IEM measurement distortion discussion (defunct)

Discussion in 'IEM Measurements' started by purr1n, Aug 17, 2016.

  1. 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
    After NOT seeing any significant differences in the usual D2, D3, and D4 distortion measurements above between the Orion and Andromeda (which sounds notably clearer and more articulate), I decided to run these on a hunch. Actually, the hunch thing is bullshit. I already knew. Those of you guys who have known me for a long time know that I always play the cards close to my chest and leave an ace up my sleeve. These are FFTs of 500Hz tones at about 89db. Graph is linear scale for what should be obvious reasons.

    Orion
    Orion FFT.png

    Andromeda
    Andro FFT.png

    Note that 7th order distortion at 3500Hz is higher on Andromeda than Orion. However, higher orders from 9th on are much higher on the Orion. In fact, harmonic distortion elements still continue on after about 15th or 16th order on the Orion, whereas on the Andromeda they finally disappear. There's a good argument that distortion so far down from the fundamental might not be audible. However, there's been the age old argument among engineers that higher order distortion is in fact more harmful than lower order distortion, and that even minute amounts is enough to have detrimental effects. As an aside, note that these results are consistent with the distortion plots in the first post, just that those plots don't go past 4th order distortion.

    The other case (in terms of measurements on this site as of today) I can think of where I have seen this behavior is here (the MicroZOTL 2 without the LPS): http://www.superbestaudiofriends.org/index.php?threads/microzotl-2-0-measurements.1724/

    Going forward, I think at least one (and preferably a few) FFT plots with test tones (maybe even several test tones to see IMD) might be worth exploring.

     
  2. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    WTF! Super high order distortions on both of them. I bet dynamic drivers don't have so much high order distortion. Still, that stuff is more than -70db down. I still think that due to the spectral response of music you wouldn't have to be Voldemike to hear it. Also what's up with the changes in the noise floor?
     
    Last edited: Aug 17, 2016
  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
    No idea on the noise floor. Could be some stuff related to crossover region. But yes, lots of higher order on balanced armature type drivers. Here's another look with an overlay. I think we are seeing reduced super high order (4000Hz onward) on the Andromeda because the treble drivers start to take over while the midrange drivers' output starts to be attenuated.

    Red = Andromeda
    Yellow = Orion
    FFT Diff.png
     
  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
    So that's actually an argument for going overboard with driver count and crossover parts. But as always, it's about tradeoffs.
     
  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
    Five way! For balanced armatures at least. This is sort of why I am working on a 4 or 5 way speaker design, but this idea is based more on lowering IMD since traditional dynamic drivers don't have all that high order garbage.
     
  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
    Sorry if I don't quite understand, but how does the crossover network attenuate the higher order distortion? Or was your comment strictly about IMD, which would then make sense. Because as I see it, only the acoustic filter would reduce the higher order distortion products, but not a crossover network.
     
  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
    There is only one midrange driver on the Andromeda (it's 2-1-2 I believe). The bass driver might also be responsible for part of the 500Hz signal (I am unsure of the crossover frequency, but I am pretty sure the crossovers are first order, for lack of internal space), therefore the bass / mid isn't combination working as hard as on the single driver Orion.

    P.S. I said it wrong initially by referring up to the mid / treble handoff instead of the low / mid handoff. But you get the idea. Letting the drivers play where they are most comfortable to reduce harmonic distortion.
     
    Last edited: Aug 17, 2016
  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
    It could very well be that the crossover is at a higher frequency like the UERM, though subjectively a lower frequency might make more sense. UERM is 4kHz/8kHz I believe. It's essentially a 2 driver with a supertweeter (that is the same driver as the tweeter). I still don't quite know how the impedance curves play into this for IEMs. Seems more complicated than with speakers.
    So you're saying that the individual drivers have to work less hard? Well, you could simply stack up on drivers without a crossover network then. Only the acoustic filter reduces higher order distortions. Stacking up on drivers also doesn't really change the relation between different harmonics.
     
  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 dunno. It's hard to say. I've seen this (the harmonic distortion extending up to 20th+ order happen) with other IEMs without crossovers, even doubled up. The Savant below is an example of this (two drivers, but no x-over, there's just a resistor). The stuff with crossovers seem to have less of the higher order junk.

    Note these were taken a while ago under different circumstances to the Andromeda / Orion 500Hz FFTs above.

    Savant (two drivers, no? crossover)
    fft savant.png

    UERR (2-2-1? multidriver)
    FFT UERR.png

    On the other hard, this could be like comparing apples and oranges since we don't know the qualities of the driver units used. But as with your question on whether stacking up on drivers would really change the relation between different harmonics: It does. These BA drivers have a peculiar behavior in that once a certain output is reached, distortion in the lower orders (say 2-5) remain mostly around the same while the higher orders (5+) start to rise. I don't have anything to show you as it was something I noticed on the spectrum analyzer while turning the volume up.
     
    Last edited: Aug 18, 2016

Share This Page