Soncoz SGD1 ASR Measurement Discrepancies

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by GoldenOne, Jun 17, 2021.

  1. Pancakes

    Pancakes Friend

    Friend
    Joined:
    Aug 13, 2020
    Likes Received:
    878
    Trophy Points:
    93
    Location:
    Atl
    Please don't stop doing what you're doing. It's the #1 reason I support on Patreon.

    This is not to say that others' measuremnts are useless. Rather, I appreciate your insights, attention to detail, professionalism and, transperancy.
     
    • Like Like x 9
    • Epic Epic x 1
    • List
    Last edited: Jun 19, 2021
  2. yotacowboy

    yotacowboy McRibs Kind of Guy

    Friend
    Joined:
    Feb 23, 2016
    Likes Received:
    6,890
    Trophy Points:
    113
    Location:
    NOVA
    Home Page:
    please, @atomicbob, don't take it this way! I don't believe this is the intent.
     
  3. crenca

    crenca Friend

    Friend
    Joined:
    May 26, 2017
    Likes Received:
    3,096
    Trophy Points:
    113
    Location:
    Southern New Mexico
    @atomicbob , I read @GoldenOne suggestion more as a reaction to ASR/Amir and his ignorance and sloppiness, a understandable reaction at that. As far as it being a more generalized effort more widely applicable to the hobby and industry, a previously well informed and agreed upon set/config can in theory be a useful step it seems to me there is more too it than that. Someone could still "game" this config by messing with other parameters such as cables, EMI and environmental noise, and other things I am not thinking of.

    In other words its worth discussing, but let's not pretend it's the end-all-be-all somehow. Any person, system, config, etc. will always need accountability from competing equipment and technique...
     
    Last edited: Jun 19, 2021
  4. GoldenOne

    GoldenOne Friend

    Friend
    Joined:
    Jul 17, 2020
    Likes Received:
    762
    Trophy Points:
    93
    Location:
    England
    This.

    If others did what @atomicbob is doing and documented exactly how everything is setup and configured it wouldn't be an issue at all.

    But when someone is putting out measurements and we have no idea if they're done fairly, consistently, or without questionable setup because none of that information is given, that's an issue.
     
  5. earnmyturns

    earnmyturns Smartest friend

    Friend
    Joined:
    Sep 25, 2016
    Likes Received:
    2,286
    Trophy Points:
    113
    Location:
    Palo Alto
    Home Page:
    @GoldenOne 's suggestion was well intended, but I'm wary of "foolproof" testing scripts. I work in a different field where measurements are critical, and I can't count the instances that someone (including me) screwed up, and not because they were using the wrong script. The measurement apparatus may be perfectly set up, but knowledge and practice in how it interacts with the measured system, and especially on how to control for confounders, are critical and often messed up. Measuring well is a hard-won craft, as @atomicbob demonstrates in his exemplary measurement threads here. A canned script can only create a false sense of certainty.
     
  6. Marvey

    Marvey Super Friend

    Staff Member Friend BWC
    Joined:
    Jun 6, 2021
    Likes Received:
    1,691
    Trophy Points:
    93
    LOL, there are much easier methods to bork scripts:
    1. Use a bad cable.
    2. Put bad / noisy switcher close to DUT or a cable connected to DUT.
    3. Parallel 60Hz mains with USB, line cables, etc.
    4. Improperly ground.
    5. Sprinkle solder dust (from the soldering iron) over the PCB of the DUT.
    6. Be lazy, do things too quickly, refuse to second-guess self, etc.
    I'm sure @atomicbob can commiserate: so many times I get funky results and take time to troubleshoot where it's often something dumb.

    BTW, $2 for the first person to tell me what's wrong here:
    [​IMG]
     
    Last edited by a moderator: Jun 20, 2021
  7. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    16,070
    Trophy Points:
    113
    Location:
    On planet
    There are many external influences that can negatively impact measurements. When looking at -140 dB on a DAC with an output of 0 dBFS = +13.5 dBu we are measuring around 350 nanovolts .... 350 billionths of a volt.

    Why would any sane person want to measure at 2Vrms for an IEM???!!! At that level
    Etymotic ER4P-T will produce 128 dBSPL
    Westone UM2 will produce 141 dBSPL !!!!!! before is expels the magic smoke
     
  8. Beefy

    Beefy Almost "Made"

    Joined:
    Apr 10, 2021
    Likes Received:
    1,086
    Trophy Points:
    93
    Location:
    Canada
    Well, it seems a bit odd that he is using the balanced input.
     
  9. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    16,070
    Trophy Points:
    113
    Location:
    On planet
    I don't intend to stop measuring but I did have a bit of a snit yesterday. My apologies for adding unnecessary drama.
    Consider how it might feel to have put in three decades of working methodically, to very high standards, producing AES, IEC, ANSI, etc compliant data which is scrutinized by regulatory bodies such as FDA, CSA, TUV etc. Applying the same methods and rigor to audio components and publishing enormous quantities of carefully vetted data over the recent several years on forums. But all the excitement is about the most expensive analyzer available with considerable discussion concerning a standardized script for the APx555 players. It kind of stings.

    Now consider that one script won't be sufficient. There are too many variations necessary to accommodate when making accurate assessments of the various components.

    Examples for DACs
    0 dBFS = ??? dBu / dBV / Vrms
    for which filter (they do deviate)
    SE or Bal output
    which input USB native driver / USB ASIO / spdif / AES / opt

    Examples for Amps
    which input SE or Bal
    which output SE or Bal
    which load
    what level
    I prefer 0 dBu, 0.775 Vrms as it reflects a reasonable level for 100 dBSPL on many headphones which is very loud.
    When more in-depth, the also include -10 dBu and -20 dBu which are 90 and 80 dBSPL on those same headphones respectively.

    Already there are many variations not including the DA Lineup for 0 dBFS.
     
    • Like Like x 8
    • Epic Epic x 3
    • List
  10. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    16,070
    Trophy Points:
    113
    Location:
    On planet
    @earnmyturns has been there, done that, got the t-shirt. You know the drill and the potential consequences.

    After every script runs, I still review the measurements and graphs looking for outliers that may indicate a problem Remember a little while back I discovered one of my scripts had somehow altered the IMD and was setting the 19+20KHz using an offset rather than a ratio for the two sine levels and created an improper stimulus. It only made a few dB difference in the IMD but it was an error just the same, requiring correction, which I boldfaced in the updates for the technical measurements.

    In my day job medical device alarm levels are critical too. Especially the ones indicating a patient is about to die. Or the ones informing rescuers to analyzing now, stand clear or stand clear, push to shock on AEDs.
     
    Last edited: Jun 20, 2021
  11. Marvey

    Marvey Super Friend

    Staff Member Friend BWC
    Joined:
    Jun 6, 2021
    Likes Received:
    1,691
    Trophy Points:
    93
    Standards first, then automation.

    Measurements need review with human eyeballs. When human eyeballs see that things are off, then we explore, dig deeper. One of the things which really bothered me about ASR was that the guy had no clue when things were off.
     
  12. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    16,070
    Trophy Points:
    113
    Location:
    On planet
    I have an adapter to use allow single ended output to connect to the analyzer Balanced input. It sometimes makes a difference over using the SE input, especially concerning mains noise.
     
  13. Marvey

    Marvey Super Friend

    Staff Member Friend BWC
    Joined:
    Jun 6, 2021
    Likes Received:
    1,691
    Trophy Points:
    93
    200-kohm load. It's unclear because ASR usually uses the built-in 300-ohm or 600-ohm load on the AP - and I've never seen a external breakout box or wires with resistors. Also, I don't think I've ever seen ASR use balanced in for SE amps on the AP. (I got worse results with two different pairs of SE-> balanced connectors into the AP input, so I'm sticking with a dedicated unbalanced cable, which among a half a dozen others I tested produced the best and most consistent results. Don't laugh. The cable came out the box of a Sony CDP and has BNC adapters on it. I'll order RCA-BNC cables from Benchmark next and see if those beat this.)

    The point is lack of consistency and sloppiness. And mostly the attitude of slam bam and done, rather than going back and correcting. Humans make errors. That's OK. But not going back and correcting is irresponsible, especially if your schtick is 1db of SINAD and 0.1db of linearity.
     
    Last edited by a moderator: Jun 20, 2021
  14. Beefy

    Beefy Almost "Made"

    Joined:
    Apr 10, 2021
    Likes Received:
    1,086
    Trophy Points:
    93
    Location:
    Canada
    Oh, I'm sure. But that sounds like the sort of thing that you would specifically report. I can't imagine it would be the default to just run every unbalanced amp into the balanced input. But I'm a different sort of scientist, so I just don't know.
     
  15. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    16,070
    Trophy Points:
    113
    Location:
    On planet
    This is why setup description and picture are important. Without we are missing significant information about the measurement. The measurements might as well be from golden toilets tracking monkeys flying out butts.
     
  16. gepardcv

    gepardcv Almost "Made"

    Contributor
    Joined:
    Nov 20, 2015
    Likes Received:
    381
    Trophy Points:
    63
    Location:
    Terra, Sol System
    Publishing the raw analyzer output data for other people to play with wouldn’t hurt, either.

    I vaguely remember asking ASR about this around the time of the Yggdrasil measurements drama (either on the ASR forum or Reddit), and he refused. I think the excuse was bandwidth cost. The amount of bullshit in that justification was my first clue that ASR engages in posturing rather than science. (There’s so much free file hosting out there. Including with his old employer. And I’m sure he can get sweetheart pricing on Azure.)
     
  17. Roget

    Roget Acquaintance

    Joined:
    Nov 17, 2018
    Likes Received:
    39
    Trophy Points:
    18
    Thank you Golden and Mr. Purrin for your work. Atomic Bob too. This is concerning. I think all gear at Audio Science review should be peer reviewed by at least one of you. Honesty is important from manufacturers. I trust Audio Science Review. I do not trust all manufacturers. I almost bought this DAC.
     
  18. purr1n

    purr1n On vacation

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    79,360
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Tell your buddy this:

     
  19. Marvey

    Marvey Super Friend

    Staff Member Friend BWC
    Joined:
    Jun 6, 2021
    Likes Received:
    1,691
    Trophy Points:
    93
    Keep the fight. Sometimes I wonder why the heck I'm doing this myself. It seems few people appreciate the nuanced and research measurement aspects of it. The fact is, we can bork the hell outta our measurements, and just write a summary. A random number. A pose of a doll. And people will focus on that. It's the nature of the universe. We need to remind ourselves that there are the unspoken few that will appreciate what we do, and that should be good enough.
     
  20. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    16,070
    Trophy Points:
    113
    Location:
    On planet
    It is very good to question if the information isn't specifically stated. That is part of the scientific method.
    Marketing is another story ...
     

Share This Page