USB Nervosa Thread Decrapifiers, pro interfaces, and bears oh my

Discussion in 'Digital: DACs, USB converters, decrapifiers' started by zerodeefex, Sep 28, 2015.

  1. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,469
    Trophy Points:
    113
    Location:
    Winnipeg
    It needs to all be copied into the first post or made a sticky of it's own on this subforum.
     
  2. earnmyturns

    earnmyturns Smartest friend

    Pyrate
    Joined:
    Sep 25, 2016
    Likes Received:
    3,218
    Trophy Points:
    113
    Location:
    Palo Alto
    Home Page:
    Mixing two very different things. 1) The USB cable conductors can convey electrical noise to the DAC. Some form of isolation may or may not help here -- after all, to get the bit waveforms across, noise of the same frequencies will have to be let through. 2) The shape and timing of the bit-encoding waveforms need only to be within some pretty coarse tolerances for the USB receiver to decode them correctly. There's no plausible argument (at least that I know of) that correlates the switching noise from the USB receiver with the shape or timing of within-tolerance incoming USB signals.
     
  3. earnmyturns

    earnmyturns Smartest friend

    Pyrate
    Joined:
    Sep 25, 2016
    Likes Received:
    3,218
    Trophy Points:
    113
    Location:
    Palo Alto
    Home Page:
    On another board I made the horrible mistake of questioning some claims about jitter with the result that I was lectured on the dangers of alleged 400 psec jitter from I2S cables between a SU-1 and a DAC. 400 psec is 2.5GHz in frequency and .14 microns in arrival shift for a sound waveform. I don't have all the latest psychoacoustics at hand, but my recollections from when I used to hang out with experts, and what I know about neuroscience, is that these numbers are orders of magnitude beneath anything conceivably perceivable. Am I full if sh*t?
     
  4. JeffYoung

    JeffYoung Friend

    Pyrate
    Joined:
    Apr 11, 2017
    Likes Received:
    714
    Trophy Points:
    93
    Location:
    Ireland
    I'm in complete agreement.

    I used to think this was how it worked too, but I believe the packets are not actually sent on request. What happens is the packets are sent out on a schedule (which happens to also be at 44.1 KHz, which is perhaps a source of all the misunderstanding).

    The asynchronous part is that the receiver updates the sender with how far ahead/behind it is, and the sender then adjusts the number of packets in the following sends (as Torq alludes to a bit farther down):

    Note that none of this is material to Torq's main point, which I believe is bang-on.

    Cheers,
    Jeff.
     
  5. JeffYoung

    JeffYoung Friend

    Pyrate
    Joined:
    Apr 11, 2017
    Likes Received:
    714
    Trophy Points:
    93
    Location:
    Ireland
    Yes, I imagine isolation from the noise on the power and ground can help (even in the absence of ground loops).

    And I'd also be very skeptical if isolation from the noise on the signal had any effect. I have to believe the physical receiver is using something like a Schmitt trigger to provide hysteresis on the input signal, which will effectively isolate the rest of the circuit from any noise present on it.

    Cheers,
    Jeff.
     
  6. Kattefjaes

    Kattefjaes Mostly Harmless

    Pyrate
    Joined:
    Sep 5, 2016
    Likes Received:
    4,521
    Trophy Points:
    113
    Location:
    London, UK
    Yeah, the biggest available non-snakeoil fix to USB audio is to launder the VBUS line, and provide cleaner replacement power.This can actually help markedly, especially when the USB is coming from a laptop (which tend to be very noisy). Even if the receiver doesn't need to draw power from the sender, they generally need the VBUS to be connected and working, to sense the connection. This is just trying mitigate a glaring* flaw of a dodgy connection choice, though.

    A competent intermediary box can also have benefits when dealing with crappy USB power management issues, too- where stuff goes into a sleep state/powers down when it shouldn't, too. However, this is also about fixing basic brokenness.

    Obviously, this is totally unrelated to any nonsense about "reclocking" though. Also, cleaner USB is still USB- a wooden spoon prize at best. If you're spending more than the price of a Schiit Wyrd or similar, then you should save your money and find a better way to do it.



    * Ho ho, I see what you did there.
     
    Last edited: Jun 2, 2017
  7. Clemmaster

    Clemmaster Friend

    Pyrate Contributor
    Joined:
    Sep 28, 2015
    Likes Received:
    3,268
    Trophy Points:
    113
    We're talking about isolating the DAC from the noisy USB source, inserting an isolation device in between. Of course there's conductors on both sides...

    And yes, there is correlation between signal integrity and noise generated by the USB receiver.
     
  8. Darren G

    Darren G Friend

    Pyrate
    Joined:
    Apr 4, 2017
    Likes Received:
    589
    Trophy Points:
    93
    Location:
    Pennsylvania, PA
    I have to admit I still can't get my head around where the dislike for USB stems from. It's understandably crappy for delivery of a real-time digital data 'stream', with low latency, but for those of us who are listening to pre-recorded music (aka not real time), and for DACs that re-generate a clock anyway, I still don't get the dislike. There is even that useful D+/D- for common noise rejection, it's not like it's an utterly terrible serial bus. Actually given how good at it is at transmitting data like mouse moves, digital video data, file data, etc., without any corruption, is it really that flawed?

    There is power and a ground, and grounds can cause ground loops (and isolators to fix that). I guess the power can be noisy but is the concern that noise there bleeds into the DAC's digital to analog conversion process?

    Open to a better understanding.
     
  9. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    Then read the bleedin' thread.
     
  10. k4rstar

    k4rstar Britney fan club president

    Pyrate BWC
    Joined:
    Jun 11, 2016
    Likes Received:
    6,942
    Trophy Points:
    113
    Location:
    Toronto, Canada
    Stems from the fact that it sounds like shit. Have you taken the time to compare USB to a decent transport? I'm not going to take the time to write a white paper on the technical fallacies of USB audio, you may want to visit Computer Audiophile for that, though you'll probably be even more confused when you leave.

    http://www.superbestaudiofriends.or...-usb-impressions-vs-cdp-vs-built-in-usb.2735/
    https://www.superbestaudiofriends.o...terfaces-aes16-and-e22.2762/page-6#post-77221
     
  11. earnmyturns

    earnmyturns Smartest friend

    Pyrate
    Joined:
    Sep 25, 2016
    Likes Received:
    3,218
    Trophy Points:
    113
    Location:
    Palo Alto
    Home Page:
    Isolation is relative. What frequency band is being blocked? For the digital signal to make it through, the "isolator" needs to let through a pretty wide frequency band. Sure, it will isolate DC and low frequencies (such as AC hum and its low harmonics), but what about the rest?
    Interesting. Can you point to any supporting experimental studies? I'm not an expert in the physical level of digital circuitry by any means, but I've worked in that neighborhood for several decades, and I can't think of any specific mechanism that would correlate the analog properties of the digital-encoding signal coming into an input buffer with the overall switching noise in the complex digital circuit downstream of that buffer. After all, a Schmitt trigger either triggers or it does not, and whatever happens downstream from it seems to be effectively separated from upstream. Of course, a very bad incoming signal could lead to mis-triggering and thus a stream reconstruction that is not bit perfect. But I don't think that's what we are talking about here. Anyway, I'm at (beyond?) my limited knowledge here, just trying to understand the claims that are made for these USB regenerator gadgets, which seem a bit hyperbolic to me.
     
  12. Darren G

    Darren G Friend

    Pyrate
    Joined:
    Apr 4, 2017
    Likes Received:
    589
    Trophy Points:
    93
    Location:
    Pennsylvania, PA
    /shrug

    I am not hearing anything difference between USB and TOSLINK, other than what I did hear before breaking the ground loop. Whether that is my lack of hearing, or that Yggdrasil handles both input sources of data equally well? If there is any diff, I am sure I'd fail a DBT.
     
  13. mrflibble

    mrflibble Friend

    Pyrate
    Joined:
    Oct 4, 2016
    Likes Received:
    358
    Trophy Points:
    63
    Location:
    United Kingdom
    Yggdrasil has Gen 3 USB chip, so that could be why you're not hearing a difference compared to TOSLINK. Gungnir Multibit, on the other hand, has Gen 2 USB, and the difference between that and the TOSLINK input (assuming a high quality source) is easily discernible to me. The Gen 3 USB in on Yggdrasil is reported to be excellent compared to some other USB implementations.

    If you are can't hear a difference, just sit back and enjoy the music :) Maybe pick up a pro audio card and S/PDIF cable (if you are using a PC) if you see the former going for cheap somewhere.
     
  14. Darren G

    Darren G Friend

    Pyrate
    Joined:
    Apr 4, 2017
    Likes Received:
    589
    Trophy Points:
    93
    Location:
    Pennsylvania, PA
    There isn't going to be much, but there can be some. Many years ago I worked on reading CD data for a gaming console, and the sideband, CD+G (graphics); CD data is crap. CD's must read those bits in real time, and read errors happen, but the +G part depended on Reed Solomon Error correction to correct. It is a computationally intensive algorithm, more so when a correction is calculated. Visually you could see those hickups, because there was no getting around the computational cycles to solve for the corrected bits. Any graphics representation that was trying to display that old standard at a smooth 30 or 60fps showed stutters at that time. CD+G is dead.

    What is more important is CD data reads are crap, and it's utterly fraught with jitter to depend on the drive to read at exactly at 44.1K and deliver it the DAC at that perfect rate. Hard drives or SSD drives do this much better.

    People will spend a ton of money to get that right, but then diss USB because that is a serialized packet format, and then also praise network solutions? But those are packet solutions too, not timely or perfect at all. The only difference is it is UDP packets (they are not guaranteed to be timely, or even in order in a complex route), or TCP/IP data, but the packets can arrive out of order too.

    Not saying all gear sounds the same, but it's very hard to cut through what we really hear from what we believe we want to hear.
     
  15. Darren G

    Darren G Friend

    Pyrate
    Joined:
    Apr 4, 2017
    Likes Received:
    589
    Trophy Points:
    93
    Location:
    Pennsylvania, PA
    Could be the Gen 3 vs Gen 2; haven't looked into the diff. It sounds amazing using either input. Both make me go, yeah, this sounds like music!
     
  16. msommers

    msommers High on Epipens

    Pyrate
    Joined:
    Nov 6, 2016
    Likes Received:
    2,750
    Trophy Points:
    113
    Location:
    Calgary, Alberta
    Home Page:
  17. johnjen

    johnjen Doesn’t want to be here but keeps posting anyways

    Pyrate
    Joined:
    Sep 29, 2015
    Likes Received:
    505
    Trophy Points:
    93
    Location:
    Well Grounded
    This may be true for you, but not necessarily for everyone.

    Much depends upon many other factors, not the least of which is the specific system setup and what I call ones degree or type of 'calibration'.
    We each listen for those cues we have been exposed to and can identify as being significant for us and what we deem desirable.
    This is an ongoing process and as we are exposed to 'better', we then can come to know what is possible to achieve.

    Some of these sonic traits we cue in on can be related to improvements we make to the system, while others may not be.
    But being able to distinguish these subtle differences, that are 'new' to our experience, which may not be recognized as such, until a further degree of familiarization/'calibration' occurs, can be all to easily claimed as non existent.

    My 2¢ anyway.

    JJ
     
  18. Kattefjaes

    Kattefjaes Mostly Harmless

    Pyrate
    Joined:
    Sep 5, 2016
    Likes Received:
    4,521
    Trophy Points:
    113
    Location:
    London, UK
  19. Darren G

    Darren G Friend

    Pyrate
    Joined:
    Apr 4, 2017
    Likes Received:
    589
    Trophy Points:
    93
    Location:
    Pennsylvania, PA
    I do absolutely agree not all solid-state/digital gear sounds the same, and a lot of it sounds crappy.

    What is spooky to me, is how good Yggdrasil sounds. What is it doing different that I've never heard before? Speakers/headphones, they have massive coloration and uneven frequency response, but that is understandable. AMPs too, but the differences are smaller (though still can be annoying or enjoyable). DACs? not that I've heard that many but the differences have been minor, but Yggdrasil is a much larger upgrade than I expected.

    Even if USB is junk (or hard to implement), Yggdrasil amazes me over and over. It gets the air and space around music right. I keep hearing details, and tonal accuracy that are lost in a one-note fuzzy presentation that is so typical solid-state/digital gear.

    For whatever it is worth I gave up on CD players except for playing lossless copies (I pay for all my music, despise theft). It's error prone at best, and in the days where RAM is inexpensive, it makes far more sense to me to read ahead, buffer, correct errors, and send the data at the correct rate (assumes a good clock here) vs depending on the motor and all error correction to be completed with perfect timing. Even better is that the DAC make the final clock correction, which removes all the timing slop in the chain before that.
     
  20. Jerry

    Jerry Friend

    Pyrate
    Joined:
    Apr 8, 2017
    Likes Received:
    1,042
    Trophy Points:
    93
    Location:
    Asia
    Hi all,

    I have a question that's been bothering me for a while. Can someone please help enlighten me?

    If my DAC is self-powered and does not require USB Power from the USB cable, do I still need the power line in the USB cable??

    For example, iFi Gemini conveniently separates the data line from the power line. Does it make sense if I only stick the data-line to my PC? Or I still need the power line for handshake or something?

    Thank you.
     

Share This Page