[LAD] Xiph.org - Video:Digital Show and Tell - No difference between analog and digitally processed sound.

Fons Adriaensen fons at linuxaudio.org
Thu May 23 12:18:25 UTC 2013

On Thu, May 23, 2013 at 09:38:40AM +0200, Jörn Nettingsmeier wrote:

> moreover, i'd expect src circuits with only -12dB at fs/s to be
> unusable in practise, because the aliasing artefacts would be
> obvious. it means the top octave from 10-20hkz would be polluted
> with junk at -24 to -12dB,

It doesn't have to mean that. If the filter gain is -12 dB at
0.5 Fs that doesn't imply it will be -24 dB at Fs - 10 kHz.

Take a filter for a 48 kHz DAC. It could be -0.5 dB at 23 kHz,
-12 dB at 24 kHz, and -100 dB at 25 kHz. Any aliasing will be
either above 23 kHz or below -100 dB, probably harmless.
Given the passband and stopband constraints at 23 and 25 kHz,
the actual value at 24 kHz is more or less irrelevant.

The potential intermodulation effects referred to in the paper
by Julian Dunn are real, but not realistic. If such signals 
(high energy well above 20 kHz) are present you'll have some
serious problems even in a completely analog system.



A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)

More information about the Linux-audio-dev mailing list