DAC Blind Test Series

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by purr1n, Feb 28, 2019.

  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
    Here is my triple blind setup. Each DAC has its own pot (JBL Nano Patch). These pots are used to match levels. The outputs from these pots are then plugged into a switch. The back of the switch is occluded so that I cannot see the wires. While I am out of the room, my wife or kids plug or change the outputs from the pots into either the #1 or #2 outputs without knowing which is which (it's a rat's nest back there). Going forward, we if feel this is not enough, we can occlude the back of the DACs and pots to protect against telepathy. Just in case one of my family members knows which wires came from which DAC and wants to me buy a specific DAC. Then they can beam "#1" or "#2" into my head to influence me. We wouldn't want that happening. I'm not against quadruple blind tests.

    The procedure will be as follows:
    1. I will learn the sound signatures of the DACs using "flawed" sighted tests at my leisure.
    2. The test will be set up
      1. I will be able to play any recording which I choose.
      2. I will be able to select either button #1 or button #2 at any time that I wish.
      3. I will then "guess" which button number is which DAC.
        1. I will also take notes that how #1 differs from #2 or vice-versa.
        2. This will be confirmed by my pressing the mute button on the appropriate pot.
    IMG_20190228_121300.jpg
     
    Last edited: Feb 28, 2019
  2. 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
    Up for battle: the EC Designs MOS16 and Massdrop Airist RDAC pre-production. A modified Marantz CD Player is used as the source. The EC Designs is fed Toslink because that's all it takes. The RDAC takes SPDIF over coaxial. The amplifier is a modest "Nobsound" JL Hood 1969 at 10W per channel. The speakers are JBL4698B. Pots are adjusted so that outputs are within 0.1db SPL at 1kHz on an FFT using a Dayton UMM-6. (Hard to get it better matching with mechanical devices and there is always some drift).


    Marv Tests:
    1. 2/28/2019 1:06pm: #2 is the MOS16. This is on account of its deeper and more expansive stage. RESULT: CORRECT
    2. 2/28/2019 1:40pm: #1 is the MOS16. k.d lang's vocal is a little further back. Also wavers front-to-back. Less stable. RESULT: CORRECT
    3. 2/28/2019: 2:01pm: #2 is the MOS16. Pharell's voice is further back and blended in with the rest of the other instruments. Bass is slightly mushier and less punchy. RESULT: CORRECT
    4. 2/28/2019: 2:12pm: #2 is the MOS16. This is too easy. RESULT: CORRECT
    5. 3/11/2019: 4:36pm: Last test before I ship out. #1 is the RDAC. RESULT CORRECT
    Julie Tests:
    1. 2/28/2019 1:49pm: #2is the RDAC Her voice is upfront. RESULT: CORRECT
     
    Last edited: Mar 13, 2019
  3. Elnrik

    Elnrik Super Friendly

    Pyrate
    Joined:
    Jan 10, 2017
    Likes Received:
    8,973
    Trophy Points:
    113
    Location:
    Denver CO
    Home Page:
    If you're not doing the listening portion while blindfolded in a pitch black room then it's not a true triple blind test, and thus is invalid, and I cannot recommend.

    :p
     
    Last edited: Feb 28, 2019
  4. 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
    You forgot about the Magneto's helmet. You wouldn't want me feeling out the electrons in metal of the wires and DACs.
     
  5. lm4der

    lm4der A very good sport - Friend

    Pyrate
    Joined:
    Sep 26, 2015
    Likes Received:
    1,461
    Trophy Points:
    113
    Location:
    Seattle, WA
    I don't want to add to the burden, but if you have any interest, it might be fun to do your blind taste test with the Topping D30 vs Schiit Modi. Just for giggles. And for Science!
     
  6. 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
    Unfortunately, the Topping D30 was so bad that cosmic forces insisted that it be stolen from someone's porch.
     
  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
    BTW, I'm stopping this test for today. This is dumb. Anyone can hear the difference. I'll test my kids tonight.

    A few takeaways:
    1. Practice makes perfect. Both Julie and I did some practice, just one or two rounds.
    2. Once you know what to look or listen for, it's easy.
    3. Using tracks not used during practice does make it harder, but the characteristics of the DACs do translate regardless of recording.
    4. No need to rush or try hard to hear the differences. The process is easier when one is relaxed and takes the time to go back and forth.
    5. SPL matching makes it much easier to distinguish the differences. Variable volume positions used during sighted tests makes it harder to distinguish the differences.
    Here are the notes, roughly in order of magnitude of difference / distinguishable factor:

    RDAC
    • Stage is upfront. Vocals are particularly up front with the instruments in the back. Soundstage is slightly more compact with most instruments between the centerline of the speakers and rarely venturing outside.
    • The overall tone is thicker with heavier midbass.
    • Less resolving, a small amount of low-level information being lost in the decays.
    • Slightly more rolled up top and bottom. Less air and less of that low sub-harmonic thud in the lows.
    MOS16
    • More expansive stage. Wider presentation with sounds emanating from beyond the width of the speakers regularly. Deeper placed vocals that are blended in layerwise with the other instruments which are also placed deeper. Image of vocals wavers front to back and isn't as stable depthwise.
    • The overall tone is less thick. Less midbass.
    • More extended both ways. More air, more low bass just below the threshold of hearing.
    • Slightly softer and mushier transients, bass slightly blurry.
    • More resolving of low-level information
    P.S. Again, I like these blind tests. There's less going back and forth to try to hear the differences because SPL matching makes the differences that much more discernable.
     
    Last edited: Feb 28, 2019
  8. mscott58

    mscott58 Friend

    Pyrate
    Joined:
    Oct 13, 2015
    Likes Received:
    2,028
    Trophy Points:
    113
    Location:
    Philly - Yo
    And here I thought the proper technique was to have one's eyes gouged out by the priests of the Temples of Syrinx...
     
  9. Mr.Sneis

    Mr.Sneis Friend

    Pyrate
    Joined:
    Oct 2, 2015
    Likes Received:
    1,321
    Trophy Points:
    93
    Location:
    Phoenix, AZ
    Suggests auditory memory is better than we think? We have more trained ears than non audio nerdites?
     
  10. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    That's just mean! No one should have to suffer the D30.:eek:
     
  11. 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
    Couple of thoughts:
    1. Audio nerdites are hypocrites. They accuse others of not doing blind tests when they themselves have not done any such blind tests themselves, much less sighted tests.
    2. Some people are actually deaf. As in they really can't hear the differences.
    3. Most people can hear the differences, but they don't care.
    4. Audio nerdites don't understand the limited scope of measurements. The state of the typical set of audio measurements really aren't any more exhaustive than a 0-60mph times and max G on a skidpad. As such, they rely on these measurements as the ultimate determiner and allow expectation bias to get the better of them
    I'm gloating at the last point. I've said so many times that expectation bias goes both ways.
     
  12. spoony

    spoony Spooky

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    651
    Trophy Points:
    93
    Boom!
     
  13. murphythecat

    murphythecat GRU-powered uniformed trumpkin

    Pyrate Banned
    Joined:
    Apr 4, 2016
    Likes Received:
    1,201
    Trophy Points:
    93
    id be interested how the MOS16 compares to Yggdrasil A2, Dangerous, Solaris, ect... any plans of including those "flagship" dac in the blind test.
     
  14. Josh83

    Josh83 Friend

    Pyrate
    Joined:
    Mar 1, 2017
    Likes Received:
    1,375
    Trophy Points:
    93
    Location:
    Charlottesville
    Number one is so crucial. It makes no sense to me when people proclaim equipment sounds identical following a blind test when they didn’t bother to learn how each piece of equipment sounds with sighted listening first. It seems to me that the people who skip that step are the people who are determined to use blind testing for the sole purpose of claiming all gear sounds the same.
     
  15. 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
    The people who are determined to use blind testing for the sole purpose of claiming all gear sounds the same:
    1. Rarely conduct tests.
    2. And if they do, provide very little of what their procedures, are other than a one-liner "I did a blind test and they sounded the same", of which their minions all nod their heads in agreement.
    3. Are heavily weighted down by expectation bias. Even if with the best-structured tests designed to help them hear the differences, nothing can stop them from either intentionally or unintentionally not hearing the differences. This is difficult to counter unless they were strapped down to a chair and threatened with a severe pummelling every time they got the wrong answer.
     
    Last edited: Mar 1, 2019
  16. skem

    skem Friend

    Pyrate
    Joined:
    Nov 1, 2017
    Likes Received:
    1,911
    Trophy Points:
    93
    Location:
    Charles River
    4. Because they believe it all sounds the same, they are happy with mediocre or crap gear in their chain—and thus when testing a quality component can’t hear the benefits because the rest of the chain is still crap.
     
  17. 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
    The MOS16 is priced appropriately and a better deal than the low-end Metrum DACs such as the Amethyst. I mention the Amethyst. because of its similar approach and sound: resistor ladder, simple functionality, NOS, etc. And the cost of the MOS16 allows enough room to get a USB to Toslink converter and still come out ahead of the Amethyst.

    First, the Dangerous Source Convert-2 is a totally different sounding DAC, not just from the MOS16, but also from most other DACs out there. Let's just skip that one.

    The tonal balance and of the Crane Sound Solaris is almost opposite to the MOS16 or classic R2R, discrete R2R sound. I'd give to edge to resolution, reproduction of low-level information to the Solaris. This isn't surprising given how resolving the new AKM chip is. In terms of transients, neither are hard-hitting, but the Solaris is a bit cleaner with less blur in the bass. The Solaris has just a slightly deeper stage with much more precise imaging with some depth layering. The MOS16 has no front-to-back layering in the soundstage. Finally, like the Convert-2, it's hard to compare Solaris and MOS16 because their tonal balance is so different. The MOS16 has got that denser thicker tone and timbre whereas the Solaris is light on its feet and can suffer from digital nasties on the wrong system.

    Coming back to the Amethyst (and also Holo Spring Level 3) in terms of soundstage, the MOS16 soundstage is deeper. Among these three DACs, there is no front to back layering localization of instruments of the stage. It's all on a flat plane more or less, just that MOS16 stage is deeper. Vocals on Solaris and the Schiit DACs, depending upon the recording, can be placed in front of the band, with the band, in the back, etc. On the MOS16 and the other discrete R2R NOS ladder DACs, the vocals are all on the same plane regardless of recordings. The MOS16 despite the slightly blurry bass, seemed to be a little bit more immune to congestion compared to the Amethyst, seemed to have more air, and also was more resolving. These last three aspects seemed to improve over time up to four hours on. The good thing is that with a proper charge, the MOS16 seems to last forever. I had it on all day yesterday - by accident.

    Again, the resolution of the MOS16 is impressive given its architecture.
     
    Last edited: Mar 1, 2019
  18. 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
    As far as the Yggdrasil and Gungnir (A2), I don't think the MOS16 is competitive. Not at least with speakers. The Schiit filter on the A2 DACs really takes them to another level in terms of space. I've been noticing this even more now than before when I only had the Oris horns. Eight-inch drivers inside big horns don't image that well - they do OK. The smaller butt-cheek tweeters on the JBL 4698b image on a whole 'another level. The dimensionality and resolution from the Schiit DACs get rather scary where I am hearing deep into the mix, probably more than the mixing engineer intended! I can recognize different chains in the recording process, microphones, boards, effects, how it's mixed-down, etc. Voices, effects, instruments float in space, all over the place, front, back, left, right, up, down, with supreme precision, never wavering, never blurry. Also in terms of attacks and transients, the MOS16 doesn't come close to the Convert-2 or Gungnir MB A2. The Convert-2 is just on a higher perhaps exaggerated level, for better or worse. Anytime there is a rapid succession of 16th notes in tom or bass drum work or bass line fills, the MOS16 kind of mushes everything together. It's not horrible in this regard, there are far worse, but let's just say tight articulate bass or punch and impact (entire range) are not the MOS16's specialty. And in any event, there are certain things that the Schiit DACs destroy most other DACs (regardless of cost) - most of this not being realized unless smaller point-source type tweeters are used.

    In terms of tone, the MOS16 does sound reasonably extended (the Convert-2, Solaris, and Schiit DACs all beat it slightly) and is better than most peers in its price range. There is a bit of mid-bass emphasis, but it's still not quite as thick as the RDAC, but it is more than the Gungnir MB A2, which can be kind of cool sounding in the wrong system (I pair the Gungnir MB A2 with a 300B no-feedback SET amp with the JBLs, or EQ the bass up on the Oris / Lowther / Eminence OB).
     
    Last edited: Mar 1, 2019
  19. 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
    The MOS16 represents a very good value for the sound it offers despite it only having a Toslink input. I think it's a very strong entry in the discrete R2R ladder deals. Possibly one of the few viable choices near the price point*, considering the delays on the RDAC and poor value of the lower line Metrum DACs (which effectively take the Metrums out of the equation).

    Here is the breakdown:

    + surprising resolving power for discrete ladder
    + deeply placed and expansive soundstage
    + well extended at the extremes, lows, and highs
    + a nice touch on the mid-bass and upper-bass warmth - not overly done, no warmpoo
    + classic R2R timbre - no digital fuckery
    + long battery time

    - battery needs to charge
    - toslink only
    - lack-of-depth / flat / no soundstage layering
    - unstable imaging (depth)
    - a tendency to blur fast bass notes
    - lack of punch down low and pop up high

    * I know some butthead is going to mention another "competitive" DAC, but it's going to be 1100 euro instead of 363 euro. Please don't do this. Literacy is your friend. Sure the MOS16 might need a converter for toslink, but get one of those Guzzard deals or even better, a UTOS for 99 euro.
     
    Last edited: Mar 1, 2019
  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
    FYI, a level matched blind test was conducted for Gungnir A2 vs MOS16. It was pretty obvious which DAC was which.

    I brought out a 300B (no-feedback) SET for this in place of the JL Hood solid-state amp.
    IMG_20190301_124517.jpg
     

Share This Page