to check again. The huge step taken by N represents the point during the test when the air conditioner kicked in. Lol.


    The mod would involve well-known techniques for damping treble resonances. Maybe a better use for that Samsung IEM is the stepped sine measurements. Toss it in the package as our throwaway IEM after you make your measurements. We can still play with the treble if we limit ourselves to simple, reversible mods.
     
    csglinux and hakuzen like this.
  5. castleofargh Contributor
    not sure the er2 comes with a custom graph, but based on my experience with Etymotic, their impedance and sensi at 1kHz are usually correct(not like the fantasy specs from some other manufacturers at least). but one thing to consider about their measurements, like those that come with the er4 to show channel matching, it that they do not use tips! they stick the IEM without tip into a mold they made so they can ensure rapid and repeatable measurements. it's obviously great for them, but it might be the cause for some more variations compared to what we get.

    I support 300 to 500Hz for voltage measurements about sensi, because that's where my dirt cheap multimeter comes closest to the correct value with low AC voltages. and I suspect a bunch of cheap meters might not be very accurate with the evil, fast frequency and low voltage combo. it's also a typically more stable area for frequency responses as noted. doesn't stop us from sharing what we get at 1kHz for those who care of course.
     
  6. csglinux
    Good & valid points raised here about the risk of destroying an IEM during measurements. I do know of somebody who recently did just that. I won't mention any names :wink:

    I think it would be fine to take any old beater set of IEMs (or IEM, singular?!) and use that instead for THD measurements. Good suggestion!

    The ER2SE don't come with any sensitivity or FR data from Etymotic; I believe that's limited to the ER4 series. Those SPL numbers came from me. BTW, I can certainly re-measure at 500 Hz - that's a good idea, as there might still be some influence from the coupler at 1 kHz. I could also supply a "reference" measurement from an RA0045 if that's helpful (using the stock triple flange tips, rather than putty). The use of quotes there is because I suspect we're all going to see significant discrepancies in the treble (~10 kHz+ region) and we'll never know which is right, because probably there is no right. I suspect treble variations are going to be our major conclusion from all this - even amongst those of us using IEC 60318-4 couplers from the same manufacturer.

    The reason why I thought the dB/mV numbers would be useful is mainly in helping folks get to roughly the same SPL for measurement purposes. For FR it's probably not critical - I'm using 90-95 dB. For THD, I try to use 80 dB, but that does need a super-quiet environment. If we agree on a dB level for THD, I wouldn't object to whatever method people wanted to use. Other than frequency resolution, I would expect/hope that we'd all converge on the same numbers, i.e., the characteristic THD for the headphone at that particular SPL, which would/should not be sensitive to the method used. (I'm not suggesting mics, external N&V, sound cards, couplers, etc., couldn't change results here, but the underlying algorithm shouldn't.) I think the focus should be THD, not THD+N, because N will be some random external pollution that'll be different for all of us, and not relevant to the DUT.

    But feel free to argue with me if I've got all this wrong :wink:

    P.S. Added the word "coupler" in that last set of parentheses. Thanks @castleofargh.
     
    Last edited: Jul 17, 2019
  7. castleofargh Contributor
    ok, I have no idea if anything I'll say makes sense, you've been warned! if a coupler has a different resonance at a given frequency, won't that logically also change the level of the harmonic content in that area? not that it will make a big difference given how THD is expressed. I'm only talking about the principle.
     
  8. csglinux
    For sure it will - the cynic in me suspects this is a big reason for certain OEMs wanting to use GRAS' new hi-res coupler. Because without them even lifting a finger, they can instantly reduce the THD on their product specs, even in the [0:10 kHz] range :wink:

    I think I should have explained myself better. When I said I hope we all "converge on the same numbers", that's just the same blind optimism I have that we'll all one day converge on the same FR curve for the same headphone. In other words, while hardware obviously matters, the choice of software algorithm (used to compute FR or THD) shouldn't. I'm ok with folks using whatever algorithm they feel comfortable with or have available.
     
    Last edited: Jul 17, 2019
  9. SilverEars
    I've conducted some measurements of the ER2SE and ER4XR, and the other iems (MH755, etc..). I will be adding a Samsung AKG EO-IG95 to this tour iem package for others to play around with. Also the fake MH750 I've obtained from Ebay. Which are all included in these measurements as well. All the measurement data are uploaded to the shared drive.

    I've posted the ER2SE and ER4XR measurement overlay in the ER4 thread here. https://www.head-fi.org/threads/if-...-thread-for-you.538615/page-922#post-15069463

    Here is everything overlayed

    All Overlay.jpg
     
    Last edited: Jul 18, 2019
    csglinux likes this.
  10. SilverEars
    Anybody have series of measurements of the ER2, ER4, or any of the Ety that depends on tip size used and/or insertion depth? I would like to take a look that, thanks.

    I've used small size triple flanges for both ER2SE and ER4XR at the same insertion depth (for the measurements I've linked on the previous post).
     
    Last edited: Jul 18, 2019
  11. csglinux
    I've forgotten (or I never bothered to dig around in the ER2SE box), does it come with three sizes of eartip, S, M & L - and the medium were on by default?
     
  12. SilverEars
    There were 2, and looks to me like L and S, but I could be wrong on the size. I can't imagine a smaller sized triple flanges. Both fit the coupler fine, but L needed to be forced, so used small.

    I'm having trouble with fit, so could it be M and S? Perhaps there was no L.
     
    Last edited: Jul 18, 2019
  13. McMadface
    The ER2 comes with L and S triple flange and a set of foamies. I used the S for measurement.
     
    csglinux likes this.
  14. csglinux
    The specs here:
    https://www.etymotic.com/er2-new.html
    just say they come with "Foam and 3-flange ear tips". I think you're right there are only two pairs of triple flange. So I think we should all use the smaller pair, which you've already done - great!

    P.S. I also have trouble with the fit with any of these triple flange tips. Shame, because I find they are the best sounding tips for these Etys - just not the best fitting or most comfortable for me.
     
  15. csglinux
    Thanks for the confirmation @McMadface!
     
First
 
Back
1 2 3
5 6 7
Next
 
Last

Share This Page