In article , David Looser
wrote:
"Eeyore" wrote in message
...


NO. Pure horse manure. The digital signal is buffered to buggery. A
bit of jitter won't bother it. This is an example of analogue-style
thinking being inappropriately attached to digital signal paths.

It's not "horse manure". No amount of buffering will cure the problem of
recovering data from a badly jittered data-stream. It's the clock
recovery that is affected. I can tell that you've never had to design
clock-recovery circuits.

Rather depends on your definition of "badly". :-)

If you look at the levels of jitter people argue about in audio mags it is
of the order of 100's ps to a ns. I'd have said is quite small in the
context of SPDIF or EBU, although they are vague about how this is
measured. I'd expect this to be easily reclocked and buffered if the
engineer knows what he is doing. In practice, my experience is that even a
decade old DAC like the Meridian 263 or 563 does this with ease.

So does a fairly cheap RX like the one in the Pioneer recorders I have been
using recently. Sample-by-sample comparisons using a player with a disc
that isn't faulty shown sample-by-sample agreement of recordings. Also
agreeing with recordings made directly to computer using a CDROM drive.

Can you define what you meant by "badly"? I can see that if the jitter is
large enough then you will get problems recovering the actual stream of
values. But I can't say I have encountered this with domestic audio kit
I've used unless the disc or equipment are faulty. No doubt, though,
someone will have made kit bad enough for this to be a problem... :-)

Of course, if the 'jitter' has significant amounts at low phase modulation
frequencies then the loop or buffer will need to alter the local clock and
this will affect the output. I can also see that an RX/DAC that doesn't
reclock but simply uses the implict clock in the data will then translate
that into phase modulation of the output. But this does not mean that the
correct series of values was not RXd. Is this what you were thinking about?
That the loop or buffer will respond to close-in phase noise by adjusting
the conversion clock frequency?

Given values for 'jitter' of the order of 100's of ps I can't immediately
see that this would cause a very significant modulation of the DAC clock
rate if smoothed with a sensible buffer and loop.


Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.audiomisc.co.uk/index.html
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html