On 2009-02-16, Serge Auckland wrote:
It's also rather interesting to see how much jitter is audible. It's all
very well to say digital is crap because of the jitter, but it takes a huge
amount of jitter before it's audible. As a reference, see

http://audiopages.googlepages.com/JitterAudibility.pdf

There are also several papers by Julian Dunn on the adibility of jitter.
See for example http://www.nanophon.com/audio/jitter92.pdf.

Considering that turntables and tape machines have hundreds of time the
jitter of CD players, nobody complains that they sound crap because of the
jitter. Maybe because it's called wow and flutter, and that sound nice and
analogue rather than nasty digital jitter.

Dunn's audibility curves (if I understand them) show that a lot of
jitter can be inaudible at low jitter modulation frequencies (the "wow"
end of the spectrum) but that at higher jitter modulation frequencies
*much* less jitter is apparently tolerable.

I have observed very wide differences in some of the jitter figures
measured by the audio magazines. From a few tens of picoseconds to the
one microsecond plus level. Unfortunately they do not qualify these
with any reference to jitter modulation frequency.

Audibility of jitter has been an interest of mine recently given that
it has acquired "bete noire" status in some audio circles. However I
still haven't acquired enough material and understanding to decide if
this is a real problem or not.

--
John Phillips