"Sonnova" wrote in message
On Wed, 14 Nov 2007 15:11:45 -0800, Arny Krueger wrote
(in article ):
"DJ" wrote in message
It's supposed to upsample CDs to SACD. Has anyone heard
about this player, and better yet, auditioned or own
one?
Reference -
http://www.emmlabs.com/html/audio/cdsa/cdsa.html
The basic premise is ludicrously flawed. No mechanical
or electrical process can accurately recreate music that
isn't already present in the recording.
True, but oversampling does tend to make Redbook CDs
sound better.
.....Right up until you level-match, time-synch, use a really good resampler,
and start trying to control bias.
Perhaps its the removal of that brick-wall
filter at 22.05 KHz that makes things sound "better",
Similar means of comparison shows that a brick wall as low as 16 KHz can be
difficult or impossible to hear.
http://www.pcabx.com/technical/sample_rates/index.htm
I don't know.
Nobody knows because it never seems to actually happen.
But something sure sounds better.
Interesting that removing trivial audible cues and the power of suggestion
has such predictable effects.
I've
performed double-blind tests with my friends, and
everyone preferred the oversampling on my outboard D/A
converter turned on rather than turned off, could dteect
the difference almost every time and I concur.
Just addressing bias isn't enough. The level-match and time-synch thing is
very important.
I also
find that
44.1KHz digital upsampled to 88.2 KHz sounds better than
upsampling it to 96 KHz, but DAT (48 KHz digital) sounds
better upsampled to 96 KHz than it does upsampled to 88.2
KHz.
If there's an audible effect, then it speaks to the quality of the
resampling. I've definately seen resampling gone wrong. Resampling down
usuallly involves two stages of low-pass filtering, and that makes two
places where audio products can and have gone wrong. Upsampling involves at
least one stage of low-pass filtering, and while there's less chance for
error, it doesn't mean no chance for error.
I don't pretend to understand why. It must have
something to do with one upsampled rate being an exact
multiple of the original sampling rate of the disc/DAT
and the other not.
It is well-known that resampling involving integer multiples or integer
fractions has no special magic involved with it, no matter what naive
intuition tells some people.