From alsa-devel-owner@alsa.jcu.cz  Tue Jan 12 22:33:47 1999
Received: from chekov.Belgium.eu.net (relay.eunet.be [192.92.130.25])
	by marvin.jcu.cz (8.9.1a/8.9.1) with ESMTP id WAA24369
	for <alsa-devel@alsa.jcu.cz>; Tue, 12 Jan 1999 22:33:25 +0100
Received: from (dialup082.liege.eunet.be [193.74.147.82])
	by chekov.Belgium.eu.net  with SMTP id WAA02586
	for <alsa-devel@alsa.jcu.cz>; 
	Tue, 12 Jan 1999 22:30:54 +0100 (MET)
Message-ID: <005301be3e72$6cbbe680$52934ac1@w98bg>
From: "Benjamin GOLINVAUX" <golinvaux@benjamin.net>
To: <alsa-devel@alsa.jcu.cz>
Subject: Re: Suggestions 
Date: Tue, 12 Jan 1999 22:27:55 +0100
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Reply-To: alsa-devel@alsa.jcu.cz
Sender: alsa-devel-owner@alsa.jcu.cz
Precedence: list

[ re 20/24/32 bit soundcards ]
:
:i've read that 20 bits gives more resolution than the human ear by
:quite some factor. 24 is way more, and 32 is almost unimaginably more.
:the human ear is so far from being able to hear 2^32 distinct sound
:levels that the processing work/storage space to support 32 bit sound
:seems dubious, from what i've read ...
:
:anyone care to enlighten me ?
:
:--p

Well... Although I think we might, under some special circonstances, be able
to hear dynamics (S/N)improvement until 24 bits, 32 bits are only used to
save extra precision for many-stages DSP whose round-off errors do "damage"
the signal a lot...

So... I don't think output DACs with 32 bit resolution are very common...
Some might dither 32 bits down to 24 or 20, which is already very good...

Benjamin.



