Text 1580, 67 rader
Skriven 2011-04-22 22:46:22 av Mike Luther (1:117/3001.0)
   Kommentar till text 1577 av Eric Oulashin (71.fido_os2)
Ärende: Poppy sound in OS/2
===========================
Hi Eric ..

 EO> I used to use OS/2 for a bit in the 90s, and I
 EO> recently installed OS/2 into a
 EO> virtual machine so I can use it again.  One thing I 
 EO> noticed, both back in the
 EO> 90s and now, is that there are quite a bit of pops 
 EO> when sound is played in
 EO> OS/2.  That doesn't seem normal..  Does anyone know if 
 EO> there's a way to fix
 EO> that? 
 EO> Eric

As was noted to you in an earlier reply, this is usually a sound card device
driver problem.  However, it also can be an issue in which the interrupt that
is being used for the sound card device driver is compromised as part of an
issue where that same IRQ is being used by more than the sound card device
driver.

To really work at this the usual awareness starting point is to know what
hardware is there for you for sound service, either as part of your
motherboard, or whatever plug-in board you are using for sound work. From there
you really need to know what IRQ is being used for the sound device, either
motherboard or plug-in board as this may be.  The best way to really start at
this issue is to use hardware for the sound device which does NOT share an IRQ
for this.

This isn't easy with later hardware.  Now for whatever as to the IRQ game. 
Yes, it is possible and OK if different device drivers, such as your audio
driver and maybe your network interface device need to share the same number
IRQ.  But ONLY if the device drivers in use for the OS/2 operations are
PROPERLY written to share a common IRQ number.  Which is where a lot of work
has gone into how to handle sound on OS/2 hardware,primarily in the last few
years with the UNIAUD project.  And here we go again on problem after problem
that seems to get fixed for one type sound device, but gets busted for another.
 A good example of that is the 'fight' that goes on between on-board audio
chipsets and SB-Live PCI plug-in audio cards.  After a LOT of contributed
development work with UNIAUD for on-board Intel and other mobo chips, what WAS
working with only certain models of the SB-Live PCI audio cards is still not
really working well with them now.  OK .. on my Intel based OS/2 MCP2 latest
motherboard systems, a later/latest version of UNIAUD does work, but really
only well if the sound device uses a SEPARATE IRQ that does not conflict with
the NIC network operations for these mobos.  I can,however, use the older 1.0.5
version of UNIAUD as long as I decide to use one of a stack of SB-LIVE 220
model PCI audio cards, with a dedicated audio IR@ that I can set with it on the
BIOS for the mobo.

As well, there has been a lot of this clicking and freezing stuff with the
audio as part of issues that are present with various versions of the ACPI OS/2
device driver operations.  Which work on some motherboards but not on others. 
And are a specially significant problem on 'modern' laptops which we now must
tackle for whatever.

I think your best path to solve this is to do the Netlabs research and reading
for the development which is going on for the UNIAUD project. Get the URL for
that interface and see what you can focus upon with that research first.  Then
with the information you can give us on the hardware, the IRQ use and so on,
hit back here to the area and see what we can really help you with.  It is
complicated..


--> Sleep well; OS/2's still awake! ;)

Mike @ 1:117/3001

--- Maximus/2 3.01
 * Origin: Ziplog Public Port (1:117/3001)