Author Topic: no oversampling vs oversampling (NOS Vs OS)  (Read 1812 times)

wusplay

  • Sr. Member
  • ****
  • Posts: 496
Re: no oversampling vs oversampling (NOS Vs OS)
« Reply #75 on: 13 June, 2022, 21:03:23 pm »
  • [applaud]
  • [smite]
  • The technical and all those graphs are beyond me. Totally catch no ball.
    Recently I like to turn and off the external master clock for DAC to hear the differences. I think clock don't change the frequency, yet the improvement in clearity and imaging is interesting. If only I can understand in layman term what have contributed the differences.

    boxerfan88

    • Hero Member
    • *****
    • Posts: 1,047
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #76 on: 13 June, 2022, 21:17:40 pm »
  • [applaud]
  • [smite]
  • Recently I like to turn and off the external master clock for DAC to hear the differences. I think clock don't change the frequency, yet the improvement in clearity and imaging is interesting. If only I can understand in layman term what have contributed the differences.

    Hopefully this helps you to understand... https://headfonics.com/what-is-jitter-in-audio/

    Super duper expensive master clock is supposed to have super duper low jitter...

    Super duper low jitter...means that the timing of the digital-to-analog voltage conversion happens at an interval that is "almost perfect".

    If still cannot understand, then I'll try to explain in even more layman terms. Imagine a cassette player playing a cassette tape. If the motor is a bit cranky, sometimes go a little faster, sometimes go a little slower, than the ideal 4.75cm/sec. One can imagine what happens to the music pitch...go up...go down...just so very slightly...  The perfect motor will make the cassette tape move at exactly 4.75cm/sec, no more no less, which of course does not exist...

    In the digital world, the clock is like the motor... Bestest clock is the one with the lowest jitter... coz there ain't no perfect clock... even those atomic clock also got jitter (but sibei super duper low lor)...

    Hope I didn't confuse u  ;)
    Usher MD2+X616 | DA&T A38  | Soekris dac1541 | foobar2000 | Elementa BAC 2500 | Samsung PS50B850 | Yammy RXA3030 | Orisun DD12 | BDP-S5100 | Chromecast | RPi4+LibreElec | Zidoo Z9X

    wusplay

    • Sr. Member
    • ****
    • Posts: 496
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #77 on: 13 June, 2022, 21:25:34 pm »
  • [applaud]
  • [smite]
  • Thank you for the explanation.
    With master clock I somehow hear more clearly, so it's the more accurate timing that's helping.

    Audio

    • Administrator
    • Hero Member
    • *****
    • Posts: 714
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #78 on: 16 June, 2022, 10:52:22 am »
  • [applaud]
  • [smite]
  • Thank you for the explanation.
    With master clock I somehow hear more clearly, so it's the more accurate timing that's helping.

    Clock makes the difference but there is no way to show you the differences in graph and charts.  That is why I say you listen to your system, you don't use your eyes to look at graphs and then tell me what sounds better and what sounds worst.

    You can use graphs and charts to guide you but you still need to listen with your ears.

    I recalled when MSB produced their DAC more than a decade ago, there is this femto crystal clock module which is an option.  It costs a cool SGD$10K, bring the total cost of the DAC to $40K+ and my friend who bought it said there is no way anyone could avoid this option without regret.

    (Audio)

    Francishuang

    • Hero Member
    • *****
    • Posts: 528
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #79 on: 16 June, 2022, 13:22:04 pm »
  • [applaud]
  • [smite]
  • Even now with the current msb models, clock upgrade still yields a significant improvement. Actually cost wise still quite sensible compared to an external clock. Ext clock need Power cord and clock cables. But if one likes to micro manage or tweak the sound, the power cord and clock cables can become a tool as well.

    Clock makes the difference but there is no way to show you the differences in graph and charts.  That is why I say you listen to your system, you don't use your eyes to look at graphs and then tell me what sounds better and what sounds worst.

    You can use graphs and charts to guide you but you still need to listen with your ears.

    I recalled when MSB produced their DAC more than a decade ago, there is this femto crystal clock module which is an option.  It costs a cool SGD$10K, bring the total cost of the DAC to $40K+ and my friend who bought it said there is no way anyone could avoid this option without regret.

    (Audio)

    AndrewC

    • Hero Member
    • *****
    • Posts: 1,928
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #80 on: 17 June, 2022, 07:32:25 am »
  • [applaud]
  • [smite]
  • Clock makes the difference but there is no way to show you the differences in graph and charts.  That is why I say you listen to your system, you don't use your eyes to look at graphs and then tell me what sounds better and what sounds worst.
    ...
    (Audio)


    Actually, Clocking improvements do show-up in measurement charts!! Albeit usually in very small amounts that border on the threshold of audibility; primarily as lowered jitter, but also as lower distortion, and lower noise floor.

    Have a read of Stereophile’s measurement of dCS’ Verona Clock in 2005; https://www.stereophile.com/content/dcs-verona-master-clock-measurements

    Shows measured improvements, not only for dCS, but even 3rd party elements used during the test;

    Quote

    The absolute jitter level was a moderately low 395 picoseconds peak–peak… For comparison, fig.2 shows the Elgar's output spectrum when it, the Verdi, and the Purcell are all word-clock–slaved to the Verona with the latter's dither switched off. The overall jitter level has dropped to 312ps p–p.


    To my surprise, the ULN-2 behaved a lot better when clocked from the Verona using the latter's dithered S/PDIF word signal (fig.4, blue trace, obtained under the same conditions). Both noise modulation and discrete tones are effectively banished


    That said, a device like a DAC has to be designed to be able to really take advantage of an external clock. Just because it accepts an external clock input doesn’t automatically mean it will perform better with one - the external clock interface/input is itself a source of jitter that could negates any benefit unless designed properly.

    The dCS folks have over 20 years heritage of dealing with Clocking. In the audio market, I can’t think of any other vendor with better knowledge on Clocking technology, implementation, and patents. On the other hand, folks like MSB depend entirely on merchant Clock suppliers, it’s not their own proprietary technology thats driving improvements.
    You'll never get to heaven with a smile on your face from me.

    Audio

    • Administrator
    • Hero Member
    • *****
    • Posts: 714
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #81 on: 17 June, 2022, 10:33:19 am »
  • [applaud]
  • [smite]

  • Actually, Clocking improvements do show-up in measurement charts!! Albeit usually in very small amounts that border on the threshold of audibility; primarily as lowered jitter, but also as lower distortion, and lower noise floor.

    Even if it show in the charts, case by case basis lah, it's still go down to the listening test.  How do you quantify the sonic differences by using your eyes looking at charts? 

    (Audio)

    AndrewC

    • Hero Member
    • *****
    • Posts: 1,928
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #82 on: 17 June, 2022, 11:08:30 am »
  • [applaud]
  • [smite]
  • Even if it show in the charts, case by case basis lah, it's still go down to the listening test.  How do you quantify the sonic differences by using your eyes looking at charts? 

    (Audio)


    Huh?? First you say cannot be measured. Now you’re saying even when it can be measured, it’s no use?? 

    The fact is, it can be measured, and the effect can be heard!

    You’re saying knowing those facts are useless??

    WTF is this forum for then??  ::)
    You'll never get to heaven with a smile on your face from me.

    BadEnglish

    • Hero Member
    • *****
    • Posts: 1,972
    • Bad Since 20th Century. Set to Expire in 21th
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #83 on: 17 June, 2022, 12:37:44 pm »
  • [applaud]
  • [smite]
  • YAHWEH of HiFi is very angry.
    To restore the goodness of HiFi, YAHWEH is going to send the flood.......




    Noah...... where are you !!!!! you owe me one  ;D ;D ;D

    Audio

    • Administrator
    • Hero Member
    • *****
    • Posts: 714
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #84 on: 17 June, 2022, 14:43:00 pm »
  • [applaud]
  • [smite]

  • Huh?? First you say cannot be measured. Now you’re saying even when it can be measured, it’s no use?? 

    The fact is, it can be measured, and the effect can be heard!

    You’re saying knowing those facts are useless??

    WTF is this forum for then??  ::)

    Please don't put your words into my mouth.  Did I say "it no use" or "useless"?
    I say again, and you did quoted me also:-

    Even if it show in the charts, case by case basis lah, it's still go down to the listening test.  How do you quantify the sonic differences by using your eyes looking at charts? 

    (Audio)

    I SAID YOU STILL NEED TO DO THE LISTENING TEST.   Which part of it do you not understand and have to put my statements out of context so that you can rebuke me? 

    (Audio)

     

    boxerfan88

    • Hero Member
    • *****
    • Posts: 1,047
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #85 on: 17 June, 2022, 23:17:08 pm »
  • [applaud]
  • [smite]
  • Listening with my ears and eyes can be quite therapeutic... :D

    Usher MD2+X616 | DA&T A38  | Soekris dac1541 | foobar2000 | Elementa BAC 2500 | Samsung PS50B850 | Yammy RXA3030 | Orisun DD12 | BDP-S5100 | Chromecast | RPi4+LibreElec | Zidoo Z9X

    Audio

    • Administrator
    • Hero Member
    • *****
    • Posts: 714
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #86 on: 17 June, 2022, 23:31:41 pm »
  • [applaud]
  • [smite]
  • Listening with my ears and eyes can be quite therapeutic... :D



    I've been to your place.  I am trying to get the McIntosh Power Meter Display in my system.....ah, listen to Hi Fi with your eyes!   :)

    (Audio)
    « Last Edit: 18 June, 2022, 00:34:39 am by Audio »

    Audio

    • Administrator
    • Hero Member
    • *****
    • Posts: 714
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #87 on: 18 June, 2022, 13:44:00 pm »
  • [applaud]
  • [smite]
  • For the past week or so, I have been trying out other upsampling software.

    Euphony
    I tried this many years ago because Lumin recommended this.  It didn't impress me and now, it still didn't impress me.  Maybe the Windows version is buggy or my hardware didn't fit.  I will try again on a Mac.

    HQPlayer
    I didn't use this because it need a USB DAC.  The interface is poor and no wonder most users used this together with Roon.  I got it to work but it sounded much worse than native Roon or JRiver at 384K.   Must be me, I guess.   JRiver seems to offer more transparency and dynamics than HQPlayer.   I will continue to work on this.

    Any other worthy software I should try?

    (Audio)

    boxerfan88

    • Hero Member
    • *****
    • Posts: 1,047
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #88 on: 18 June, 2022, 20:47:55 pm »
  • [applaud]
  • [smite]
  • I'm now using Secret Rabbit Code...

    That time when you visited, I was using SOX. IIRC u didn't like SOX.
    Usher MD2+X616 | DA&T A38  | Soekris dac1541 | foobar2000 | Elementa BAC 2500 | Samsung PS50B850 | Yammy RXA3030 | Orisun DD12 | BDP-S5100 | Chromecast | RPi4+LibreElec | Zidoo Z9X

    BadEnglish

    • Hero Member
    • *****
    • Posts: 1,972
    • Bad Since 20th Century. Set to Expire in 21th
    Re: no oversampling vs oversampling (NOS Vs OS)
    « Reply #89 on: 24 June, 2022, 09:07:43 am »
  • [applaud]
  • [smite]
  • May suggest a blind test  ;D

    Participants provide the music files to Audio for over-sampling then controlled blind test at where participants have no control over play back, but just to take note which they feel sounded better e.g. playback no.1, no.5, no.6 while the host controls the playing sequences  ;D