NwAvGuy's Dream Cums Tru: Schiit Magni 3 Heresy

Discussion in 'Headphone Amplifier Measurements' started by purr1n, Nov 22, 2019.

  1. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    73,416
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The time spent is because of curiosity, It's not everyday that @atomicbob runs into amps that don't register any distortion on his dScope. Could be because the signal generator on the AverLAB is the cause of the distortion that we see on the AverLAB results, or that the dScope ADC doesn't have low-enough distortion to see it on its own, or there's some other reason attributable to the "measurements are hard" thing. Would be interesting to see AP-5xx measurements, which I am sure ASR will eventually provide.

    I've also gotten to the point where I feel I have a consistent methodology, cables, layouts, power outlets, loads, etc. to get repeatable results. For instance, I held on to headphone FR measurements for almost a year before even posting anything on the Internet. Little snippets is one thing, but you want to be sure before getting more into the game. You don't want to be that dude who is known for borking measurements or who is talking out of their ass (like Jude early on before he got schooled by others and then trained up by AP Jedi to take on Amir's borked linearity measurements.) Finally, Avermetrics finally released version 3 (beta) of their software which has usability improvements. I was hoping to get dual tone, IMD, or multitone, but for the time being, I am stuck with the Modi 3 as the signal generator.

    I can do most of the measurements, it's just that AP with their super software makes it x100 easier. It's not too different from ethical hacking. Nothing replaces a good penetration tester, but the expensive five-six figures pen test software makes it x100 easier and faster.
     
    • Like / Agree Like / Agree x 6
    • List
    Last edited: Nov 25, 2019
  2. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    15,447
    Trophy Points:
    113
    Location:
    On planet
    It's not that the amp doesn't register distortion but rather the amp has low enough distortion that the analyzer's signal generator is now the limiting factor in obtaining an accurate reflection of the amp's true measurement. dScope has an incredible Gain Staging / ADC / SW combination to resolve way below anything that we know to matter at this time. The same issue exists with the Avermetrics signal generator and also the APx555, though the limit is slightly lower for each respectively.
     
    • Like / Agree Like / Agree x 6
    • List
  3. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Friend MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    9,978
    Trophy Points:
    113
    Location:
    Irvine CA
    No active components. Shit Bob, are you measuring that souless piece of shit also known as the NoAmp?
     
    • Like / Agree Like / Agree x 14
    • List
  4. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    73,416
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Hahaha
     
    • Like / Agree Like / Agree x 1
    • List
  5. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    15,447
    Trophy Points:
    113
    Location:
    On planet
    There weren't enough zeros in the previous measurement. Here it is again run with higher resolution:
    20191125 mystery analyzer measurement - t2.png
     
    • Like / Agree Like / Agree x 9
    • List
  6. Biodegraded

    Biodegraded Friend

    Friend
    Joined:
    May 28, 2017
    Likes Received:
    5,028
    Trophy Points:
    113
    Location:
    Vancouver BC
    But eight billion zeros
    Is still zero
    If you got no heart

    (Robyn Hitchcock)
     
    • Like / Agree Like / Agree x 5
    • List
  7. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    15,447
    Trophy Points:
    113
    Location:
    On planet
    And yet I've spent, and will continue to spend many enjoyable hours using this amp:
    guess this amp A04 THD+N  THD  nth-HD FFT 300R - 4+HD+N with 60Hz 0dBu.png

    Go figure.

    Can you guess which one this is?
     
    • Like / Agree Like / Agree x 2
    • List
  8. Tim Thomas

    Tim Thomas Friend

    Friend
    Joined:
    Sep 27, 2015
    Likes Received:
    272
    Trophy Points:
    63
    Location:
    Detroit Area
    Bottlehead Crack?
     
    • Like / Agree Like / Agree x 4
    • List
  9. Vtory

    Vtory Illogical Spock

    Friend MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    7,348
    Trophy Points:
    113
    Location:
    Greater Philadelphia Area, PA
    Probably the same one I am listening with..

    PS. SW51+
     
    • Like / Agree Like / Agree x 4
    • List
  10. Pogo

    Pogo Friend

    Friend
    Joined:
    Jun 14, 2018
    Likes Received:
    642
    Trophy Points:
    93
    Location:
    SoFla
    • Like / Agree Like / Agree x 1
    • List
  11. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Friend MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    9,978
    Trophy Points:
    113
    Location:
    Irvine CA
    With that amount of measurement resolution, it's no good Bob. The measurement is not "Excellent", just "Good" bordering with "Poor" actually. You are not close enough to absolute 0.0 dBFS in amplitude for channels A and B cuz it's Monday.

    Have some decency and make it at least somewhere around +/- 0.0000001 dBFS.
     
    • Like / Agree Like / Agree x 3
    • List
    Last edited: Nov 25, 2019
  12. Jinxy245

    Jinxy245 Vegan Puss

    Friend
    Joined:
    Nov 17, 2016
    Likes Received:
    4,848
    Trophy Points:
    113
    Location:
    Montgomery, New York
    Is that made by Schrödinger??
     
  13. mscott58

    mscott58 Friend

    Friend
    Joined:
    Oct 13, 2015
    Likes Received:
    2,142
    Trophy Points:
    113
    Location:
    Philly - Yo
    [​IMG]

    Best Schrödinger quote ever was him talking about some of his quantum work, where he said:
    "I don't like it, and I'm sorry I ever had anything to do with it." |\/|
     
    • Like / Agree Like / Agree x 2
    • List
  14. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    15,447
    Trophy Points:
    113
    Location:
    On planet
    @ultrabike and @purr1n NoAmp: https://www.kvraudio.com/product/noamp_by_mokafix_audio

    Good guess. That is among my favorite OTL amps.

    Correct. This isn't a FOTM amp but destined to become a classic. Acquired out of curiosity, now staying with a vengeance. One of my prized amps in the lab inventory.

    Actually it is necessary to keep the stimulus level this low to avoid intersample overs from circuit ringing and spline fit algorithms. May the BS be with you. :D
     
    • Like / Agree Like / Agree x 8
    • List
  15. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Friend MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    9,978
    Trophy Points:
    113
    Location:
    Irvine CA
    Schrödinger was a mortal. So no.
     
    • Like / Agree Like / Agree x 1
    • List
  16. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    73,416
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    upload_2019-11-26_14-51-7.png

    upload_2019-11-26_14-51-18.png

    NOTE: gain makes a difference for THD+N measurements.

    For any scientific comparison of THD+N, we need to do an apples-apples comparison and keep the gain the same. In other words, anyone who lays out chart comparing SINAD (THD+N) of different amps should insure that the gain is the same across all of the amps being tested. The reason is that lower gain typical results is lower distortion because of increased feedback, and it would be wrong to "punish" the amp in terms of distortion measurements because of a design decision.

    The amount of gain that an amplifier has is a choice on the part of the designer. For example, a designer may intend the amplifer for use with inefficient speakers or headphones like the HE-6, henceforth implement higher gain. Inversely, a designer may intend the amplifier for use with efficient headphones like the Grados or ATH-AD2000s or IEMs, and decide to use unity gain. Heck, heaven forbid, a designer may even feel that more open loop designs sound subjectively better!

    Finally, SINAD or THD+N does not really work in the context of amplifiers. There are two reasons: 1) different headphones or speakers will tend to run at different power (voltage) levels for normal listening; 2) in order to get a THD+N figure, we need to specify a voltage; 3) there's no guarantee that this "reference" voltage for THD+N is suitable for the particular transducer we are using, e.g. voltage necessary to run an HE-6 at normal listening levels will likely light an efficient IEM like the Andromeda on fire.

    Moral of the story:
    Anyone who lays out SINAD of headamps to compare on a simplistic chart is a moron
    Such a person does not understand the relationship of gain to distortion, design requirements or goals that affect distortion, or the wide range of operating voltages from the most efficient IEMs to the least efficient planar headphones.
     
    • Like / Agree Like / Agree x 20
    • List
    Last edited: Nov 26, 2019
  17. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Friend MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    9,978
    Trophy Points:
    113
    Location:
    Irvine CA
    YUP.

    On top of the headphone/speaker support, there is also support for the variety of DAC products out there, all with somewhat different output voltage ranges.

    With most amps, I prefer to use the lower gains unless the system starts to run out of headroom.

    YUP. Looking at THD+N sweeps for a few loads can give a sense of performance for a range.

    On top of that, we have that transducers are not as linear as resistors.

    I think Class A/B benefits greatly from relatively high impedance loads, not only in FR but for also in distortion, also due to increased feedback.

    Class A may still be FR sensitive with relatively small loads, but perhaps not as much in terms of distortion as Class A/B.

    Most of these things would not be obvious with any kind of resistive load (which transducers in general are not). Furthermore, transducers are all over the map as a function of frequency, and from product to product.

    Indeed, anyway one may want a look at it, synergy is not BS. AFAIK, the awesome-sauce product that is bestest for everything does not exist (with the exception of the NoAmp).
     
    • Like / Agree Like / Agree x 2
    • List
    Last edited: Nov 26, 2019
  18. Psalmanazar

    Psalmanazar Most improved member; A+

    Slaytanic Cliff Clavin
    Joined:
    Sep 27, 2015
    Likes Received:
    4,749
    Trophy Points:
    113
    @atomicbob how is the noamp plugin versus the legendary Simulanalog made by the overloud guy and the TSE BOD emulating a sansamp?
     
  19. wadec22

    wadec22 Facebook Friend

    Contributor
    Joined:
    Sep 28, 2018
    Likes Received:
    233
    Trophy Points:
    43
    Location:
    USA
    This is annoying as shit. For a science site, there is zero checklist or consistency when testing products. Certainly not just amps. Minutes after reviews are posted members clamor for tests that were not run on X device that were just run in the previous review. It often shows bias towards what the reviewer is rooting for to test well (maybe even subconsciously).

    Unfortunately most members or guests are searching for help in selecting product and think they find a great, unbiased reviewer to trust but are actually getting a guy in his basement drinking wine, measuring whatever the heck he feels like at the moment. while that isn't necessarily a big deal, as long as the "reviewer" is being transparent about that in every review but I do not think that is how it is communicated.

    Of course I say all that, and I do still find some useful information on there. You just have to dig for it and occasionally make some adjustments when trying to compare two products. And then apply context to what tests actually even mean anything...
     
    • Like / Agree Like / Agree x 8
    • List
  20. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    73,416
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    indeed, the inconsistency and lack of adherence to science in terms of applicability is disturbing. In terms of ASR "SINAD" measurements for headamps, I can't figure out if the SINAD or THD+N measurement is based on a reference of 1Vrms, 2V, or 4V or something all together different. It seems to change with every other amp, and as we've seen the above plots, the THD+N changes depending upon the output level. On ASR, I've seen 4V typically used across the board with a lot of balanced headamps, but I'm sure many know here how loud 4V is into most headphones. A 0.2Vrms to 0.4Vrms 1kHz signal into a RAD-0 or HD650 is going to cause some very serious ringing after 10 seconds of exposure and probably permanent hearing damage after a minute. I definitely wouldn't use 4V because its not normal listening volume.

    Most guests are going to see a single number and zone it on it. it's easier. I'm not against single numbers. 0-60mph times or 1/4 mile trap speeds tell us a ton about a car. What ASR is providing is something along the lines of max speed for a car running on random grade inclines. (Max speed is already next to useless because of factors such as gearing,)
     
    • Like / Agree Like / Agree x 3
    • List
    Last edited: Nov 26, 2019

Share This Page