[slinkelist] Sony CDP-CX300 Digital of Analog Out?

Steve Weed weed@lexmark.com
Fri, 4 Aug 2000 11:30:19 -0400 (EDT)


> How would you recommend setting up a system with Sony CDP-CX 90ES, 255, 
> 235, 270, and 225 CD changers and a Yamaha RX-V995 receiver?

I have no experience with that Yamaha,
but since this is the slinkelist,
have you considered the DXS?

Jitter was really more a problem for early separate D-A products
that had neither buffering nor particularly sophisticated
clock extraction.  Early enthusiasts who were sufficiently unhappy with
the quantizing artifacts, filtering and/or other analog output limitations
of CD players were presumably satisfied with some improvement, then
could try for even more by adding other boxes for jitter reduction.

I feed the digital output of my CDP-CX300 to a Lexicon MC-1
and its analog output distributed via a headphone amp to systems in
other rooms, including a Denon 5600.  There is nothing particularly
wrong with the CDP-CX300 analog output, but the Lexicon redigitizes
all inputs, so avoiding two conversions is an easy choice.

Some of the music modes on my older Yamaha surround processor were
very nice; digital inputs should work better if you ever exploit
those on your RX-V995.

The fellow who previously owned my Denon also had a 90ES.
Because he paid the ES premium for it, he uses its analog output.
He moved on to a Lexicon DC-2 and Sonic Frontier Power 3 tube amplifiers.

I felt the 90ES digitial link to the Lexicon sounded better without question,
while the Denon was a tossup between digital and analog,
but my friend always chose the analog link.  On the other hand,
he has played electric guitar loudly and for a long time,
and I suspect that he favors a somewhat compressed dynamic range.

With your 5, going on 6, CD changers, I would if possible suggest
TOSlink for all, to minimize ground loop problems.  At the
other extreme, the probability of at least one marginal RCA ground
out of 24 is too high for a pessimist.

My cheap recommendation is to modify Sony TOSlink outputs to disable
them when not playing, then use the Radio Shack optical splicers
to double up on inputs.  This works not so good with the Denon,
because it has such a long delay when recapturing digital lock,
but you already presumably experience some delay when switching
inputs on the RX-V995. I have modified out-of-production Audio
Alchemy dejitter boxes to switch between two TOSlink signals,
but they are not smart enough to distinguish whether
an unmodified Sony changer is playing or simply powered on.
Sony and MSB have TOSlink switchers with which I have no experience.
I understand that the MSB is equally dumb and uncontrollable,
and the SBD30 appears to only switch manually.
You could integrate Slinke with X10 to power off changers not
playing, if you are willing to live with the restriction of not
immediately switching from one Sony to another on the same Yamaha
input.

On a slightly different topic, allow me to disagree with some of

http://www.digido.com/jitteressay.html

... in particular the seeming suggestion to reducing jitter
by going through multiple D-A and A-D conversions.
The bad effects of jitter are manifested when converting
the digital signal to analog, and once that has been done,
any jitter artifacts are there and will continue to be part
of the signal when it is re-digitized.

On the other hand, it is certain that a digital copy can have
less jitter than an original, and the copy can be digitally perfect unless
the jitter is so bad that the data is misclocked (unlikely and detectible).

Uncorrected jitter typically causes a D-A to generate high frequency artifacts
in the resulting analog signal.  In many cases, these will be above
the frequency response of the rest of the system, so will not be
directly heard.  However, through the magic of intermodulation distortion
and heterodyning, side effects can be audible.  Since there are different
kinds of jitter and ways for the artifacts to be manifested, there is
no single 'sound of jitter'.  If you have access to a digital recording
of pure single frequencies (sine waves), some jitter artifacts can be
more easily heard on some systems.

Unfortunately, jitter artifacts can sound like quantizing artifacts.
If you believe in blind testing, then focus on sibilants
and listen for an irritating 'edge' on stringed instruments.
If I can hear realistic 'spits' and 'blats' from good recordings
of brass and woodwinds without having shimmer and sheen of
cymbals and brushes dissolve into pink noise,
then I stop worrying about jitter and start enjoying the music.

- Steve Weed