Topping D10s Review and Measurements

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by purr1n, Dec 25, 2022.

  1. Hrodulf

    Hrodulf Prohibited from acting as an MOT until year 2050

    Pyrate
    Joined:
    Sep 28, 2015
    Likes Received:
    4,309
    Trophy Points:
    113
    Great review! I've steered clear of ESS DACs since I heard the Twisted Pear Buffalo 2 a friend of mine built. For me it was nigh unlistenable, yet unlike anything I've heard before technicalities wise. Even after heroic efforts from the said friend the 9018 chips proved very hard to tame. I think most of DIYA forum ended sticking huge Nelson Pass designed output stages and had their DACs work in 3U heatsink cases to properly cool them. Mad kudos to Schiit for what they've done.

    The Qudelix 5K I'm using right now has two ES9218P all in ones and they're pretty okay for what the device is trying to do. Can't really detect any of the ESS treble nasties.
     
  2. Tone?

    Tone? Acquaintance

    Joined:
    Dec 23, 2022
    Likes Received:
    29
    Trophy Points:
    18
    Location:
    USA

    The new Cirrus CS43198 isn’t better than the CS4398? ( my Marantz uses the CS4398 and I love it more than any DAC I’ve tried.)

    the Topping D30pro DAC I tried a couple weeks ago had a very similar tonality to my Marantz, but because of the cheap analog output stage and other parts the topping was flatter and didn’t have much weight at all. As all cheap Shenzen audio products have.
     
  3. Tone?

    Tone? Acquaintance

    Joined:
    Dec 23, 2022
    Likes Received:
    29
    Trophy Points:
    18
    Location:
    USA
    By far the worst DAC I’ve heard was the topping D90SE

    Horrible. And I had no idea of Amir and his cult back when I tried that DAC.

    was unlistenable.
     
  4. Tone?

    Tone? Acquaintance

    Joined:
    Dec 23, 2022
    Likes Received:
    29
    Trophy Points:
    18
    Location:
    USA

    I think you might like the Topping D30pro which uses 4 Cirrus chips.

    if I knew more about electronics I would mod that DAC to sound great.
    IMHO the Cirrus chipsets have the most natural timbre by far.
     
  5. goodvibes

    goodvibes Facebook Friend

    Joined:
    Jun 21, 2017
    Likes Received:
    35
    Trophy Points:
    30
    Location:
    Chicago
    Questyle is Chinese but competes with the west. My fav of the AKM was the 4490 pre 'R'. I still like what what some have accomplished with the PCM 1792 and Cyrus chips as well.
     
    Last edited: Dec 26, 2022
  6. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,634
    Trophy Points:
    113
    Location:
    Winnipeg
    I thought the Prism/Tracktion thing was just a joint effort thing mostly on the software side?

    I still have my Prism Lyra and it was my favourite until the SPL Mercury came around which is an AKM chip. My other dac with the CS4398 is the Bryston B60 which is the same but simplified architecture from the BDA-1. I actually compared the two at one point and funny enough preferred the B60 built-in over the dedicated dac.
     
  7. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    19,082
    Trophy Points:
    113
    Location:
    On planet
    Why do you assume the results on ASR are correct and @purr1n needs to retest?

    Let me elucidate.

    Look at that graph above. Especially the y-axis which is labelled Level (dBrA). That means the y-axis is a relative measurement independent of signal generator stimulus level. Keep that in mind while viewing the results I obtained with a Modi+ on my dScope independently.

    Imaging Modi+ USB input SE output 48KHz sample rate with stimulus set to -6 dBFS
    20221228 Modi+ Imaging 48KHz -6dBFS USB SE annotated.png
    This result is substantially similar to that obtained by @purr1n

    Imaging Modi+ USB input SE output 48KHz sample rate with stimulus set to -46 dBFS
    20221228 Modi+ Imaging 48KHz -46dBFS USB SE annotated.png
    This result is substantially similar to that shown on ASR.

    When relative level is used, the signal generator can be set to any level with results normalized to 0 dBr.
    You won't know what level was used unless a setup parameter full disclosure is provided.

    Here is the animation of the two results:
    20221228 Modi+ Imaging 48KHz -6  -46dBFS USB SE.gif

    Distortion sweep 1 KHz -1dBFS Modi+ USB input SE output 80 KHz BW
    20221228 Modi+ Distortion vs Freq 48KHz 1dBFS USB SE  80KHz BW.png
    This result is substantially similar to @purr1n result.

    I'm going to suggest something was amiss with the other website's results, not accurately reflecting objectively the true performance of Modi+.
     
    • Epic Epic x 21
    • Like Like x 15
    • List
  8. GoldenOne

    GoldenOne Friend

    Pyrate
    Joined:
    Jul 17, 2020
    Likes Received:
    1,051
    Trophy Points:
    93
    Location:
    England
    +1

    I've got a Modi+ here and my results align with @purr1n 's

    Seems something was up with ASR's test.

    @Roget whilst not ALWAYS the case, usually in the instance that two results differ, it's the one that shows worse performance that needs to be revisited.
    It's quite hard to screw something up to make it look better than it is. It's very easy to screw something up and make it look worse than it is.

    upload_2022-12-29_15-24-34.png

    And as @atomicbob mentioned, with stuff like this you can't know what the actual values being sent to the DAC are or what the Y axis is normalised to unless it is disclosed.

    Amir never discloses much about his test config and doesn't provide a report so we can't verify.

    I've attached a full report with all info here: https://drive.google.com/file/d/1MN5lB2I_5oaz7V3RSGEAdZChKw77vk__/view?usp=share_link

    (I feed -3dBfs white noise to the DAC rather than 0dBfs just to ensure there's no intersample overs clipping as some DACs are susceptible to that. Y axis is normalised to that level. 1.248 million point FFT, 50 averages)
     
    • Like Like x 26
    • Epic Epic x 2
    • Agreed, ditto, +1 Agreed, ditto, +1 x 1
    • List
    Last edited: Dec 29, 2022
  9. Maximillion

    Maximillion Acquaintance

    Joined:
    Dec 15, 2018
    Likes Received:
    58
    Trophy Points:
    18
    Location:
    Warwickshire
    The above discrepancies are why , imo, measurement is fast becoming the new snake oil, which is ironic.

    I still don’t understand why any measurement of something we can’t actually hear is in any way relevant (when it comes to consumer hifi) - maybe I am wrong but surely measurement should be limited to a simple pass / fail. i.e no defects that make it sound crap in the frequency range you can actually hear = pass. Otherwise, fail.

    Then it’s all about HOW it sounds, not whether it’s broken.
     
    • Like Like x 3
    • Agreed, ditto, +1 Agreed, ditto, +1 x 1
    • List
  10. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    90,317
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    This is exactly the Changstar and SBAF approach - and why we just didn't bother measuring most DACs or amps until recently. Most of the DAC and amp measurements had zero correlation to what we were hearing. SINAD / THD+N really only tells us if something is broken, and even then, as said earlier, it needs to be really poor and according to the realistic use case.

    SINAD (or AmirNAD), especially way past audibility (that is 55db) becomes relevant only because of religious belief. Or as many have surmised, because it's being used as a marketing tool: https://www.superbestaudiofriends.o...g-and-smsl-with-asr-punking-us-on-dacs.12808/
     
  11. Entropy

    Entropy Facebook Friend

    Joined:
    Dec 3, 2022
    Likes Received:
    204
    Trophy Points:
    33
    Location:
    Portland OR
    ... Yikes. I didn't know that this issue ran so deep. Fighting disinformation with disinformation, eh? This Really seems to highlight the perils of trying to interpret measurements without fully understanding their meaning and the context behind them, setting aside the fact that prominent subjective commentary will likely also sway opinions... makes me want to know exactly what's going on in ASR, what Amir's game plan truly is. But in the meantime, please excuse me to re-watch GS's video on measurements again.
    Thanks for the insight.
     
  12. Psalmanazar

    Psalmanazar Most improved member; A+

    Pyrate Slaytanic Cliff Clavin
    Joined:
    Sep 27, 2015
    Likes Received:
    5,359
    Trophy Points:
    113
    Yeah everything decent (not something like Presonus or Focusrite) since multibit delta-sigma chips were invented has had low enough thd+n. NObody realyl wants to measure components or phase or imd. Then the loser gear is revealed and yes it can be high priced and worse than the mid gear.

    Let's consider what I would consider to be still produced modern mastering hardware. Here are some noisefloors:
    Daking -80ish dbu pretty clean
    Manley -85ish dbu cleanish to lofi goop soup
    Dangerous -90ish dbu a little color
    Maselec -100ish dbu super duper clean
    API something stupid like -120 dbu but the sound might as well be bandpassed

    I don't see the noise floors having anything to do with the perceived sound of the CD or 16-bit file you get after it's reconverted back to digital and run through a limiter that can be clean to dull to WTF distorted and has even more noise stuck on it from the necessary dither. 24-bit as a distribution format is WTF unneeded.
     

Share This Page