1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Jitter is not audible, sighted and blind tests.

Discussion in 'Sound Science' started by sb, Nov 25, 2009.
First
 
Back
1
3 4
Next
 
Last
  1. Gbjerke
    Well if you have to train your ears to hear jitter, why would you do so? Just to spend money on removing it afterwards?
     
  2. AdamWysokinski
    Quote:

    Originally Posted by Gbjerke /img/forum/go_quote.gif
    Well if you have to train your ears to hear jitter, why would you do so?



    Easy, to let other people sell you their "state-of-the-art" digital cables and other jitter-reducing equipment [​IMG]
     
  3. Dan Lavry Contributor
    Quote:

    Originally Posted by jcx /img/forum/go_quote.gif
    I believe he lost track of the context - 250 nS jitter is only ~1% of the 44.1 sample time - which is where this jitter # applies - not at SPDIF bit edges

    the only limit on digital domain jitter is that all the bits arrive in sequence and can be correctly assembled into the original value for the DAC every ~22uS

    further, jitter spectrum is important to understanding effects - 250 nS rms jitter with a 20KHz bandwidth limit will not corrupt > 1 MHz self clocked serial data




    I thought the conversation was about jitter in a usb and spdif, and given that the spdif format simply can not tolerate 177nsec jitter. So I pointed that fact out.

    I am not sure what those guys were doing. It seems to me that they pre distorted the signal to emulate the effect of jitter. The paper did not have enough detail.

    Well, something is way off here. You can find credible double blind listening tests where people could hear far bellow 1 nsec jitter. There is some good material out there, and I would start by looking at papers by Hawksford and Jullian Dunn. Don't you think that tt is remarkable to have such a huge differences in findings?

    Regards
    Dan Lavry
     
  4. nick_charles Contributor
    Quote:

    Originally Posted by Dan Lavry /img/forum/go_quote.gif

    Well, something is way off here. You can find credible double blind listening tests where people could hear far bellow 1 nsec jitter. There is some good material out there, and I would start by looking at papers by Hawksford and Jullian Dunn. Don't you think that tt is remarkable to have such a huge differences in findings?

    Regards
    Dan Lavry




    Dan, do you have a reference for these sub-ns listening tests, I am genuinely interested, Dunn's jitter work was centered on mathematical models of audibility, Dunn never did any controlled listening tests, I do not think the good Dr from Essex has either afaik, but Benjamin and Gannon who are perhaps more recognized than Ashihara did and they still found thresholds in the 20ns region for signal correlated jitter. Even if we disregard Ashihara and colleagues US empirical research (Dolby Labs) seems to put the thresholds well above 1ns.

    Nobody is disputing the objective damaging effects of jitter , the big question is how bad does it have to be before we can actually hear it.
     
  5. oldson
    "Nobody is disputing the objective damaging effects of jitter , the big question is how bad does it have to be before we can actually hear it. "

    thats what i looked on this thread for. wished i hadn't now. all this is waaay above my head!
     
  6. nick_charles Contributor
    Quote:

    Originally Posted by oldson /img/forum/go_quote.gif
    "Nobody is disputing the objective damaging effects of jitter , the big question is how bad does it have to be before we can actually hear it. "

    thats what i looked on this thread for. wished i hadn't now. all this is waaay above my head!




    LOL - if it is any comfort I have been trying to track down a definitive answer to this question for over 3 years and so far the only controlled listening tests that I have found after rigorous searching *ALL* put the thresholds way above what you would actually get in commercial audio gear, well gear that was not utter pants anyway [​IMG]

    Paradoxically the very worst audio device out there , for measured jitter according to Stereophile, is the not inexpensive McIntosh Music server , well I'll go to the foot of my stairs [​IMG]
     
  7. oldson
    Nick
    thanks for that comment , you just saved me hours of searching.
    i shant dwell on the jitter subject any longer. [​IMG]
     
  8. Dan Lavry Contributor
    Quote:

    Originally Posted by nick_charles /img/forum/go_quote.gif
    ...Nobody is disputing the objective damaging effects of jitter , the big question is how bad does it have to be before we can actually hear it.



    I need to spend less time on the web, but given that you have been looking for a “jitter number” for a long time, let me suggest again to review Hawksford paper, and also Adams paper.

    http://www.essex.ac.uk/csee/research...e%20flawed.pdf

    I do not have a link for Adams paper but I believe you can find it.

    There is much to be read in that articles. Much of what Hawksford and Dunn did and talked about does have to do with keeping the jitter "below the threshold of audibility", not some made up figures but pretty well known levels for what real people can hear. As always, follow the references at the end of the paper…

    And I also highly recommend reading “Clock Jitter, DA Converters, and Sample –Rate Conversion” by Robert Adams published at the Audio Critic (or better yet, for any AES members, look for print # 3712). Bob Adams is an exceptionally brilliant engineer, highly respected by the engineering community!

    If nothing else, you may gain some understanding WHY there is no single number for jitter audibility.

    In Adams paper, he explains why a 1 bit 64fs (such as DSD) is vastly more susceptible to jitter then say a multi-bit sigma delta with a switched cap output stage. You will see that a resistor based R-2R DA has a different sensitivity to jitter… Add to that the facts that the type of jitter plays a role, and the frequency content for each type also plays a role.

    You can not have a meaningful single number for jitter audibility. There are too many variables, from jitter type and content to circuit architecture. In my view, the folks that do come up with papers, listening tests or comments ending with a single universal jitter number, are in fact showing much lack of depth regarding the subject of jitter.

    Regards
    Dan Lavry
     
  9. nick_charles Contributor
    Quote:

    Originally Posted by Dan Lavry /img/forum/go_quote.gif
    I need to spend less time on the web, but given that you have been looking for a “jitter number” for a long time, let me suggest again to review Hawksford paper, and also Adams paper.

    http://www.essex.ac.uk/csee/research...e%20flawed.pdf

    I do not have a link for Adams paper but I believe you can find it.




    I have this paper (or a close variant) but it is worth pointing to anyway.


    Quote:

    There is much to be read in that articles. Much of what Hawksford and Dunn did and talked about does have to do with keeping the jitter "below the threshold of audibility", not some made up figures but pretty well known levels for what real people can hear. As always, follow the references at the end of the paper…



    This is a different Dunn by the way C not J. Also I never referred to Dunn or Hawksford making up figures , I refer them to using models, this is not the same thing, however these figures are still predictions based on models. I do not think they fully take masking into account but even if they do they are yet not supported by listening tests with humans, this is the gap in knowledge. And an opportunity for someone to get a handful of easy AES publications [​IMG]


    Quote:

    And I also highly recommend reading “Clock Jitter, DA Converters, and Sample –Rate Conversion” by Robert Adams published at the Audio Critic (or better yet, for any AES members, look for print # 3712). Bob Adams is an exceptionally brilliant engineer, highly respected by the engineering community!



    Know it, agreed, nice paper and very useful for debagging Stereophile magical thinking journos. But iirc Adams makes the point that separately measuring jitter may be (mostly) unnecessary as it will manifest in downstream distortions anyway ?


    Quote:

    If nothing else, you may gain some understanding WHY there is no single number for jitter audibility.

    In Adams paper, he explains why a 1 bit 64fs (such as DSD) is vastly more susceptible to jitter then say a multi-bit sigma delta with a switched cap output stage. You will see that a resistor based R-2R DA has a different sensitivity to jitter… Add to that the facts that the type of jitter plays a role, and the frequency content for each type also plays a role.

    You can not have a meaningful single number for jitter audibility. There are too many variables, from jitter type and content to circuit architecture. In my view, the folks that do come up with papers, listening tests or comments ending with a single universal jitter number, are in fact showing much lack of depth regarding the subject of jitter.

    Regards
    Dan Lavry



    Well as I say above, this is would look like an excellent opportunity for someone to make a name by carrying out some serious listening tests varying one thing at a time in the established empirical tradition.

    A good starting point though might be blind tests of some add-on jitter reducers.

    To say there are too many variables and leave it at that allows manufacturers of specific jitter devices to escape from the imperative to prove their claims ?
     
  10. SB
    Quote:

    Originally Posted by b0dhi
    Incorrect. The correct conclusion is that it isn't detectable by conscious comparison by untrained listeners. "Audibility" is a seperate issue.

    Also, they had a mouse interface. They don't mention any keyboard control, which is strange considering how important it is to be able to perform the test with eyes closed. They also used a 100ms crossfade when the user wanted to switch samples. Also, the test only addresses frequency or amplitude domain effects of jitter. Time domain effects are not explored with their digital simulation method. The utility of this study is limited.




    Wrong, why not mention Period jitter, Absolute jitter/wander, Peak-to-peak and RMS, Wideband jitter, Baseband jitter, Long-term jitter, ect. You just picked something at random to sound intelligent and have yet to back up anything you have said.

    Quote:

    Originally Posted by ert
    FTA "A total of 23 audio professionals or semi-professionals participated as the listeners. They were audio engineers, audio critics, sound engineers, and musicians." What is your definition of "untrained listener"? Also, I'm not sure that your distinction between "audibility" and "detectable" has any practical implication.

    An interesting study. They seem to discuss some more details of the simulation wrt actual observed hardware jitter in other papers.




    I agree.

    Quote:

    Originally Posted by AdamWysokinski
    Oh, is it? How can you determine sonic qualities of something that is not detectable by the hearing system?



    It seems most people do not listen and make up stories about what is audible and what is not.

    Quote:

    Originally Posted by nick_charles
    There are several different types of jitter, random jitter as in the paper you cite is indeed far less audible, Benjamin and Gannon used correlated jitter in their 1998 study and found the thresholds to be from about 20ns in music but getting as high as 300ns in some cases. Correlated jitter creates distortion sidebands while random jitter raises the noise floor.



    20ns is still 20,000ps which is much higher then the current jitter measurements.

    Quote:

    Originally Posted by leeperry
    well, I said "IMO"...I don't quite have the same kind of equipment the XXhighend coder has to measure jitter...I merely use my two ears to judge on that.



    So you don't know and are just assuming.

    Quote:

    Originally Posted by b0dhi
    If the test depends on conscious comparison and detection, it is a skill. It is therefore required to train the listeners if the result is to be extrapolated as "audibility" and not just conscious comparative detection skill.

    Audio-pros, musicians and audiophiles are not people trained in detecting jitter-caused distortion. Being slightly more likely to be able to do so than the general population due to one's profession/hobby is hardly "trained".




    Trained to listen to jitter, now that is funny.

    Quote:

    Originally Posted by Dan Lavry
    Well, something is way off here. You can find credible double blind listening tests where people could hear far bellow 1 nsec jitter. There is some good material out there, and I would start by looking at papers by Hawksford and Jullian Dunn. Don't you think that tt is remarkable to have such a huge differences in findings?



    Post up the link.
     
  11. DannyBuoy
    Quote:

    Originally Posted by punkaroo /img/forum/go_quote.gif
    Maybe my ears need a cleaning, but I've done a listening test using an iBasso D10, and hooked it up via USB and optical. I could not tell a difference in SQ using either or. I was shocked to say the least. I was really expecting optical to be on top, but nope!



    Not sure if I need to spend extra cash on D10 as it looks like multi-interface is the strength of D10 but I notice the chips used are different. Is the D10 sonically superior to iBasso D4 Mamba in your opinion?
     
  12. nick_charles Contributor
    Quote:

    Originally Posted by SB /img/forum/go_quote.gif
    20ns is still 20,000ps which is much higher then the current jitter measurements.



    There is one device that gets to over 11ns. I guess we cannot say jitter is definitively always inaudible without more testing but all the controlled listening tests with whatever methodological issues they may have to date do point in the same direction... i.e the don't worry about it direction.
     
  13. SB
    Quote:

    Originally Posted by nick_charles /img/forum/go_quote.gif
    There is one device that gets to over 11ns. I guess we cannot say jitter is definitively always inaudible without more testing but all the controlled listening tests with whatever methodological issues they may have to date do point in the same direction... i.e the don't worry about it direction.



    What component is that?
     
  14. nick_charles Contributor
    Quote:

    Originally Posted by SB /img/forum/go_quote.gif
    What component is that?



    The McIntosh Music Server.
     
  15. SB
    Quote:

    Originally Posted by nick_charles /img/forum/go_quote.gif
    The McIntosh Music Server.



    Really? Hahaha so much for the high end being better designed and performing better, mcintosh is such a joke.
     
First
 
Back
1
3 4
Next
 
Last

Share This Page