Opps, I confused 'jitter' with 'dither'! Definitely not the same and of course my comments on 'dither' don't apply here for 'jitter'. My bad!
Bob P.
Bob P.
I think jitter sounds like the jingle of loads of money in audiophile manufacturers pockets for expensive products. The marketeers love the sound of "jittery" and fearful audiophiles who come forth, wallet in hand, with a concern that they have some intractable and immeasurable jitter problem with their system. (immeasurable problems are immeasurably hard and expensive to fix!) Depending on which "experts" you believe it is either still a problem today or it is something that was caused by poor clock design circuits (or sharing of clocks) in the distant past of digital circuit designs like the early CD. In theory over sampling is meant to drive jitter problems well outside the audible band where it can be filtered out. Also, accurately clocking things out using a buffer with a dedicated and accurate clock is cheap and is known to minimize jitter. In theory it can sound pretty much like any other noise or distortion ....as the jitter creates low level frequencies that were not in the original signal. The problem is exacerbatted if jitter is significant and repetitive (as opposed to random) and therefore correlates to a specifc noise signal rather than whiteband noise. This should not normally be the case with most recent designs, even low cost ones. This is what Bob Katz has to say on what it sounds like; Here are some audible symptoms of jitter that allow us to determine that one source sounds "better" than another with a reasonable degree of scientific backing: http://www.digido.com/portal/pmodule_id=11/pmdmode=fullscreen/pageadder_page_id=52/ |
there is some misinformation in one of the posts above:- >> Depending on which "experts" you believe it is either >> still a problem today or .... yes, it still remains a problem today esp. in inexpensive CDPs & DVD players of which there are plenty in the market. >> it is something that was caused by poor clock design >> circuits (or sharing of clocks) in the distant past of >> digital circuit designs like the early CD. yes, it was caused by poor clock design (one of the reasons was using a high variable xtal for the clock source to keep BOM cost down) in CDPs of yester years. It remains an issue (maybe to a lesser degree) today. >> In theory over sampling is meant to drive jitter >> problems well outside the audible band where it can be >> filtered out. this is clearly not right! You seemed to be on the correct track citing clock designs & then ........... where did you pull out over-sampling to alleviate jitter? Over-sampling was introduced to reduce the analog filtering requirements post-DAC. >> Also, accurately clocking things out using a buffer >> with a dedicated and accurate clock is cheap.... well not cheap enough for Magnavox ($35 at Walmart) or CyberHome ($38 at BestBuy) to put a highly accurate clock in these DVD player models! Cheap is a relative term - you'll probably find a good quality clock in some "ES" Sony model & other models costing several 100 dollars. So, yes, if you are one to consider a $600-$1200 DVD or CD player to be cheap, you have a higher probability of being correct. >> ....as the jitter creates low level frequencies that >> were not in the original signal. by low level, I assume you mean "low amplitude"? Jitter raises the overall noise floor of the instrument thereby masking lower level details. >> The problem is exacerbatted if jitter is significant >> and repetitive (as opposed to random) since we are talking audio here where the jitter issues are almost always clock related, jitter in audio is always repetitive. It is described as periodic jitter or cycle-cycle jitter. It is quantified as RMS jitter or peak-peak jitter. >> This should not normally be the case with most recent >> designs, even low cost ones. How very mistaken are you!! I personally do not think that there is a "sound" to jitter per se. Maybe there is but I have never heard it. What I've heard is its manifestation: loss of low-level details, reduced D.R. - music loses its snap, imaging not being precise, high freq being edgy/sharp/brittle. |
"I would say that it makes everything less distinct" That's the best description, and matches my experience. And it's a problem in both low and mid priced equipment. In low cost it's a problem due to low price components, and in mid priced equipment due to poor or careless design (a great design can be ruined if the PCB is incorrectly laid out). |
Bombaywalla, Thanks, I stand corrected. Oversampling does NOT help drive jitter outside the audible range (I was definitely confused there). I believe, at most, oversampling or upsampling can reduce jitter wideband noise effects by the factor of over or upsampling. Also provided the output clock is not synchonized to the input clock...you can get further jitter benefits from a good upsampling circuit. I agree with your definition of low cost...definitely Walmart is not where I would start for an audiophile system. I meant that you don't necessarily have to spend several thousands of $ to avoid jitter problems these days. Your clarifications helped. |