Actually how you rip your audio has a lot to do with how good the final
output stage will be.

If you don't use a quality ripper with excellent error correction (such as
EAC in "secure" mode) then you risk the same problems with your ripped audio
that you would get if you threw the disc in a cheap transport.

-----Original Message-----
From: discuss-bounces (AT) lists (DOT)
[mailto:discuss-bounces (AT) lists (DOT)] On Behalf Of Mark Bennett
Sent: Thursday, February 26, 2004 3:12 PM
To: Slim Devices Discussion
Subject: [slim] [OT] Audiophiles: Can the audio reproduction of the CD
be equalled or bettered?

Pat Farrell wrote:

> At 04:12 AM 2/26/2004, Simon Turner wrote:
> To my view, the beauty of a SqueezeBox system is to remove all
> attractiveness of the audiophile $5000 "cd transport"
> The $300 Sbx probably has a $20 DAC, but I can replace it with one
> costing $200, $1000, or even $20,000 if I want to, and not have to
> touch my CDs or server. Very cool.

This is exactly my position and I can see no reason why it shouldn't
be true.

>> In addition is it certain that the streaming process itself will not
>> degrade
>> the sound, or even that the digital-out on the Squeezebox outputs exactly
>> the stream it receives?

> I can't talk about how bit accurate the SqueezeBox outputs are, but
> streaming is
> just data transfer. Using TCP/IP, it is error protected and ordered.
> It is just a pipe, as long as there are no big drop outs.

Completely seconded on the network front. If this wasn't true then
computer networks as we know them would never work, and they
clearly do.

I also can't believe that the squeezebox or server is corrupting
the data - if it is then it's pretty subtle. It would be very
interesting to capture the digital stream coming out of the
squeezebox and then comparing it with the original. As long as
you're using WAV as the source and get the data alignment right
it really should be identical.

This is theoretically possible if someone has a PC with S/PDIF
in. Unfortunately I do not, although I'll add it to the spec
of my intended SlimServer machine to allow me to find out.

> The beauty of digital audio is that it is digital, it flat works or does
> not.
> No room for subtle degradations.

I'm not so sure I agree with this statement in this context.
In absolute terms you're completely right, but this is different.

The problem with CD player's is that they do occasionally read
the disc wrongly, and when the error correction is overloaded
they send out bad information. This damages the works/doesn't
work argument, because a few bits here and there wrong in the
data stream, which the mechanism interpolates to fill in for,
would be very hard, if not impossible, to detect subjectively.
As a whole I would expect them to appear as a subtle degradation.

The really old CD players used to indicate when the error
correction failed to correct errors, most players now hide that

I suspect that in most cases the uncorrected error rate is pretty
low, if not zero, but I wouldn't like to put money on it.

The squeezebox really should have zero errors all the time, since
hard disks and computer networks have so much error correction
built in that it isn't likely to happen. If you get glitches on
the squeezebox it's probably going to be because the buffer empties
than any other reason.