Schiit Yggdrasil Stereophile Review + Measurements

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by purr1n, Jan 20, 2017.

  1. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    OK, let me do this out loud:

    Fig.7 Chord DAVE, waveform of undithered 1kHz sinewave at –90.31dBFS, 16-bit TosLink data (left channel blue, right red).
    [​IMG]
    1. Per above plot, that's -1.5mV to 1.5mV peak-to-peak for -90.31db signal
    2. A 90.31db gain to 0dbFS (full output of DAVE) is effectively a voltage ratio gain of 32772. [ Voltage gain in dB = 20 × log (V2 / V1) ]
    3. 0.0015V * 32772 = 49.158V peak
    4. -49.157V to 49.158V peak-to-peak = 34.76 Vrms.
    34.76 Vrms should be the calculated full output of DAVE assuming the plot above is correct. So something is wrong because max output from the balanced jacks of DAVE is 12.35V peak, not 34.76Vrms or 49.159V peak.
     
    Last edited: May 25, 2017
  2. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    I wouldn't even fully blame JA on that one, except oversight.

    JA did seem to adjust his equipment to capture LSB toggling which yielded 1.5 mV peak. But that means CHORD is doing some funky shit because MSBs are going to clip. Who knows WTF CHORD is doing, but it sure as hell makes no sense.

    It's like driving things loud while cooking numbers to look great.

    EDIT: I'm actually giving JA the benefit of the doubt for not calling this out in his review. Unless I'm mistaken, this issue is pretty bad.
     
  3. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
  4. aufmerksam

    aufmerksam Friend

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    1,337
    Trophy Points:
    93
    Location:
    E. Lansing, MI
    I am reminded of a certain German auto manufacturer...
     
  5. noshortcuts

    noshortcuts New

    Joined:
    Nov 16, 2016
    Likes Received:
    2
    Trophy Points:
    3
    Location:
    Bay Area, CA.
    "20-bit chips are obsolete as they become the limiting factor in resolution when used in real-word audio circuits, ending up with <20-bit resolution. The very best current D/A processors achieve between 20 and 21-bit resolution, limited primarily by the thermal noise of the resistors used.

    John Atkinson
    Editor, Stereophile"​
    I think JA's limiting factor is that he's stuck on this idea and not using his ears. Yggdrasil lacking resolution? - not to my ears.
     
  6. Larry Megugorac

    Larry Megugorac Craps on Filipino accents to ease inner poverty

    Joined:
    Jul 27, 2016
    Likes Received:
    31
    Trophy Points:
    18
    Location:
    Brentwood
    Home Page:
    That's what my feeling was with his comments about the Yggdrasil measuring poorly.....But with Tubes it's "euphonic distortion" and that's ok? JA needs to put down the O scope and just listen...
     
  7. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    None of these DAC measurements where we see shit below -90, -100, -110db matters. None of it is correlatable to sonic perception phenomena. DAC measurements are a magnitude or two better than amp measurements, which are a magnitude or two better than transducer measurements. Transducers make gross errors, and even then, measurements do not tell the entire story.

    The only times it might matter is when measurements are egregiously bad, like on the Singularity 19. Atkinson is actually doing a disservice by not making this clear to readers. An example of this would be that dude on Computer Audiophile who claims he can hear the glitch on the Yggdrasil. I could process a 24/96 wave file to simulate that glitch, and I'll bet $10,000 he would not be able to hear that glitch on any other DAC.
     
  8. winders

    winders boomer

    Banned
    Joined:
    Feb 13, 2017
    Likes Received:
    1,596
    Trophy Points:
    113
    Location:
    San Martin, CA
    Good ol' manisandher.....in the following thread manisandher and GUTB ganged up on Yggdrasil:

    https://www.computeraudiophile.com/forums/topic/31284-just-got-a-yggdrasil/

    Haters.....
     
  9. earnmyturns

    earnmyturns Smartest friend

    Pyrate
    Joined:
    Sep 25, 2016
    Likes Received:
    3,218
    Trophy Points:
    113
    Location:
    Palo Alto
    Home Page:
  10. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    I totally understand if people like MadamSandler and GUTB don't like the Yggdrasil. As RH said in TAS, it's not a relaxing sound. (I don't think GUTB has heard the Yggdrasil or MSB DACs that he refers to.)

    The problem is when people like them, or maybe even more innocent people, start mis-attributing Atkinson's measurements and irresponsible commentary to what they hear. If they don't like the Yggdrasil, they just don't like it. Has nothing to do with the measurements.

    Heck, I've even warned folks to take amp distortion / FFT measurements with a grain of salt.
     
  11. winders

    winders boomer

    Banned
    Joined:
    Feb 13, 2017
    Likes Received:
    1,596
    Trophy Points:
    113
    Location:
    San Martin, CA
    I agree. I don't care if you don't like the Yggdrasil. Just don't make stuff or talk out of your ass about it.

    Yggdrasil does not have a relaxing sound. But, to me, it is not fatiguing either. It's engaging and that is what I want from my music. I don't want it to put me to sleep!!
     
  12. Delayeed

    Delayeed Facebook Friend

    Joined:
    Aug 19, 2016
    Likes Received:
    165
    Trophy Points:
    43
    Location:
    Funland
    "It's a 20-bit DAC in a 32-bit world. But more to the point, it doesn't support DSD or DXD.

    Native DSD is very very good and DXD sounds amazing. A Yggdrasil owner will never get to enjoy those.

    Also, high end DACs can resolve 24 bits of high definition audio...the Yggdrasil can't." -GUTB
    [​IMG]
     
  13. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    To be fair, he does exhibit an extreme talent for maximizing how much of an ass he can make himself look with very few words.

    One could sum up the bulk of the truly native DSD catalog as "Music you don't care about, by people you've never heard of." (for DXD that might become "Music you don't care about, by people whose names you can't pronounce.").

    Beautifully performed, recorded and mastered dross is still dross. Though in practice there is so little of it anyway, that it really doesn't matter.

    And DXD? Well, that's just 24 bit/352.4 KHz PCM. How much of that content is legitimate AND interesting? There's a dearth of proper 24/96 content, let alone anything better.

    And enjoy that lovely USB connection, since that's what you're going to need to play those formats properly.
     
  14. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    Yup. Saw JA's response.

    Well I have to give him something. Whatever his convictions and personal biases are, he definitively is a gentleman and will engage with reasonable discussion.

    My read on this is that JA is in agreement with our observations.

    For manufacturer's own measurements, I can only guess JA means these ones:

    http://chordelectronics.co.uk/wp-content/uploads/2016/09/DAVE-Technology-Presentation.pptx

    Going through the presentation, before hitting the characterization plots there is a discussion about:

    1) How amazing their WTA filter is (164000 tap FIR).
    2) How out of this world their noise shaper is (17th order with 46 integrators).

    Given the high order of their noise shaper, they probably need that crazy long FIR to mitigate out of band noise. Too many filter taps may require some random form of optimization and lots of bits. Alternatively, the noise shaper may be more controlled. Who knows.

    The characterization there shows THD and noise with 0 dBrA set to about 5 Vrms using a tone at -6 dBrA (2.5 Vrms) which seems reasonable, along with jitter plots, IMD (19/20 kHz), and 16/24-bit -90 dBFS tone characterization plots. All of which obviously do not match JA's characterization at all.

    Based on that CHORD presentation, I agree with JA's comment about not being able to "replicate the manufacturer's own measurement at lower levels". Indeed, JA's characterizations shows a piece of shit relative to CHORD's claims.

    In the face of so much fail, relative to CHORD's claims, I guess that's something to look at. But that would only tell me something I already know: That their overkill FIR filters are linear phase, go low in frequency, and behave relatively well if we can ignore the massive delay they are likely to produce.

    Please note I'm qualifying the CHORD DAVE as a piece of shit relative to their alleged characterization and marketing claims. In more absolute terms, I don't think the CHORD DAVE is a miserable piece of shit. Likely other things down the chain will affect performance more than the DAC itself.
     
  15. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,463
    Trophy Points:
    113
    Location:
    Winnipeg
    I felt like JA was artfully dodging with his second reply...

    ultra: hey, isn't this kinda deceitful the way you've buggered the numbers?
    JA: nah dude, you're not supposed to look at the actual numbers, but like, the shape of them man
     
  16. Kattefjaes

    Kattefjaes Mostly Harmless

    Pyrate
    Joined:
    Sep 5, 2016
    Likes Received:
    4,521
    Trophy Points:
    113
    Location:
    London, UK
    @ultrabike - the delay that they absolutely do produce. Even the Mojo carries a warning from Chord that it's unsuitable for live use.

    (Can't find the Chord warning, sod's law, but this Moon audio FAQ also discourages live use..)
     
    Last edited: May 26, 2017
  17. ultrabike

    ultrabike Measurbator - Admin

    Staff Member Pyrate MZR
    Joined:
    Sep 25, 2015
    Likes Received:
    8,960
    Trophy Points:
    113
    Location:
    Irvine CA
    LOL! Yup.

    164000 taps is kind of long. I think JA said his AP could not fit the full response.
     
  18. Kattefjaes

    Kattefjaes Mostly Harmless

    Pyrate
    Joined:
    Sep 5, 2016
    Likes Received:
    4,521
    Trophy Points:
    113
    Location:
    London, UK
    Yep. that's bordering on "Riverdance".

    Probably a pain if you're watching video of a concert or something, you'd need to tweak the offset to make it line up.
     
  19. winders

    winders boomer

    Banned
    Joined:
    Feb 13, 2017
    Likes Received:
    1,596
    Trophy Points:
    113
    Location:
    San Martin, CA
    Well, shouldn't JA be called to task for not publishing apples to apples graphs? In other words, for being deceitful?
     
  20. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    JAWTF.png
    [​IMG]

    This was totally cheating and setting up the measurement on JA's part of make the DAVE look good. The volume should have been set to a realistic output level at 0dbFS (max output) to take the -90.31db 16-bit measurement.
     
    Last edited: May 28, 2017

Share This Page