From ultra-devel-owner@ultra.jcu.cz  Mon Feb 23 16:11:32 1998
Received: from entry.jcu.cz (perex@entry.jcu.cz [160.217.1.111])
	by marvin.jcu.cz (8.8.8/8.8.8) with SMTP id QAA26776;
	Mon, 23 Feb 1998 16:11:02 +0100
Date: Mon, 23 Feb 1998 16:11:02 +0100 (MET)
From: Jaroslav Kysela <perex@jcu.cz>
To: Bradley Ward Allen <ulmo@Q.Net>
cc: ultra-devel@ultra.jcu.cz
Subject: Re: One feature lacking in USS Lite Linux sound drivers is full duplex SB16/AWE32
In-Reply-To: <199802230422.XAA18577@Q.Net>
Message-ID: <Pine.LNX.3.96.980223121456.15106A-100000@entry.jcu.cz>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: ultra-devel@ultra.jcu.cz
Sender: ultra-devel-owner@ultra.jcu.cz
Precedence: list

On Sun, 22 Feb 1998, Bradley Ward Allen wrote:

> One feature lacking in the USS Lite Linux sound drivers is
> full duplex in the Sound Blaster family, especially the 16 & AWE32 models.
> 
> Check out how Creative Labs uses their full duplex.
> 
> Apparently, they use the 8bit in one direction and 16bit in the other.
> Sometimes it is hard to decide which one to do in which direction,
> but oh well.  At least it *can* do a form of full duplex.

OK. I see some sources for SB 16 - I must say that "dsp" from
Creative Labs isn't very good chip. Full duplex can be used only for
buffered playback/record. Duplex can't be used in mmaped access (or
application should know, that output card is SB and setup one direction as
16-bit and second as 8-bit)...

> Will you support any of the new 5 speaker (what are they called?) cards?
> You should generalize the API so that the program can figure out what
> to do with the available cards.

Yep, this is one from goals of new project. We can use PCM subdevices for
one soundcards (for example GUS PnP have codec and synthesizer chip - both
can do PCM) or we can use 5 channels PCM device.

> Also, another shortcoming with the USS Lite drivers at least wrt AWE32
> support is that annoying repeating and not finishing (basically bad flushing)
> of the data stream, and extremely poor buffering causing lots of skips
> that are really unacceptible for a P133c system.

-----
Jaroslav Kysela <perex@jcu.cz>
Academic Computer Centre, University of South Bohemia
Branisovska 31, C. Budejovice, CZ-370 05 Czech Republic



