Audio-GD Singularity 19 DAC - Measurements

Discussion in 'Source Measurements' started by Hands, Mar 23, 2017.

  1. Senorx12562

    Senorx12562 Case of the mondays

    Pyrate
    Joined:
    Oct 26, 2015
    Likes Received:
    3,274
    Trophy Points:
    113
    Location:
    Bird-watcher's paradise
    Assuming the demand for r2r/nos dacs continues to increase (although I think that outside of the market probably well represented by the membership here, even the concept of an external dac is pretty esoteric i.e., it's a small part of a miniscule market) there will appear a disproportionate number of poor ones. I would think that the cost floor to design and build a good dac of this sort is much higher than one using an off-the-shelf chip.

    But attempts to build it for less will result in what you see here. And widespread and contagious cognitive dissonance, expectation bias, and confirmation bias will result in many gushing reviews. Many hipsters will buy them to listen to mp3s at 128kbps. It will share pride of place with the (dusty) turntable next to it. I've already seen one ad (for what, I wish I could remember ) comparing the advertiser's r2r dac to vinyl, as in the way digital used to be converted when music was still good or some such crapola. Don't start without me, I just have to freshen this up.:drunk:
     
  2. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    The measurements aren't so bad to suggest the unit is defective.
     
  3. Walderstorn

    Walderstorn Friend

    Pyrate
    Joined:
    Mar 20, 2016
    Likes Received:
    1,904
    Trophy Points:
    113
    Location:
    Europe
    So its hopeless, how can they sell this? Makes me wonder about their "testers".
     
  4. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    • They have a rabid fan base that would gobble up anything and everything that they make
    • Discrete R2R ladders are "en vogue"
    • Randos with really bad gear or limited experience, making pronouncements of awesomeness
    • FPGAs are also en vogue.
    Making an R2R ladder for a 16 bit audio DAC using 0.1% parts is nuts. Take for example the Soekris which uses the 0.05% parts at the low end and the 0.01% parts at the high end. Or the TOTAL DAC guy who uses up to 12 of these ladders (using 0.01% parts) to average out the errors in order to get better INL.

    Now, the the screwy INL can be corrected with calibration, but I seriously doubt this is done by AGD.
     
    Last edited: Mar 27, 2017
  5. Grahad2

    Grahad2 Red eyes from too much anime

    Pyrate
    Joined:
    Jan 26, 2017
    Likes Received:
    1,162
    Trophy Points:
    113
    Location:
    Singapore
    You missed the flavour du jour FPGA too. S19 seems to be a collection of buzzwords.
     
  6. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    LOL. Fixed. Edits made to my prior post.
     
  7. Psalmanazar

    Psalmanazar Most improved member; A+

    Pyrate Slaytanic Cliff Clavin
    Joined:
    Sep 27, 2015
    Likes Received:
    5,345
    Trophy Points:
    113
    AGD fans are mostly lofi and uncritical listeners, id est ricer pc gamers, music as background noise guys, low level listeners, "MP3s ARE TRANSPARENT!", anime fans, and indie Macbook music recorded through a Focusrite or Apogee Duet guys.
     
  8. SSL

    SSL Friend

    Pyrate
    Joined:
    Nov 12, 2015
    Likes Received:
    1,239
    Trophy Points:
    93
    I think we get the picture by now.
     
  9. landroni

    landroni Friend

    Pyrate
    Joined:
    Aug 12, 2016
    Likes Received:
    1,164
    Trophy Points:
    93
    Up to which levels do you define "low level" listeners?
     
  10. Walderstorn

    Walderstorn Friend

    Pyrate
    Joined:
    Mar 20, 2016
    Likes Received:
    1,904
    Trophy Points:
    113
    Location:
    Europe
    That's what I don't get, I should be on the same boat but I (like others) have something called common sense, how can (I presume) grown men behave in such a teenage-girl-loves-boyband-way I will never understand.

    That said I hope AGD does better next time.
     
  11. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    I have never gotten the fanboism around Audio-GD's stuff either (or anything else, really). And it tickles me the most when I see people buying them, selling them very shortly thereafter - while still proclaiming how absolutely amazing they are. And then I don't know anyone that owns one that hasn't modified it on some level. The amp I heard, which I think was a "Master 9" was clean, powerful, and otherwise thoroughly sterile. Which might be technically desirable, I suppose, but wasn't a great deal of fun to listen too. I came away from that meet much more interested in the Decware Taboo III than any of the Audio-GD stuff that was present.

    But it was more their DACs that had crossed my path.

    And their DACs have always left me rather cold ... the constant call, whining, pleading, pissing AND moaning in the original DAC comparison thread I did on Head-Fi for comparisons to the Master 7 were quite funny. No matter how many times I said I wasn't that big a fan of the PCM1704 in the first place, and that telling me you'd put MORE of them in the box wasn't helping matters, the muppets kept on and on about it.

    Having seen how poor the Singularity 19 is though, measurement wise, and looking at @Hands eloquent summary of it's musical properties, I'm actually quite interested in hearing whatever turns out to be their top of the line DAC that winds up saddled with those DAM modules. I like a nice comedy show.
     
  12. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Measurements rolling in...

    AGD Singularity 19 1kHz -1dbFS.png

    I didn't want to believe @Hands measurements, but my measurements here are very consistent with his. 0.12% THD. Higher third order. Lots of harmonic products that go on endlessly at ever 1kHz interval.

    AGD Singularity 19 1kHz -1dbFS below 1kHz.png
     
  13. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    1kHz at -60db and -90db. Note that my analyzer has a higher noise floor than @Hands.

    AGD Singularity 19 1kHz -60dbFS.png
    AGD Singularity 19 1kHz -90dbFS.png
     
  14. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    19kHz signal. Should be around -1dbFS, but the S19 has some HF rolloff here, so doesn't quite get that high. The extraenous stuff seems kind of high in level. Consistent with @Hand's results, except without the pyramid sidebands (well, mine are slightly different).

    AGD Singularity 19 19kHz -1dbFS.png
    AGD Singularity 19 19kHz -1dbFS wide.png
    AGD Singularity 19 19kHz -1dbFS narrow.png
     
  15. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Two tone 19 and 20kHz. I get a lot more random spurs, but some consistency with @Hands on the IMD at 1kHz intervals. (20-19 = 1)
    AGD Singularity 19 19k and 20k.png

    Slight roll-off starting early. Seems odd for an oversampling DAC to do this.
    AGD Singularity FR.png

    Let's look closer: (ignore the jaggies - the signal is white noise)
    AGD Singularity FR 2.png
     
  16. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    60Hz -1dbFS. Similar distortion characteristics as the 1kHz signal.
    AGD Singularity 60hz.png

    Multiples of 60Hz go on forever
    AGD Singularity 60hz wide.png
     
  17. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    -90dbFS sine waves

    16 bit. As expected. There is some low frequency stuff that isn't being filtered, so the sine wave will see-saw a bit.
    AGD Singularity 19 16bit -90db.png

    24 bit. There is some low frequency stuff that isn't being filtered, so the sine wave will see-saw a bit.
    AGD Singularity 19 24bit -90db.png
     
    Last edited: Mar 30, 2017
  18. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    OK, this one is experimental. Triangle waves at -72db to get a sense of DAC accuracy. This will only get the last 4 bits or 16 values of a 16 bit ladder. (Someone correct me if I screwed up my calculations) There will be errors introduced by noise and the analog circuitry, but I still think it's worth a shot. Why triangle? LOL, it's easier to see any errors because the slope will be consistent. All measurements at 96kHz.

    Singularity 19 triangle wave -72db
    Singularity 19 triangle -72db.png

    For reference, Modi 2 (USB powered)
    Modi 2 triangle -72db.png
     
  19. Serious

    Serious Inquisitive Frequency Response Plot

    Pyrate BWC MZR
    Joined:
    Sep 28, 2015
    Likes Received:
    2,594
    Trophy Points:
    113
    Location:
    near Munich, Germany
    Do the fundament always look so broad, or is that jitter?
    Also the HF rolloff seems to start very early - something like 3kHz.

    The wiggles seem sort of consistent, so it could really be DAC inaccuracy. Wouldn't you need something like 0.001% resistor accuracy for 16 bits? And 0.000005% for 24 bit accuracy?
     
  20. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,778
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Limitations of FFT measurements / math, especially at lower frequencies. Super large windows needed to get low frequency sines to show up as lines. Also smaller FFT points (smaller windows) will result in broader lines.
     
    Last edited: Mar 30, 2017

Share This Page