Patrick Dixon;325171 Wrote: 
> Solving the problem with a buffer is not straightforward.  The read and
> write clocks will never be at exactly the same frequency (and won't be
> constant either), but they are at the same nominal frequency.



Consider the following scenario:  suppose you have a local clock in
your DAC with an adjustable frequency (such clocks exist and are used
in some DACs).  For simplicity we'll ignore jitter in that clock (and
any other local jitter source) and worry only about jitter in the
incoming S/PDIF stream.  Now you record the incoming data stream in a
buffer for (say) 1 second.  Then you start to play out the buffer using
your local clock.  The analogue signal coming out is jitter-artifact
free given the above assumption.  

But now we will run into a problem - our buffer will start to either
fill or empty depending on the mismatch in average timing between the
local and source clocks.  But since our local clock is adjustable, we
can monitor the buffer state and make a (tiny) adjustment.  We'd need
to do that at most every second, but in reality much less often (since
the clock mismatch will be smallish).

As far as I can see this scheme removes all jitter with a frequency
higher than some very low cutoff (which in this example will be around
1 Hz times the fractional clock mismatch).


-- 
opaqueice
------------------------------------------------------------------------
opaqueice's Profile: http://forums.slimdevices.com/member.php?userid=4234
View this thread: http://forums.slimdevices.com/showthread.php?t=50147

_______________________________________________
audiophiles mailing list
[email protected]
http://lists.slimdevices.com/lists/listinfo/audiophiles

Reply via email to