Denafrips Ares 2 R2R DAC Measurements and 'NOS' Investigation

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by GoldenOne, Jul 12, 2021.

  1. Baten

    Baten Friend

    Pyrate
    Joined:
    Mar 18, 2018
    Likes Received:
    1,130
    Trophy Points:
    93
    Location:
    EU
    Very interesting!! Question, I believe iFi in 'BitPerfect' does the same thing. Any chance you have one of those lying around to verify that? Would be very cool. I really like iFi's 'NOS'/BitPerfect sound. If Denafrips/Ares does the same thing that'd be good for me to know.
     
  2. Marvey

    Marvey Super Friend

    Staff Member Pyrate BWC
    Joined:
    Jun 6, 2021
    Likes Received:
    1,712
    Trophy Points:
    93
    I would tend to agree that bandwidth for distortion measurements should be extended past 20kHz. Up to 88-100kHz seems reasonable without going crazy. I know it's easy to say that past 20kHz, we technically can't hear so no big deal. Well, we technically can't hear -100db SINAD either. Speaking to many audio designers from the big names (I did live in the audio mecca - all of these guys are retired or dead), stuff up to the 50-60kHz range was important to them. Ultrasonic oddities, distortion, frequency response, up this high - it still mattered. They believed it affected the sound that we heard. Not as outright measurement phenomena from the steady sine wave, but more as indicators of what happens in more complex signals or music.

    Now as far as up to the MHz and GHz regions, that's probably needed for FCC Class B.
     
  3. NekoAudio

    NekoAudio Acquaintance

    Joined:
    Apr 18, 2016
    Likes Received:
    50
    Trophy Points:
    18
    Location:
    San Jose, California
    Home Page:
    When I spoke with Dan Lavry many years ago, he mentioned making sure that ultrasonics did not exist on the analog output of his DACs because that "noise" could impact the performance of the downstream circuitry and chips. I don't recall him mentioning anything about those ultrasonic frequencies themselves being perceived by the listener. I can understand the EE reasoning behind concerns that these frequencies might end up affecting the audible spectrum below 20kHz if they get through to or are picked up by other circuitry. Although I can't point to any specific evidence to support it.
     
  4. GoldenOne

    GoldenOne Friend

    Pyrate
    Joined:
    Jul 17, 2020
    Likes Received:
    1,049
    Trophy Points:
    93
    Location:
    England
    As an update, I'm currently talking with someone else who has an APx555 and Denafrips terminator plus with them currently.

    I've sent them my project sequence and they're running it through in NOS and in OS.

    Similar, but different behaviour is going on with the termi plus. This is the "NOS" impulse:
    506ECDF8C4D847F3A6DDAE4A531FAB40.png

    And as with ares 2, getting the IR from the transfer function measurement instead of an actual IR test shows something resembling normal sinc.

    Definitely not NOS either, but it seems they're running the DSP differently to the ares 2.

    Full measurements coming soon
     
  5. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    So what's going on with the 'slow' filter impulse response? It really shouldn't be ratcheting from zero to full output like that in the middle of the pulse, should it?
     
  6. GoldenOne

    GoldenOne Friend

    Pyrate
    Joined:
    Jul 17, 2020
    Likes Received:
    1,049
    Trophy Points:
    93
    Location:
    England
    It definitely shouldn't no.

    Unfortunately the unit has been sent back to the owner now so I can't test further. Probably should have tested a -3dBfs impulse or something to see if it was 'wrapping around' or something
     
  7. Beefy

    Beefy Friend

    Pyrate
    Joined:
    Apr 10, 2021
    Likes Received:
    1,738
    Trophy Points:
    93
    Location:
    Canada
    Good, nobody mentioned it, so I'm glad I'm not going crazy. Reminds me of what the O2 DAC was doing here: https://superbestaudiofriends.org/i...-how-nwavguy-fooled-us-all.10173/#post-326363
     
  8. 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:
    Going off the Metrum/Sonnet DACs, where each module generation sounds different than the last and shown that this glue point hasn't remained static, I'd slap a big, fat "It Depends" sticker on the topic.

    Having this point as high up as -40dBFS, though, that seems peculiar to me. Would be interesting to have the ability to adjust this point on the fly and listen for audible changes.
     
  9. Marvey

    Marvey Super Friend

    Staff Member Pyrate BWC
    Joined:
    Jun 6, 2021
    Likes Received:
    1,712
    Trophy Points:
    93
    Well, I didn't care for the Sonnet and pulled my punches big time for that one. Would be curious as the methods for Metrum, of which all their entry level and first gen units I didn't like either.

    Would also be interesting to see data for the DACs and line them up with what I thought was poop and good. I bet some ways of gluing are better than others.

    And really, I'm not a discrete resistor ladder guy either, the sole exception being Soekris. The DAC2541 remains the only discrete ladder DAC which I loved (enough to want to own one - that is sidegrade from my other DACs). The RDAC with tweaked power supply the other.

    Seriously, I would rather take ancient PCM63 over thus glue crap, because I can be assured of non-shittiness.
     
    Last edited by a moderator: Jul 14, 2021
  10. Marvey

    Marvey Super Friend

    Staff Member Pyrate BWC
    Joined:
    Jun 6, 2021
    Likes Received:
    1,712
    Trophy Points:
    93
    upload_2021-7-14_11-34-25.png

    15kHz sine? Bwhahahahahahahahahaha!

    I just noticed this. WTF. Am I really reading this correctly? Did Denafrips seriously play connect the dots instead of coding in a proper sinc interpolation formula?

    This is exactly why DACs should not be designed by people who believe in 15x15 arrays of small caps for PS filtering.

    LOL, I guess this is "NOS". The purest form of NOS where aliasing defects are forced upon the listener! (Normally without a brickwall analog filter, we can count on bandwidth limited components such as the amp or transducers to anti-alias for us. However, we don't have a choice here. :p

    This is horrid. This is utter garbage. This is a fricking nightmare.
     
    • Like Like x 3
    • Epic Epic x 2
    • List
    Last edited by a moderator: Jul 14, 2021
  11. GoldenOne

    GoldenOne Friend

    Pyrate
    Joined:
    Jul 17, 2020
    Likes Received:
    1,049
    Trophy Points:
    93
    Location:
    England
    Unfortunately terminator plus is doing the same thing :(

    upload_2021-7-14_17-41-49.png

    Definitely an.....odd choice.
     
  12. Marvey

    Marvey Super Friend

    Staff Member Pyrate BWC
    Joined:
    Jun 6, 2021
    Likes Received:
    1,712
    Trophy Points:
    93
    Relevant discussion for non-math people with homework exercises (actually completed by a few SBAF members):
    https://www.superbestaudiofriends.o...an-inconvenient-truth.9480/page-7#post-307070

    TL;DR: One cannot connect simply the dots from a sampled sound wave, especially as get near Nyquist. If we do, we get aliasing. That is we could end up with a different frequency at a different amplitude than the original. Think of it as moire on badly downscaled images of stuff with fine patterns.

    16kHz at 48kHz sampling rate
    upload_2021-7-14_14-41-20.png
     
    Last edited by a moderator: Jul 14, 2021
  13. Walderstorn

    Walderstorn Friend

    Pyrate
    Joined:
    Mar 20, 2016
    Likes Received:
    1,904
    Trophy Points:
    113
    Location:
    Europe
    If @GoldenOne wants I can send him my metrum Jade, if it helps with anything...
     
  14. dasman66

    dasman66 Self proclaimed lazy ass - friend

    Pyrate Contributor
    Joined:
    Mar 13, 2018
    Likes Received:
    2,462
    Trophy Points:
    113
    Location:
    NW Pennsylvania
    I can send an Amethyst to someone in the US if there is interest (at least, I can as soon as I get the ECP Walnut that @TomB is building for me)
     
  15. GoldenOne

    GoldenOne Friend

    Pyrate
    Joined:
    Jul 17, 2020
    Likes Received:
    1,049
    Trophy Points:
    93
    Location:
    England
    Would definitely love to test it and give it a listen!
    Are you in the UK?
     
  16. toddrhodes

    toddrhodes Friend

    Pyrate
    Joined:
    Sep 13, 2017
    Likes Received:
    585
    Trophy Points:
    93
    Location:
    South Bend, IN
    I will say that I really enjoy my Venus II DAC. I pair it with an Ian Canada reclocker and have thoroughly enjoyed its sound. However I do have an odd "click" every now and again, over i2S. I've tried multiple sources for i2S including the DF Hermes.

    And given how some of these things are coming to light with NOS not being NOS (and I'm an HQPlayer upsampling user), and now this connect the dots stuff which I don't fully understand, admittedly, but it just feels off.

    I've put in for a Holo May KTE, because as I understand it, despite being $2k more than Venus II - it does have the detail and refinement and layering and impact that I want out of a DAC, and it does true NOS so I can really stretch its legs with Roon and HQPlayer.

    So yea.

    Funny enough, I just posted a postively glowing review of Venus over on Hoffman forums today. And I meant and still mean every word of it. But I'm ready to move up to May.
     
  17. Jerry

    Jerry Friend

    Pyrate
    Joined:
    Apr 8, 2017
    Likes Received:
    1,042
    Trophy Points:
    93
    Location:
    Asia
    What would this do to actual sound quality? Would this mess up the whole frequency spectrum? Some frequency appearimg louder than it should?
     
    Last edited: Jul 15, 2021
  18. ushanka

    ushanka Facebook Friend

    Contributor
    Joined:
    Nov 30, 2020
    Likes Received:
    121
    Trophy Points:
    43
    Location:
    PNW
    In the particular example you are basically getting a bunch of harmonic distortion, and not the pleasant even order kind. Resolution loss in treble, which would match @GoldenOne’s impressions.
     
  19. Jerry

    Jerry Friend

    Pyrate
    Joined:
    Apr 8, 2017
    Likes Received:
    1,042
    Trophy Points:
    93
    Location:
    Asia
    GoldenOne still gave a positive review...sonic wise. And more all-rounder than Bifrost 2 too.

    Being a total noob in measurements, very interesting to understand more the correlation between the two. On paper, this shouldnt sound good.
     
    Last edited: Aug 12, 2021
  20. GoldenOne

    GoldenOne Friend

    Pyrate
    Joined:
    Jul 17, 2020
    Likes Received:
    1,049
    Trophy Points:
    93
    Location:
    England
    To be clear though my Impressions were formed almost entirely in the sharp os mode.

    I didn't like the nos mode at all

    In terms of measurements, it still does pretty well. And in fact measures better than many other r2r options.
     

Share This Page