Re: [linux-audio-user] Re: 192kHz

From: Brent Busby <brent@email-addr-hidden>
Date: Fri Jan 27 2006 - 20:31:44 EET

On Fri, 27 Jan 2006, Florin Andrei wrote:

> Some people (sound/music professionals) do report an audible improvement
> of the sound quality when using 192.
> Even from an objective standpoint, wave forms of the high end of the
> audio spectrum are slightly different when using 192. It is conceivable
> that those differences can be perceived in some rare instances.

I think one of the earlier posters in this thread got it exactly right
when he said that it's not so much that we want to reproduce frequencies
that nobody can hear, it's that we'd like to have enough resolution to
reproduce the waveforms of the ones we can hear well enough to be able
to tell what shape they were supposed to be. As you go into the higher
octaves, having only a few samples-per-wave resolution gives horribly
approximated versions of each wave. Sure, a nice filter can smooth off
the peaks and nastier bits, but you're still just trying to repair a
wave that's had some of the information taken out of it -- and this for
frequencies that are within our hearing range.

It never makes sense to sample the analog world as "just enough"
resolution. Good flatbed scanners scan at 45-bit color so they can give
you 24. Good audio gear samples at 24-bit resolution so you can master
at 16. It's never a bad idea to leave some room on sampling frequency.
Things like that contribute to that "dithering noise" metallic sound
that analog recordists never had to worry about. They were cursed with
more background hiss and a lower dynamic range, but since the human ear
already has to deal with noise and signal strength issues in daily use
of our sense of hearing, I think things like that on analog recordings
have always been easier for a listener to psychologically tune out than
that metallic "sheen" in the upper octaves that bad digital recordings
(especially over-processed ones) seem to always have. Nature never has
that particular kind of distortion, so it's harder for the listener to
overlook the way we've always forgiven phonograph and tape recordings in
the past.

-- 
+ Brent A. Busby,   UNIX Systems Admin	 +   "It's like being	+
+ James Franck / Enrico Fermi Institute	 +    blindsided by a	+
+     The University of Chicago		 +    flying dwarf..."	+
Received on Sat Jan 28 00:15:07 2006

This archive was generated by hypermail 2.1.8 : Sat Jan 28 2006 - 00:15:07 EET