Re: [Openvortex-dev] feature news

[Ryan Underwood]

Hi Manuel,

On Thu, Jul 03, 2003 at 07:34:26PM -0400, Manuel Jander wrote:
> 
> > Don't forget the vortex modem!   Only half kidding, ALSA is talking
> > about generic softmodem support implemented soon, mainly waiting for
> > opensource implementations of e.g. v.92/v.44.
> 
> The modem interface is just another audio DMA engine, as far as i
> understood the code i've read about the modem so far. I don't have such
> hardware... would interesting, specially the modem part itself: QAM
> constelations :)

Okay, there were two types of the Aureal modems, and I only remember
seeing them with AU8810 chipsets.  One was the "Soundcom" combination
soundcard and modem, and another was just "Vcom" which was nothing more
than a modem all by itself.  These pretty much sucked as winmodems, but
maybe better as a linmodem ;)

> > Also, does the vortex have a OPL3 cell in hardware?  It always had a
> > decent sounding Adlib emulation, but I don't know if it was actual hardware
> > or just something running on the DSP.  I have a nice disassembly of the
> > au30dos, although with tons of nice mysterious 0x29xxx addresses in it. :(
> 
> The aux0dos.com program sets the IO ports for the legacy devices.

I see that.  It also hooks Int 21/67/2F, probably to do some evil things
when DOS software uses them.

> routes some ADB DMA channels in a static fashion for audio playback. It
> looks like that the generic ADB functionality can't be used while at
> least the SB pro emulation is enabled... wait a moment !! maybe there is
> some fuzz with the 0x220 port on peoples affected by the stereo swapping
> problem ? Someone once mailed me claiming that the aureal driver
> interfered with his NE2000 ISA board (port 0x300 -> MPU401 port). Maybe
> we should take a look if they are really disabled ?

Hmm, on my system at least there is nothing on 0x220, but I do have a
mpu401 card at 0x330.  Are those addresses hard-coded or can they be
changed?  (EnableHardCodedxxxPort makes me think the former ;) )

> OK back to the legacy device issue, the final question is, how does the
> FM synth audio reach the codec ? Knowing this, probably will give us
> some light about how those 64 WT audio streams get into the output mix.
> The only thing i can think about is that there is maybe another static
> wired mixer we just dont know about yet.

What are the OPL related addresses?  ALSA has a driver for opl3 that we
could point to those, to implement a direct-FM interface & opl3 synth.
That would be fun to play with, if nothing else but to discover more
about the vortex.

> > An independent rear PCM device would be excellent of course but I don't
> > think there's anyway to do that outside of the A3D engine.
> 
> That doesnt seme to difficult. It semes just a matter of routing.

Really? That would be great; some of us use Xine or mplayer for playing
DVDs, which decodes AC-3 in software.  Enabling an independent rear PCM
would allow for surround output 5.1 -> 4.0 (center channel gets mixed
into the stereo channels, LFE gets mixed into both rears).  This way you
do not need a separate surround decoder and can use a nice set of
speakers with analog inputs only (like Klipsch 4.1) or simply two pairs
of stereo speakers.  mplayer/xine handle the dirty work, they just need
to be pointed at the second (and/or third) PCM device in order to place
the sound properly.

> > > I am still working on getting a reassmeblable version of the win2k/nt4
> > > drivers a la tui.
> > 
> > What disassembler are you using?
> 
> Jeff's IDA diassembly is very readable. I've made big profit from it.

I'm a big fan of IDA; still having trouble getting readable VxD
disassembly from it though. (specifically for the au30sbfm.vxd)

-- 
Ryan Underwood, <nemesis at icequake.net>, icq=10317253