From alsa-devel-owner@alsa.jcu.cz  Mon Jan 18 21:50:35 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 VAA09187
	for <alsa-devel@alsa.jcu.cz>; Mon, 18 Jan 1999 21:48:41 +0100
Received: from (dialup145.liege.eunet.be [193.74.147.145])
	by chekov.Belgium.eu.net  with SMTP id VAA25207
	for <alsa-devel@alsa.jcu.cz>; 
	Mon, 18 Jan 1999 21:46:04 +0100 (MET)
Message-ID: <001001be4323$1daf59e0$91934ac1@w98bg>
From: "Benjamin GOLINVAUX" <golinvaux@benjamin.net>
To: <alsa-devel@alsa.jcu.cz>
Subject: Re: RTLinux/ALSA mix
Date: Mon, 18 Jan 1999 21:42:48 +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

:(I'm leaving off RTLinux; that's a whole different platform.)


...

:For comparison, Irix has a process-scheduler latency somewhere under 5
:msec.  The audio API imposes no noticeable latency (and at its core is
:charmingly simple: basically just "write(int nsamples, sample*)",
:"poll()", and "select()").  So by running our multi-priority
:application as shared-space processes, we got an audio latency of
:about 7 msec.  Which is tolerable, and which I'd love to be able to
:get on a platform that isn't circling the drain.
:

of course, tolerable, versus 40ms for many Windows apps (however good they
are told to be for pro musicians)...

(btw, aren't you disgusted by all this hype surrounding virtual studios when
all you get is something really useless for anything but off-line mastering)

BUT we ain't gonna make an FX box of a Linux box if latency <2ms isn't
achieved... I know this is severe but I feel we can do it.....

For me, latency of 7ms is not much better than 40ms : both let you record
stuff, but not give any fx return to musicians playing on stage or in a
studio.....

I consider this discussion as useful because many people would love to be
able to use their computer as a good fx box, since most of us can't afford a
computer, two or three rack-mounted fx boxes, and a multitrack recorder...

So... does anyone feel interested about this integration of ALSA drivers in
a real-time linux kernel extension ?

I don't claim it IS possible, I'm just wondering... BeOS engineers told me
they were able to reduce soft-latency (excluding any sound-hardware latency
such as fifo's, ...) to less than 1ms...

So, why not Linux ?

Regards,

Benjamin.



