Audio-GD Singularity 19 DAC - Measurements

Discussion in 'Source Measurements' started by Hands, Mar 23, 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
    Last edited: Mar 31, 2017
  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
    Right, I just thought your Yggdrasil results looked less broad, but going back it's pretty much the same thing. Looks like both are limited by FFT size.
    With a super long FFT size you could probably see some broadening of the fundamental with more jitter, sort of how MSB visualized it here:
    I think this is a very zoomed in look at the fundamental of the 11.025kHz tone, but I could be wrong.
    [​IMG]
    You can sometimes see the effect in some of the stereophile measurements. I've also seen the same thing for turntables, but they will obviously look way worse.

    It was expected with the AGD S19, but I'm actually surprised the Modi 2 looks as bad. Might be interesting to do it with other DS DACs to see if there are different patterns with different chips. I doubt it's the filter.
     
  3. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    I'm not sure if I'm more, or less, anxious to get my hands on this thing now. Those measurements aren't any more inspiring than those @Hands posted. Maybe I need hearing-insurance before listening to this thing.

    It does make me want to get around to adding a useful measurement capability/practice though. However, if it comes in this direction still, @atomicbob expressed a desire to measure it, so we can probably take care of that while its in the Seattle area as we're both located here - assuming that works for @Marvey (not sure if you're just thinking of sending it back to the manufacturer and recouping costs - don't know what their return-policies are).
     
  4. shipsupt

    shipsupt Admin

    Staff Member Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    1,119
    Trophy Points:
    93
    Location:
    Bay Area CA
    Is this signal or noise?
     
  5. 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
    @Torq: I will ship it to you next Tuesday. After warm-up, it does sound better. I want to continue the warm-up process for a few more days, flipping bits.

    I'm using the QA400 with QA190 diff probe (best results) and the Focusrite 2i2. The Focusrite needs resistors to reduce the output from DACs. @Hands uses XLR attenuators, but those tend to be designed for microphone impedances, so they will attenuate much less than what they say.
     
  6. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    Thank-you for the measurements @Marvey . When @Torq receives the unit in our region he and I can coordinate so I can provide dScope data for comparison.
     
  7. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    No rush at all, was just curious what was going to happen to it.

    Lovely, I'll take a look at the QA400 (a cursory glance suggests it's a Windows only unit, which won't work for me, but I'll do some digging). Thanks!
     
  8. Psalmanazar

    Psalmanazar Most improved member; A+

    Pyrate Slaytanic Cliff Clavin
    Joined:
    Sep 27, 2015
    Likes Received:
    5,345
    Trophy Points:
    113
    I wish I was joking but I'm not.
     
  9. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Pyrate MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    12,285
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    Thanks, @Marvey. Some differences in our results, some similarities, but that's to be expected based on hardware, including the source feeding the DAC in this case (ugh), and methodology. BTW, I only need to use an attenuator when using XLR output on some DACs. Single ended usually doesn't clip the 2i2 (first gen) or my Creative SB1240. I usually have to raise the gain when measuring on 2i2 via single ended, or if the attenuator brings XLR too low, so that can affect the noise floor.

    Either way, my DAC measurements are taken with cheap-but-decent gear, so I hope everyone knows not to take them too seriously. Just in case someone didn't catch that implication.

    But it's good to see the S19 has seemingly verifiable performance issues in measurements from different individuals. You don't need high end ADCs to see the faults in the DAC...And still doesn't change that I was able to pull decent results from the Holo Spring, new Metrum gear, and so on.

    Warm up did improve the sound, but it didn't move it out of the "worst DAC I've heard" slot. I had it running 24/7 for about 5-6 days before my "review."
     
  10. Hands

    Hands Overzealous Auto Flusher - Measurbator

    Staff Member Pyrate MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    12,285
    Trophy Points:
    113
    Location:
    Colorado
    Home Page:
    Oh, and, @Marvey, your -90dB tests look a good bit better than mine, 16 and 24 bit. I wonder why I got so much noise in mine? Haven't had that issue with other DACs (usually).

    I just can't believe the SPDIF input was so broken with my D100 source. Unbelievable, really. In fact, it was so screwy, I couldn't run a Rightmark test with the S19 using SPDIF. The results got screwed up so much that RMAA had no idea how to interpret half of the readings. The -6dB crosstalk result was one oddity! I know you aren't a fan of RMAA, but I still found it funny.
     
  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
    Maybe the S19 wasn't fully broken in yet for the -90db tests? LOL.

    I went SE from the S19 into XLR on the 2i2. Balanced inputs / probes tend to have less random crap, but sometimes at the expense of a higher noise floor.


    I still need to hear this DAC on my most resolving setup to make an accurate determination of how it sounds subjectively. At best, it's meh. At worst, it's a disaster. At least right now. Would like to give it as much benefit of doubt as possible, so still baking it in, flipping bits, despite that it's been broken-in by AGD and then you for at least 400 hours now.
     
  12. Azimuth

    Azimuth FKA rtaylor76, Friend

    Pyrate
    Joined:
    Nov 9, 2015
    Likes Received:
    6,833
    Trophy Points:
    113
    Location:
    Nashville, TN
    Home Page:
    Remember, it is not just burn in of the unit in total, it is how long it has been on. Temperature fluctuations affect the resistors in the ladder greatly. Even the PCM1702 spec sheet calls for one minute warm up time, and those resistors at embedded in the chip and laser trimmed. Didn't many people say their Schiit MB DACs taking 48/72 hours to settle down. I remember when Bifrost Multibit hit and saying that 4 hours shied improvement, 24 hours was better, 48 hours is best. With something with MSB to LSB ladder SMD 0.1% resistors...you know what I mean.

    I also think having the DAC itself on a separate board is no good. I remember recently reading DIY DAC forums that it actually is not good practice to isolate the boards as the leads, ribbon cables, and whatnot can cause issues. Not saying that is or has anything to do with the abnormalities, just thinking out loud.

    This is also their first generation. I have still would have liked to see something a bit more promising. It just goes to show how hard it actually is to build a discrete DAC that performs well.
     
  13. Psalmanazar

    Psalmanazar Most improved member; A+

    Pyrate Slaytanic Cliff Clavin
    Joined:
    Sep 27, 2015
    Likes Received:
    5,345
    Trophy Points:
    113
    Why the hell commercially produce the prototype except to prey on the wallets of fanboys ignorant to the fact that AGD make shitty gear?
     
  14. 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
    It's been our experience that warm up makes a difference. Less so with burn-in.

    The dumbass thing is that AGD fanboys and even AGD apparently subscribes to the burn-in thing. Regardless, we've done both. By the time I intently listened to it, the S19 easily had over 150 hours of burn-in (not counting factory burn-in) and two to three days of warm up. I'm sure @Torq added many more hours.

    Also, in our experience, warm-up or burn-in has never transformed a turd into something mediocre, or something mediocre into something good.

    Burn-in is mostly disingenious bullshit from folks trying to convince others that they should like gear that they don't. It would be a lot more honest to tell people to spend more time to familiarize themselves with the gear or try other components to obtain better synergy.
     
  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
    I remember a statement from AGD to this effect as to why the PCM1704 chips were soldered to the board on their M7 DAC.

    As for the discrete parts not measuring to snuff (our measurements had three magnitudes more distortion than AGD's), it could be possible that AGD received a shipment of resistors that had 1% parts instead of the 0.1% parts. You never know in China with all the fake / relabeled opamps. AGD could likely had been a victim themselves. But as I've said before: not our problem.
     
    Last edited: Apr 15, 2017
  16. Walderstorn

    Walderstorn Friend

    Pyrate
    Joined:
    Mar 20, 2016
    Likes Received:
    1,904
    Trophy Points:
    113
    Location:
    Europe
    That would be as wrong as it just being a bad piece of gear. I mean, come on, their QC already seems to suck donkey balls if they decided to release this as a final product, so if it was a bunch of bad resistors that would just mean it also sucks dirty donkey balls.

    It may be possible but with what they charge for each piece of equipment and with the known cheap hand labor in production this all just screams "just profits" to me. I hope I'm wrong but those are my thought on it.

    I'm just glad people like you guys do this so less informed people like myself avoid this type of (very expensive) mistake.
     
  17. Azimuth

    Azimuth FKA rtaylor76, Friend

    Pyrate
    Joined:
    Nov 9, 2015
    Likes Received:
    6,833
    Trophy Points:
    113
    Location:
    Nashville, TN
    Home Page:
    Good to know. I was only posting my comment due to what I have heard from @baldr and Jason about MB DAC's settling after a few hours. Jason mentioned this for any MB DAC, regardless of architecture and to a certain extent DS DAC's. And also, my earlier comment on what the the PCM1702's datasheet mentioned about warm up time.

    So the fact you cannot measure any changes could be confirmation bias or just whatever it is, this warm up/settling whatever, is something difficult to measure. I read many forums on burn in on equipment and even down to filter capacitors and burning paths in the caps, but for DAC's I know it just a little different as it is more of a warm up rather than a burn in deal. Still odd that measurements don't change over time for any DAC's. The AGD has many other issues going on for sure.

    I think it is more than just bad resistors. It looks like a bad design altogether. The caps are not holding the voltage at all either. That trace is all over the place. Looks like some bad interference or HF noise going on.

    If it did not sound so bad, I would say maybe your scope equipment interfering, but that ain't it. It is just a bad design with some possible bad parts. I am not smart enough to speculate any further. Just maybe discrete isn't better for this application. He did make a full discrete DAC - not anything off the shelf.

    They only have one guy in their A&D department. Why not sell it as a DIY kit and let people experiment with it? Maybe someone with a better receiver chip or better better implementation can make something better. I think this might push the technology forward instead of being proprietary and in a flashy box and selling it 10 times more just because it's discrete. However, like I said, given the measurements, I am starting that discrete digital circuits may not not actually be better and cause more problems than solve them.

    Damn straight! I need an D-scope myself. I am sure many companies would rather not this information be published. I was kind of interested in the ADG unit - not anymore. I'll keep enjoying my vintage PCM1702 Denon DAC.
     
  18. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    The Singularity S19 is now in the hands of @atomicbob ...

    Hopefully it doesn't leave him sonically traumatized!
     
  19. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    It will take a bit longer than my usual time for such measurements as this DAC keeps changing. The corporate day gig will also be consuming extra resources until I get a project over a hurdle this week. Hope to be finished with the S19 by end of this coming weekend.
     
  20. Grahad2

    Grahad2 Red eyes from too much anime

    Pyrate
    Joined:
    Jan 26, 2017
    Likes Received:
    1,162
    Trophy Points:
    113
    Location:
    Singapore
    I suppose the charts aren't going to look like this then
    [​IMG]
     

Share This Page