adamdea;699021 Wrote: 
> John, I have two questions
> 1) given your objection to ASRC, does it follow that not only is it
> necessary to use a different method of cleaning up the clock, it is
> necessary to use a multibit dac chip- won't a delta sigma chip require
> rather more drastic sample rate conversion? Is it just the
> asynchronousness (asynchronicity?)which you object to? I appreciate
> that ASRC may end up slightly changing the information in extreme
> cases, whereas all SRC will alter the nominal data values. 
> 2) assuming that we use the buffer and slowly adjustable clock option,
> why now should jitter and/or noise in the S/PDIF signal make any
> difference to the output- is it just the ground plane noise? presumably
> you agree that it is possible to read the data perfectly and to have a
> clock which is not derived from the S/PDIF stream.

My objection to ASRC is that the current implementations out there
don't seem to sound good to me. I don't have an objection to the
concept, but I haven't heard an actual implementation that I like. My
guess is that it's the issue I have with almost all digital filters in
commercial audio chips in general: they seem to have compromised
implementations, my guess is in order to cut costs in the chips. 

So far I have not heard a delta-sigma DAC chip that can beat a 1704,
BUT if done properly they can come quite close. The big issue is to
bypass the internal digital filter and instead use a properly
implemented digital filter.

For example I have taken a 1794 (or was it the 1792, I don't remember
right now) and fed it from an FPGA, I could either send it I2S and use
its own digital filter or put my own digital filter in the FPGA and
bypass the internal filter. Note this filter was a straight forward
SINC brickwall filter, nothing fancy. The FPGA filter sounded much
better. I did this listening with a bunch of people under blind
conditions and everyone prefered the FPGA filter. So even though both
filters were supposed to be doing the same thing something was
different. 

Two obvious possibilities for the difference: the internal filter is
somehow not doing what it should, or the extra processing of the
internal filter is generating more noise on the internal PS and ground
traces causing noise which is messing up the sound. 

I tested the second possibility by putting the DAC chip in external
mode and using a DF1704 digital filter instead of the FPGA filter, it
sounded almost identical to the internal filter. This tends to rule out
the extra processing going on in the DAC chip as the culprit. This
leaves the implementation of the digital filter itself as a prime
suspect.

I am guessing that the designers of these chips cut corners in the
implementation of the digital filters. An implementation of the proper
filter for 44.1 sample rate takes a fair amount of horsepower, which
can lead to a chip costing more than the management prefers. Thus the
designers are under preasure to figure out how to cut corners to
decrease cost, but still meet the specifications. I know from
experience that this is quite common in the industry, I have personally
worked on quite a few DSP chips (mostly for image processing) and there
has always been preasure from management to figure out how to make
things "good enough" to keep costs down. 

Another telltale sign is the graphs that the manufacturers publish for
their filter functions: they don't look anything like what a simple
SINC filter should look like. They are doing something else, I'm not
exactly sure what, I'm not good enough at DSP theory to reverse
engineer the hardware from the graphs, but I do know its not a simple
SINC. 

When I do a simple SINC in an FPGA it sounds much better, but does take
a lot of resources, so my guess is that the designers of these chips are
getting fancy, using something other than a simple SINC in order to get
the spec sheet numbers they are after and still fall within the cost
parameters the management wants, and somehow this does not sound as
good. 

And no I have not done this with every DAC chip on the planet, but I
have done it with several. In order to do my test it has to be a chip
that allows you to turn off the internal filter and use an external
one. Many DAC chips do not allow this. But I have done enough of this
and heard the difference between my simple filter and what is inside
the chips to get a good feeling for difference in sound. I have
listened to quite a few of the chips that don't let you disable the
internal filter and I hear a similar sound to what I was getting from
the chips where I could use an external filter. 

So my conclusion is that this compromise is pretty pervasive in the
industry.

I guess all this is a very long winded way of saying that yes a I think
you can get very good results out of a delta sigma DAC chip, you just
need one that lets you use an external digital filter.

And BTW I think this is also a major reason for  the NOS movement.
Going back to old DAC chips without digital filters  seems  to let
parts of the music "through" which normally somehow get messed up by
the internal digital filters. Of course then you have aliasing all over
the place. For some people its worth the tradeoff. 

Of course the proper solution is to use proper digital filters so you
can have the best of both.  

John S.


-- 
JohnSwenson
------------------------------------------------------------------------
JohnSwenson's Profile: http://forums.slimdevices.com/member.php?userid=5974
View this thread: http://forums.slimdevices.com/showthread.php?t=94352

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

Reply via email to