-=> Quoting T Owen to Steve Mctague <=-
-=> Quoting Steve McTague to Bonnie Goodwin <=-
 SM> topic.  Do you think the Nyquiest theorem is fair in saying
 SM> a sampling rate of 44kHz can be used on 20kHz signals?
 TO> Certainly not. This is audibly not sufficient. It *can* be used
 TO> to repro 20K, but not with enough resoloution to sound natural.
 TO> 10 times that rate would be more appropriate, though we aren't
        I can tell instantly that you don't understand digital
technology.  I've said before, it doesn't work that way.  Resolution
in audio is DEFINED by Nyquist's thereom.  It's not just some
"minimum" standard to reach a given frequency.  Higher sampling rates
get you absolutely positively *nothing* but higher frequency response.
There is *NO* higher resolution at lower frequencies by using higher
sampling rates.
        Adding more bits *only* adds more dynamic range.  That's it.
There are no other benefits to adding more bits.  If 96dB is more than
you'll ever use (to even realize 96 dB, you MUST be playing your
stereo louder at 96dB+room noise, which will give you ear damaging
levels.  The average room has 60dB of noise.  One can treat a room to
reduce room noise, but short of playing your stereo in a lab chamber,
you're not likely to need more than 18-bits before you hit the 120dB
"no more dynamic range possible in human hearing" range, not to
mention it's the instant damage range.
        
 TO> there yet in terms of processing and storage. The 16/44 standard
 TO> was adopted in order to get CDs into production, and start making
 TO> money for the record companies. :( 
        This is not true either.  Most audio hardware today doesn't
have low enough noise levels to get even 18 bits of resolution, let
alone 20-bit, 24-bit, or "10 times" 16-bit resolution, as you suggest.
        20 and 24-bit word lengths *are* useful in the studio because
it will allow a recording engineer a *large* leeway in recording
levels before they red-line (clip) it.  This makes sure one will get
the maximum range when shaped down to 16-bits.  It is not really
needed in the average home environment.  This is in no way saying that
digital is perfect.  But word length and sampling rate are not the big
problem areas.  Internal noise, jitter, and studio recording and
mixing techiques are the reason everyone isn't happy with CDs.  LPs
have a natural tendency to mask many of these problems and even
"sweeten" up the sound with 2nd order harmonics.  Even big fat analog
tape has higher noise levels than CD and can bias the sound more.
Also, really good A/D converters and recording equipment is
*expensive*.  A typical DAT deck isn't going to offer the extra
recording range of a Pro Studio setup.  Therefore, recordings that
don't push right up to the red line without crossing aren't being
optimally recorded.  You can red line analog, you *can't* redline
digital without seriously screwing up the sound.  And if you drop down
for safety, you're not getting optimal dynamic range.  I'm all for
24-bit recorders.  I'm not concerned about 24-bit CD Players.  With a
really good DAC and noise-shaping, you can get almost 18-bits out of a
regular CD.  I think this is more than adequate for almost everyone on
the consumer side.  The trick is to improve the recording side.
Better microphones, lower noise equipment, less mixing after the
recording, less multi-mic recordings, better quality digital
recorders.  No messing with the digital master after it's sent to the
CD factory, etc.
 TO> Nope! Anyone with ears can hear the inadequacies in the CD
 TO> format. (No offense intended Bonnie.) Take care.
        No, I think anyone with ears can hear the inadequacies in a
given CD recording.  I've heard many CDs that have no significant
problems what-so-ever, short of the natural problems associated with
the STEREO format itself. 
 * AmyBW v2.14 *
... "Earth:  Mostly Harmless."
--- FLAME v1.1
---------------