To what degree is jitter audible?
Nov 29, 2009 at 6:13 AM Post #2 of 32

nick_charles

Headphoneus Supremus
Joined
Feb 26, 2008
Posts
3,180
Likes
333
Quote:

Originally Posted by 3602 /img/forum/go_quote.gif
I mean, the jitter needs to be at least how high to be heard by the majority of the population?


There are three fundamental sources of evidence on this question.

One is mathematical models such as Julian Dunn's work which places audibility based on the point at which effective resolution is degraded by 1/2 LSB. Dunn places the threshold at about 20ps at 20khz.

Two is sighted evaluations based on ill controlled listening tests or pure speculation, thresholds under such "conditions can be as low as 2ps.

Three is actual controlled blind listening tests published as academic papers, the two most important papers are..

‘‘Theoretical and audible effects of jitter on digital audio quality,’’ Preprint of the 105th AES Convention, #4826 (1998).E. Benjamin and B. Gannon.

Detection threshold for distortions due to jitter on digital audio Kaoru Ashihara, Shogo Kiryu, Nobuo Koizumi, Akira Nishimura,
Juro Ohga, Masaki Sawaguchi and Shokichiro Yoshikawa. Acoust. Sci. & Tech. 26, 1 (2005)

Benjamin and Gannon using signal-correlated jitter place the thresholds at around 20ns. Ashihara et al using random jitter place the thresholds at around 250ns.
 
Nov 29, 2009 at 4:22 PM Post #4 of 32

nick_charles

Headphoneus Supremus
Joined
Feb 26, 2008
Posts
3,180
Likes
333
Quote:

Originally Posted by 3602 /img/forum/go_quote.gif
Seems that I need to dig much deeper into this...thanks.


Be my guest but I have been researching this question for 3 years plus now and I have summarised our current state of knowledge from 3 perspectives, theoretical , subjectivist and empirical, you can take your pick of any of the three.

Personally my money is on door no 3. if you take the subjectivist stance you end up infinitely tending towards zero but never quite getting there and spending vast sums of money in doing so. Once the sub-ps jitter barrier is broken there are several more orders of magnitude that can be called into play as needed. One prominent member here has already stated that he can not only detect 2ps of jitter but different 2ps jitter spectra.

Door no 1 tells you very accurately what damage jitter does in an objective manner but does not help you with actual audibility, since they do not take masking into account and assume that you are listening to full scale 20khz tones at 120db above the hearing thresholds , as if.

Door no 3 says, hey relax your bog-standard audio kit has such low jitter anyway it is not worth worrying about, why go looking for problems unless you are obsessional.
 
Dec 4, 2009 at 7:18 PM Post #5 of 32

SB

Guest
Joined
Jul 20, 2009
Posts
130
Likes
14
Quote:

Originally Posted by 3602 /img/forum/go_quote.gif
I mean, the jitter needs to be at least how high to be heard by the majority of the population?


It isn't or as in the current equipment measures far below what is audible.
 
Dec 4, 2009 at 9:11 PM Post #6 of 32

sahwnfras

500+ Head-Fier
Joined
Oct 17, 2004
Posts
938
Likes
11
Who here can tell me what jitter sounds like? How does it affect your music.

Me personally I cannot tell at all. Now my gear isnt high end, but still on high end gear i cannot hear or tell.
 
Dec 4, 2009 at 9:53 PM Post #7 of 32

fenixdown110

Headphoneus Supremus
Joined
Oct 21, 2009
Posts
2,555
Likes
14
It's very hard to tell. It really isn't much of an issue to worry about.
 
Dec 5, 2009 at 3:47 AM Post #8 of 32

SB

Guest
Joined
Jul 20, 2009
Posts
130
Likes
14
Quote:

Originally Posted by sahwnfras /img/forum/go_quote.gif
Who here can tell me what jitter sounds like? How does it affect your music.

Me personally I cannot tell at all. Now my gear isnt high end, but still on high end gear i cannot hear or tell.



Funny you bring that up since it shows just how silly the high end is in that they think jitter can only affect their equipment. unless your equipment is doing 50-100ns I would not worry about it.
 
Dec 5, 2009 at 7:56 PM Post #10 of 32

AdamWysokinski

100+ Head-Fier
Joined
Mar 18, 2008
Posts
235
Likes
12
Quote:

Originally Posted by 3602 /img/forum/go_quote.gif
I mean, the jitter needs to be at least how high to be heard by the majority of the population?


The most important question that one should honestly answer to [him/her]self - can I hear the jitter in my setup?
As for me - I can't and therefore I don't care
k701smile.gif
 
Dec 6, 2009 at 3:45 PM Post #12 of 32

nick_charles

Headphoneus Supremus
Joined
Feb 26, 2008
Posts
3,180
Likes
333
Quote:

Originally Posted by Stereo_Sanchez /img/forum/go_quote.gif
I try to listen to the totality of a system. Assuming high-end components, jitter is usually pretty negligible, and I have never been able to hear it on a well designed system.


Have you heard it on lesser systems ? and how would you know ? on a poor system there may be layers of noise and distortion how would you know what was jitter - honest question
 
Dec 11, 2009 at 5:54 AM Post #14 of 32

Punnisher

Headphoneus Supremus
Joined
Dec 4, 2007
Posts
2,655
Likes
40
Quick question:

Why is jitter even an issue? Couldn't the device simply cache the data for a short time to negate the effects of even severe jitter? Some kind of buffer I suppose.
 
Dec 11, 2009 at 8:35 AM Post #15 of 32

AdamWysokinski

100+ Head-Fier
Joined
Mar 18, 2008
Posts
235
Likes
12
Quote:

Originally Posted by DanD /img/forum/go_quote.gif
In loudspeaker system moving your head will change the sound more than jitter of digital system. Or in headphones, moving the headphones around ever so slightly.


This is a very good point.
 

Users who are viewing this thread

Top