DAC-OFF: Soncoz SGD1 vs. Geshelli JNOG

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by purr1n, May 5, 2021.

  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
    I picked these two DACs because they seemed comparable in price and implementation, with the Soncoz SGD1 using one of the latest ESS parts ES038Q2M (x2) and the JNOG using 4493EQ. Both offer balanced outputs and designed with measurements in mind.

    JNOG v2 impressions here: Geshelli JNOG V2 Impressions | Super Best Audio Friends

    @Raimei Templar was written a JNOG vs Modius comparison here: https://www.superbestaudiofriends.o...eshelli-labs-jnog-dac-akm-4493-version.10506/ I thought @Vtory did do, but I can't find it. These impressions should provide you guys with a frame of reference if you have either JNOG or Modius and allow you to triangulate the sound of the SGD1.

    @Bill-P was reviewed the Soncoz SGD1 here: https://www.superbestaudiofriends.org/index.php?threads/soncoz-sgd1.10907/

    Anyway, I more or less agree with their comparative assessments. However, my view of the Soncoz SGD is a bit more negative than @Bill-P's. Maybe Bill was pulling his punches?

    First of all, the Soncoz DAC is a usability disaster. The JNOG by contrast was x500 simpler to use. Plug in stuff into JNOG. Press power button. Press input button to cycle inputs until I hear sound. The Soncoz SGD1 I found extremely frustrating to the point where I wanted to throw it out onto the street and run it over with my Jeep. There's some part of me that still wants to do that and film it to put in on YouTube. Words cannot express the level of frustration I had with this Soncoz DAC. Basically, there are no buttons on the front. There is a large knob that can be rotated and pushed in like a button - and that's it - well except for a remote control which is just as confusing. All functions arise from this knob, input selection, filter selection, volume, mute. Who knows.

    PXL_20210505_194038187.jpg

    To add insult to injury, Soncoz provides a manual in the worst kind of Chi-Fi style where the print is illegible unless you are 16-years old with better than normal vision. Unfortunately, I'm 50 now and my eyes don't focus close anymore. As if Soncoz couldn't say f**k you enough, they plaster 25% of the paper manual with f'ing measurements. Yes, they waste space on their manual to post measurements, instead of more effectively using the space as an easy start up guide to help their users. I blame Amir and ASR for this shit. And BTW, "This design is unique for a desktop DAC" : ASR

    PXL_20210505_190953112.jpg

    And yes, it's easier to read the lettering on my knife than the lettering in the manual. This isn't the first time I'm run into this situation. The other time was with the Topping D50 DAC. At least I know when it comes to Chi-Fi DACs, next time I should buy one of these: https://www.amazon.com/NEWACALOX-Ma...ds=magnifier+soldering&qid=1620245639&sr=8-50
     
    • Like Like x 9
    • Epic Epic x 3
    • List
    Last edited: Jun 8, 2021
  2. rhythmdevils

    rhythmdevils MOT: rhythmdevils audio

    Pyrate
    Joined:
    Apr 15, 2020
    Likes Received:
    12,237
    Trophy Points:
    113
    Location:
    Bay Area, CA
    Home Page:
    Run over it with your jeep! That’s what they’re made for! (Both products)
     
  3. 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 Soncoz SGD1 has a million filters: APOD, BRIC, FAM, FAL, HYER, SLM, SLL. This makes me want to murder someone. Even moreso because the manual says absolutely nothing about what these filters mean. The two where I think I may have some idea are APOD and BRIC. A designer may argue that providing this capability gives the customer options. Personally, I think it's cowardice. In my experience with hundreds of DACs, most of the filters sound like shit (or do nothing discernable) except maybe a handful. The designer should either pick a filter or two that sounds best (I like what iFi does with their "measure" and "listen" or "bitperfect"). Or if the designer believes in measurements, which seems to be the case here, they should pick the filter that reconstructs the signal the most correctly in the time domain with the least amount of high-end rolloff in the frequency domain. Even better yet, the designer should program their own filter parameters into the ESS part. Based on the datasheet, I know this is possible, along with many other things. This is worst kind of Chi-Fi, so creativity gets thrown out the window.

    The JNOG on the other hand handles things for you when it comes to filters. This makes my life that much easier. The JNOG isn't True Hi-End, but it sounds damn good enough without me having to think about everything. FWIW, while I do have better DACs, I didn't have a problem inserting the JNOG into my main headphone system and listening to it for a week. I didn't even have to force myself to listen to it either during this time. In a nutshell, the JNOG is a good AKM Velvet implementation. More "velvet" than Modius, less "velvet" than RME. Neutralish sounding, just a tiny bit soft/fuzzy (AKM thing), obviously delta-sigma, very good resolution (amazing how modern DACs are in this regard), and quite involving. This took me by surprise, but then again, we already know these good these next gen (and expensive) OPA1656 opamps sound. Very competent, maybe lacking slam, but this would be expecting too much. Also the SE and XLR seem to sound more or less the same. Thank you Geshelli for not making me waste the time in my life dicking around with filters.

    For the Soncoz, I took a more measured approach. I performed level-matched blind listening tests (not double-blind) in assessing the sound of the SGD1. System is this: Freya S, Nexus 4x gain mode, Aegir, and JBLs with big drivers for the lows and mids and horn at the top for lowest distortion. My listening tests where with the SPLs cranked up to 95db at the peaks. XLR outputs were used for both DACs.

    PXL_20210505_191658827.jpg

    Again, I think @Bill-P's impressions sums it up pretty good, but I will summarize with my own words in a nutshell:
    • Just because it's a ES9038 part doesn't mean it's gonna sound good. What I've realized is that ESS DACs need a DAC whisperer (LH, MOTU, etc.)
    • ssssh, ccchhh. WTF. OK, like @Bill-P said, it's not as offensive as the worst ESS Sabre DACs of yesterday. Still it's bad because my speakers are pretty laid back and smoother in the highs compared to headphones. I do crank it up, so any issues with the highs will tend to be exacerbated. Bottom line is that digititus should largely engineered out today in 2021, even with delta-sigma DACs.
    • Lean tonality. The Geshelli JNOG had more body, but not "phat" like Modius SE.
    • Limp-dick bass. The JNOG had more rhythm, more grunt, more explosiveness with bass drum and tom hits. The Soncoz lows were not only soft and muted, but kinda indistinct.
    • Kind of "meh" in the engagement department.
    Measurements soon...
     
    • Like Like x 15
    • Epic Epic x 5
    • List
    Last edited: May 5, 2021
  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
    Soncoz SGD1 0dbFS BAL
    upload_2021-5-5_17-34-32.png

    -30dbFS
    upload_2021-5-5_17-36-28.png

    -60dbFS
    upload_2021-5-5_17-37-10.png

    -90dbFS
    upload_2021-5-5_17-37-43.png

    That spur at 850Hz is troubling.
     
    • Like Like x 1
    • Epic Epic x 1
    • List
    Last edited: May 5, 2021
  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
    Geshelli JNOG 0dbFS BAL (SPDIF in)
    upload_2021-5-5_17-43-49.png

    -30dbFS
    upload_2021-5-5_17-45-53.png

    -60dbFS
    upload_2021-5-5_17-46-39.png

    -90dbFS
    upload_2021-5-5_17-47-15.png
     
  6. Merrick

    Merrick A lidless ear

    Pyrate
    Joined:
    Jan 6, 2016
    Likes Received:
    12,568
    Trophy Points:
    113
    Location:
    Portland, OR
    Your comments on the JNOG make me extremely curious how the J2, which uses an ESS chip, will sound. I’ve found I prefer the sound of tamed ESS chips over semi-unfucked AKM chips.
     
  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
    Totally whispered ESS = Exemplary AKM 4399 (prior gen) > Unfucked ESS > AKM Velvet > Bad kind of Chi-Fi* ESS

    * Not all Chi-Fi is bad. There's some really good stuff out there, just not the SGD1.
     
  8. penguins

    penguins Friend, formerly known as fp627

    Pyrate
    Joined:
    Sep 26, 2018
    Likes Received:
    3,451
    Trophy Points:
    113
    Location:
    SouthernCA
    @purr1n - Very minor point but just curious - did you get any RC noise with the JNOG DAC? I consistently heard a little audible background noise with it when it was stacked with the ERISH - only audible when there was no music. Noise went away when I put them side x side or separated them, but I know a lot of people make stacks with their gear so.

    Otherwise, I agree, based on my brief listen, the JNOG was not "True Hi-End" or anything mind blowing, but was pretty enjoyable. Especially "for an inexpensive DAC". I would have no problem recommending this to people new to the hobby or anyone looking for a relatively inexpensive solution.
     
  9. 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
    Yes, I had to reorient the JNOG and move stuff around to avoid very low level buzzing noise infecting my tube amp. It would be cool to have a gadget that measurements RF emissions.
     
  10. Roget

    Roget Acquaintance

    Joined:
    Nov 17, 2018
    Likes Received:
    46
    Trophy Points:
    18
  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
    Can you be more specific?
     
  12. 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
    OK, I will take a shot. Measurement differences could be attributed to the following:
    1. Different power at output / different environment. This is my first precision instrument measurement since back in December of last year. I have moved to a new place. While I have kept the PC the same, the AverLAB the same, the cables the same, you never know what other factors may come into play. We are talking about stuff -125db -145db down where lots of shit can go wrong. Even if you have an AP555, shit will go wrong. I've played with the AP stuff recently and I absolutely love the features and ease of use (the Aver is downright clunky). However, shit still goes wrong for oddball reasons. As such, I think it's fair that if any of these results show anything below -125db, just call it a day and say they are the same, rather than line them up on an Excel bar graph in an attempt show us what's the bestest. What's the bestest one day may not be the next day when the power line or RF conditions change. What's best in one person's home may be different in another person's home.
    2. The 850Hz spur is suspect with the SGD1 measurements I took. However, I have never ever seen a random spur appear like that before a fundamental tone in any DAC measurement - at least I cannot remember any. So perhaps it's not an artifact and actually there. Regardless, the level is too low for it to be of any realistic concern.
    3. I used the SPDIF input. ASR could have used USB / ASIO. This can sometimes account for different results.
    4. This SGD1 was purchased by @Bill-P. I have no idea what the source of it was. I would argue that the distortion pattern of the Soncoz is similar enough to ASR's results: lots of harmonics at each and every order harmonic at about the same level. The main difference is that my harmonics are higher at -120 to -130db and ASRs are about -135db to 145db. This could be the result of the AP555 being a better analyzer. However I doubt this as with what I've seen the AverLAB is just as good on the analyzer (however, the generator on the AverLAB isn't as good). Note that results with other DACs that we have both measured have been similar. This Soncoz is the only one that seems way off.
    5. I believe ASR received his SGD1 from the manufacturer directly? ASR could have received a special "hot-rodded" unit.
    6. We never know when it comes to stuff made in China and possible parts substitution. The 9038 could have been replaced with a lower line chip with the same pinout! The opamps could have been changed for cheaper ones with fake markings. Resistor values could have been slightly different by accident.
    I'll take another measurement later, maybe with different cables, or using a different power strip.
     
    • Like Like x 4
    • Epic Epic x 4
    • List
    Last edited: May 5, 2021
  13. Bill-P

    Bill-P Level 42 Mad Wizard

    Pyrate
    Joined:
    Sep 25, 2015
    Likes Received:
    4,801
    Trophy Points:
    113
    Location:
    San Jose, CA
    For transparency, I got this one off of Amazon recently. It was brand new in box.
     
  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
    PXL_20210506_001753921.jpg
    PXL_20210506_001816969.jpg
    PXL_20210506_001956105.jpg

    Markings on the ESS chip are obscured. However, pinout is consistent with 9028Q2M / 9038Q2M. Opamps are TI OPA1612 and RT6863.
     
    • Like Like x 3
    • Epic Epic x 1
    • List
    Last edited: May 5, 2021
  15. Raimei Templar

    Raimei Templar Friend

    Pyrate
    Joined:
    Dec 8, 2019
    Likes Received:
    775
    Trophy Points:
    93
    Location:
    Osaka
    Good to know I am not completely insane with my assessments :)

    I am planning on doing a comparison between the JNOG AKM and the J2 ESS whenever I receive it and have time to do it. I am quite curious after reading their blurb on the website as it seems like they are doing some unusual stuff with the 9026 chip.

    I haven't had any issues with EM noise MOST of the time with the Erish/Jnog stack, however I just got a new monitor and after plugging it in I started getting some noise/buzz going on. It seems that the JNOG is the issue, as I was able to resolve the issue by plugging it into a different power outlet than the one my monitor was connected to. I guess it is vulnerable to noise generated by other electronics plugged in the same outlet. The Erish seems to be fine on the same power strip as the monitor and a pc. Maybe JNOG has some issues with noise rejection going on.

    Unfortunately I will be returning the monitor as it sucks (Who takes a ultra awesome Samsung VA panel and cranks the gamma so hard it breaks the contrast ratio to shit seriously!?) so I wont be able to test if the J2 resolved this issue or not.

    Another interesting point with the JNOG is that the power output over balanced is quite low, according to Geshellis measurements its only about 2.5v via XLR (at least the J2 is, I assume original is roughly same) as opposed to the 4.0 Modius and most output. The Soncoz looks to output 4.1v unless that was just Amir messing around with its output as he is know to do. I actually kind of like the lower output of the Geshelli as it makes it easier to use with sensitive headphones (also has the 3 gain settings which is really nice). But if you want MOAR power that could be a concern.
     
  16. Psalmanazar

    Psalmanazar Most improved member; A+

    Pyrate Slaytanic Cliff Clavin
    Joined:
    Sep 27, 2015
    Likes Received:
    5,345
    Trophy Points:
    113
    Totally whispered AD and Cirrus are also really good. You just pay well north over 1k for them new.

    The Dangerous stuff is semi-whispered but the high treble... I couldn't live with. It's inoffensive and fun but... too Burr Brown.
     
  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
    OK, another day. Turns out I may have bumped the volume knob. This time I made sure the volume was at full. With the volume at full, the higher order harmonics are a bit lower than the prior measurement (admittedly slightly borked setting). However the third harmonic looks markedly worse. I don't know how ASR got these measurements. They look sus.

    Soncoz SGD1 0dbFS BAL
    SPDIF in
    upload_2021-5-6_18-20-6.png
     
  18. Roget

    Roget Acquaintance

    Joined:
    Nov 17, 2018
    Likes Received:
    46
    Trophy Points:
    18
    Purrin, You need an AP. AverLAB is not good enough.
     
    • Dislike Dislike x 3
    • Respectfully Disagree Respectfully Disagree x 3
    • Miss Information Miss Information x 1
    • List
  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
    Touche, mothafucka!
    [​IMG]
    sgd1.jpg

    I didn't want to reveal this as I'm still in the process of evaluating various AP solutions. As such, I shall invoke a beg-a-thon (I promise I won't do this with every review like Ken Rockwell or Amir). Although I have a reasonably comfortable living, I still have growing kids and am hardly a 1%er. APs aren't cheap. If you haven't already done so, please considering donating to SBAF because the more measurements from different people, the better: https://www.superbestaudiofriends.org/index.php?threads/donations-to-superbaf.331/

    P.S. This Soncoz SGD1 / ASR deal is looking incredibly sus. Maybe @Bill-P has a bad unit, maybe there was some unauthorized parts substitution, fake parts, or some resistors got scrambled. TBH, I suspect ASR got special hot-rodded units which were not reflective of production models here.
     
    • Like Like x 6
    • Epic Epic x 2
    • List
    Last edited: May 6, 2021
  20. penguins

    penguins Friend, formerly known as fp627

    Pyrate
    Joined:
    Sep 26, 2018
    Likes Received:
    3,451
    Trophy Points:
    113
    Location:
    SouthernCA
    And what is your experience with measurements?

    I don't deny that AP stuff works for most audio gear on the market, but just having the AP 555 kajillion or whatever is not an end all be all.
     

Share This Page