clock jitter

Discussion in 'Converters / Interfaces' started by audiokid, Apr 18, 2010.

  • AT5047

    The New AT5047 Premier Studio Microphone Purity Transformed

  1. Ryan Edward

    Ryan Edward Active Member

    Joined:
    Aug 9, 2010
    I see ads for very expensive power supply cleaners that ensure only clean regulated power goes to your equipment. Will this reduce jitter?
     
  2. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    There is no fixed answer to this as it depends totally on your particular soundcard. As I have said already on this thread the biggest problems from clock jitter arise when you are using multiple synchronised soundcards or an external clock master. Every soundcard will have a minimum jitter performance that is defined by its design and is presumably what you will see in the spec. sheet. Noisy power can only add to this minimum and whether it will be a significant degradation depends on the ability of the power supply of the particular soundcard to reject that noise. This is very variable.

    That said, clean power in a studio is always a good idea, not just from the point of view of clock jitter but also just about all the other sensitive gear around. Field recording is just the usual headache of what gear to take!
     
  3. Ryan Edward

    Ryan Edward Active Member

    Joined:
    Aug 9, 2010
    Thanks for the clarification.
     
  4. jasonthomas

    jasonthomas Guest

    Having read the explanation of Clock Jitter, I honestly can not say I have ever noticed it. At least not whilst watching Blu-rays. I don't recall hearing it in a studio, though I'm probably wrong there. As I am in no way an expert, unlike most of you guys and gals.
     
  5. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    I would think there are very very few people that have heard what can be directly attributable to clock jitter. I also think that there is an element of "panic" in the audio world as another source of noise/distortion is recognised and accepted (although engineers in other disciplines have been battling with jitter for many decades!). There we go, now I'm expressing an opinion. That's almost bound to get a reaction so I'd better try and justify my opinion!

    Unless a locked oscillator has some very specific audio frequency sidebands which would create audio noise at that specific frequency, then jitter noise should approximate to pink noise. Although the worst case figures I gave would (should) be quite noticeable, in practice the "real" levels of the jitter noise will be very much lower and therefore either amid the normal system noise floor or at least well below a strong signal.

    The usual manifestation of truly random excessive jitter noise should be a signal dependent elevated noise floor. This is why very few will be able to positively determine that an elevated noise floor is caused by jitter as the noise floor will also reduce with the signal level. This also means that jitter noise will generally be unobtrusive - unless of course your clock has some really nasty sidebands.

    In hindsight, I should perhaps have mentioned this much earlier. Thanks for prodding!
     
  6. Ryan Edward

    Ryan Edward Active Member

    Joined:
    Aug 9, 2010
    So it's not some the average Joe would spot?
     
  7. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    Well, in my opinion no! First, most people will be just using a single soundcard, at least in the domestic environment. As this does not involve locked oscillators and uses a simple crystal oscillator, it is unlikely to have a large jitter. This is just the nature of crystal oscillators and it would take some very poor design for such an oscillators to have jitter in the order of the 25 ns I measured with locked oscillators on my own system.

    Secondly, the figures I gave using a 20 kHz sine give a maximum error which is - 56 dBc, where dBc is the measurement of the noise relative to the signal. For a 1 kHz sine, this reduces to -82 dBc. Compare these figures to say a vinyl disk. Due to the required RIAA compensation, this is just about the maximum theoretical S-N available for a disc input and that noise is constant and independent of programme material unlike jitter noise. I give these figures to try and put some meaning to the numbers and it is not within the capabilities of many systems (or ears) to truly reveal jitter noise.

    In an industry where we can measure many many things, we constantly get theories about why A sounds different to B which usually involve some speculative mechanism which defeats the ability of objective science to measure (as IMHO the mechanism probably didn't exist in the first place). This really isn't one of them. We know about clock jitter and can measure it quite readily. Having measured it we can easily enumerate what the effects will be.

    While it is good engineering practice to minimise any clock jitter and with the current apprehension about it, then any modern converters should really have battled the effects down to insignificance. I believe they have achieved this and it is not something that concerns me at least, even with my ageing equipment with (as Boswell said) HUGE 25 ns jitter. I have a pretty decent monitoring set up, and I certainly cannot claim I have ever heard anything I could ascribe to clock jitter. Most of my time mixing is concerned with much bigger problems than clock jitter! Personally, I never give it a second thought other than to try and answer audiokid's original question.
     
  8. Ryan Edward

    Ryan Edward Active Member

    Joined:
    Aug 9, 2010
    Thanks for the clarification.
     
  9. Ryan Edward

    Ryan Edward Active Member

    Joined:
    Aug 9, 2010
    I read reviews of very hi end Blu-ray player boasting of the fact of almost zero clock jitter and charging a premium for it. Sounds like these manufacturers are trying to make money on some thing that is negligible anyway.
     
  10. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    Well, marketing departments are well aware of the debate and controversy surrounding clock jitter so I guess this is to be expected!
     
  11. jasonthomas

    jasonthomas Guest

    One company were promoting their ultra hi end Blu-ray player costing over $100K and all they were going on about was ultra hi quality clocks to prevent jitter. Seems like marketing BS to me. My $500 PS3 plays Blu-rays beautifully and I observe no jitter.
     
  12. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    Just to pick up on that, it is IMPOSSIBLE to prevent clock jitter. All electronic circuits have noise so the jitter can only be reduced - and I also indicated what levels of reduction might be achieved. I have already given figures on what reducing jitter will achieve in terms of audio performance.

    Talking about different Blu Ray player claims is not the main thrust of what I have been trying to explain and I think I have already well covered this in responding to other previous comments just a few posts back. Now we're just going round in circles! Please bear in mind that there is a big difference when using a single clock and not slaving soundcards. Unless you slave an external D-A to a Blu Ray player then most of the problems with clock jitter are avoided anyway.
     
  13. Ryan Edward

    Ryan Edward Active Member

    Joined:
    Aug 9, 2010
    All this talk of jitter is making my head spin.
     
  14. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    Well, after Jasonthomas paraphrased your earlier post, I did say we were going round in circles! Maybe you just got dizzy! :<)
     
  15. jasonthomas

    jasonthomas Guest

    My apologies. Didn't mean to derail the thread.
     
  16. Boswell

    Boswell Moderator Distinguished Member

    Joined:
    Apr 19, 2006
    Location:
    UK
    Home Page:
    Here's a link to an on-line seminar entitled "Anatomy of Jitter", presented by Tektronix and due to go out on 11 Nov 2010. I suspect it may deal more with data transmission than with sampling of audio, but there is a lot of common ground.
     
  17. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    @jasonthomas. No apology needed and sorry if I seemed a bit grouchy! This is not even my thread so it could easily be me that has derailed it!

    Jitter in any digital music system is relevant especially if it is not well controlled and I was not trying to dismiss your point even though it had been covered a few posts earlier. In fact it gave me the chance to spout off a bit more... :<)



    As you say, I'm sure this will be mainly aimed at data transmission (in view of Tek's recent scope releases!) but the brief synopsis does at least promise a bit more on general jitter examination.

    Just a word of warning for others though, this presentation is going to be aimed squarely at engineers (of various abilities!) and not the lay person. As such it is highly likely to entail some slightly more advanced maths (mainly statistics). You may end up even more confused than before if you're not careful!

    What I hope you should learn is that my presentations of peak clock jitter are somewhat simplified (but quite valid) from the mathematical analyses of the various aspects of clock jitter.
     
  18. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    I took the time out to look at the Tektronix seminar (it is still available on the link and will normally stay around for a while).

    Sadly, as Boswell and I suspected, it is solely focussed on data transmission. It does however give the "standard" analysis of the various factors contributing to jitter. This is of little interest to the general forum when I have already given some measurements of a typical audio system and also indicated where different causes of jitter could be problematic.

    If you want to know more of the maths and statistics involved, by all means take a look but I don't expect you will find anything that really adds to what we have already discussed here.
     
  19. djmukilteo

    djmukilteo Well-Known Member

    Joined:
    Nov 23, 2008
    Location:
    Rainy Roads WA USA


    "Solely focused on data transmission" is the topic here
    I don't see any reason why the Tektronix white paper isn't completely relevant as that's what your talking about.
    After all the only thing the clock is syncing is data transmissions...nothing else!
     
  20. MrEase

    MrEase Active Member

    Joined:
    Jan 19, 2006
    Location:
    West Suss ex, UK
    Well there are certainly different aspects to what I have discussed but I do not regard any aspect of this as (digital) data transmission rather than clock transmission. I may have discussed recovering a clock from S/PDIF or ADAT but I have not concerned myself with the associated data integrity, only the clock recovery and mainly the relative jitter between the recovered clock and the transmitted clock. In terms of data integrity of S/PDIF and ADAT transmissions then of course the seminar is relevant, however, with the data rates involved with these formats we have relatively huge margins for noise and in the eye diagrams and hence BER's are vanishingly low. This is why there is no real need for any error correction with these formats.

    However the main thrust of what I have been trying to explain is the effects of clock jitter on the Analogue to Digital conversion, which has nothing at all to do with data transmission. Of course the elements of clock jitter are the same, as I have pointed out and discussed but there is no relevance to eye diagrams or BER.
     

Share This Page

  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