Todd said:
Hi All,
I have two virtual machines. One is XP and one is W7. On W7,
my sound works. The Device Manager says the audio device is
a "High Definition Audio Device" and the drive is by M$.
On XP, the Device Manager says it is an "Audio Device on
High Definition Audio Bus". And, I get a bang mark.
Any idea where I can get this audio driver for XP?
Many thanks,
-T
You realize, of course, there is more than one Virtual Machine
hosting software. Virtual PC 2007, for example, emulates SB16,
and I can tell immediately from your question, you're not using VPC2007.
It sounds instead, like you're using the more fully-featured
VirtualBox. VirtualBox tends to offer enough options, you can
"shoot yourself in the foot"
http://en.wikipedia.org/wiki/VirtualBox
For a sound card, VirtualBox virtualizes Intel HD Audio,
Intel ICH AC'97 device and SoundBlaster 16 cards.[26]
http://www.virtualbox.org/manual/ch03.html#settings-audio
http://www.virtualbox.org/manual/ch03.html#ftn.idp11141632
"Intel HD Audio support was added with VirtualBox 4.0 because
Windows 7 (32-bit and 64-bit versions) as well as 64-bit Windows Vista
do not support the Intel AC'97 controller."
HDAudio generally consists of two pieces. There is a Microsoft
component, and a hardware component. When you download a driver
from a hardware site, generally the driver contains a "hotfix"
for the missing Microsoft piece.
If you install SP3 for WinXP, that should pick up the missing
piece for you. It seems pointless to be running WinXP, without
patching it to SP3 first, as that can save time when doing
Windows Update.
Once you do that, things like a RealTek HDAudio driver (if you
had a RealTek chip), will no longer need to apply the "hotfix".
Whereas if the driver package queries the OS and sees SP1 or SP2,
then that triggers the hotfix.
Since the VirtualBox documentation does not mention what they're
emulating for the hardware layer, I can't begin to guess what
is going to be needed for that.
Once you've installed SP3 (or alternately, installed the
UAA hotfix separately), the hardware layer may remain a
mystery.
http://support.microsoft.com/kb/888111 (UAA description,
bundled in SP3)
What I would do in that case, is look in setupapi.log, for
an hdaudio/<devicename> kind of entry, and then Google that
device name for more hints. When you look at the plug and
play information in setupapi.log, each system bus has a name.
PCI cards might be pci/Ven&Dev, while USB might have usb/VID&PID.
Things like HDAudio, get their own bus name entry. Based on
that, and looking at the end of setupapi.log (inside your
WinXp Guest), you should have ho trouble identifying the
plug and play numbers being fed to the system.
This is from my system, with an Intel Southbridge and an ADI1988
HDAudio chip. This corresponds to the real hardware part of
my sound.
[2011/08/14 09:00:38 3720.226 Driver Install]
#-019 Searching for hardware ID(s):
hdaudio\func_01&ven_11d4&dev_198b&subsys_1043829c&rev_1004,
hdaudio\func_01&ven_11d4&dev_198b&subsys_1043829c
#-018 Searching for compatible ID(s):
hdaudio\func_01&ven_11d4&dev_198b&rev_1004,
hdaudio\func_01&ven_11d4&dev_198b,
hdaudio\func_01&ven_11d4,
hdaudio\func_01
#I063 Selected driver installs from section [A1988] in
"c:\windows\inf\oem19.inf".
I think this might be related to UAA, the Microsoft part.
#I063 Selected driver installs from section [HDAudio_Device] in
"c:\windows\inf\hdaudbus.inf".
In that hdaudbus.inf, I can see things like this. A PCI entry with
its own Class Code, implies hardware designed to a particular
standard. And the Microsoft HDAudBus.sys then takes care
of driving that. The Class Code, may have been passed in a
BIOS table, so this "detection" does not use runtime detection.
The guy preparing the BIOS tables, knows he's supposed to
report the computer has HDAudio, and this CC_0403 is
what is passed to the OS.
%HDAudio.DeviceDesc% = HDAudio_Device, PCI\CC_0403
ServiceBinary = %12%\HDAudBus.sys
The chip specifics, like the oem19.inf part, would handle things
like 5.1, 7.1 or whatever channel mix the hardware has.
This isn't guaranteed to solve your problem, but if I was working
on it, I'd be looking at the tail end of "setupapi.log" with a
text editor.
Best guess,
Paul
