via 82c 686 a driver probably broken

Logged In: YES
user_id=50766

Oops. Okay, I just now figured out it's NOT my hardware.

More info:

Driver version is

#define VIA_VERSION "1.1.14a"

Player is either ESD or XMMS playing via ESD or via OSS
sound driver.

Here's a very precise description:

1: Playing a piano tune, I hear grating distorted sound
underneath the normal sound.

2: Stop playing piano tune, go to Gnome sound events, and
play some event.

3: I hear in the 'grated distorted' area of the sound the
piano tune I heard before, and in the 'normal' area of the
sound, the short beep that is the system sound event.

What this means to me is that there are two buffers (or one
buffer with two pointers into it?) and that one buffer is
accidentally being played (very wrongly) at the same time as
another buffer is playing (probably correctly, possibly
slightly wrong).

If you play loud, distorted music like Gravity Kills, White
Zombie, or Garbage, you won't notice it so much until they
play something simple and quiet. I REALLY noticed it first
on Rammstein's Angel song, where there's essentially a sine
whistle melody at the start.

If you play soft, subtle music like Chopin or Mozart, you
will notice it like a wart on a supermodel's nose. I can't
even listen to my Chopin without cringing.

Logged In: YES
user_id=50766

Wow. Okay, so whatever the bug is, it's gonna be
intermittent. I just powered down my box and brought it back
up again.

Everything sounds fine.

Now!

If I try to use "mpg123" to play the MP3s, it sounds really,
really bad. I'm not sure, because it's been a week or so,
but I think I had run mpg123 right before these problems
began.

But other variables that may have affected things? I
recompiled my 2.4.1 kernel to 2.4.2 (I know, doubtful this
made any difference) and changed a couple of variables in
the driver .c file (including turning on /proc/driver/via
support).

Since I doubt any of the tweaks I made to the driver
actually fixed anything, I won't go into detail.

I'm pretty sure the driver worked correctly after boot, just
somehow I put it into this nasty broken mode.

So my question becomes finally: is there a way to reset the
state of the audio device other than a hard power down?

Thanks!

Logged In: YES
user_id=217044

Yes. Same problem on my motherboard (Biostar VKB v1.1). I
don't know if this matters but here is part of lspci -vvv

00:07.5 Multimedia audio controller: VIA Technologies, Inc.
AC97 Audio Controller (rev 20)
Subsystem: Sigmatel Inc: Unknown device 7609
Control: I/O+ Mem- BusMaster- SpecCycle- MemWINV-
VGASnoop- ParErr- Stepping- SERR- FastB2B-
Status: Cap+ 66Mhz- UDF- FastB2B- ParErr-
DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Interrupt: pin C routed to IRQ 15
Region 0: I/O ports at ac00 [size=256]
Region 1: I/O ports at b000 [size=4]
Region 2: I/O ports at b400 [size=4]
Capabilities: [c0] Power Management version 2
Flags: PMEClk- DSI- D1- D2- AuxCurrent=0mA
PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D3 PME-Enable- DSel=0 DScale=0 PME-

00:0a.0 SCSI storage controller: Adaptec AHA-294x / AIC-7871
(rev 03)
Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV-
VGASnoop- ParErr- Stepping- SERR- FastB2B-
Status: Cap- 66Mhz- UDF- FastB2B+ ParErr-
DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 32 (2000ns min, 2000ns max), cache line size 08
Interrupt: pin A routed to IRQ 15
Region 0: I/O ports at bc00 [disabled] [size=256]
Region 1: Memory at db000000 (32-bit,
non-prefetchable) [size=4K]
Expansion ROM at <unassigned> [disabled] [size=64K]

So audio is sharing irq with scsi-controller.

And sound works once after modules are loaded. So rh7.1
sndconfig gives clear sound. But playing with xmms (esound
plugin) or mpg123 gives noise with sound.

Logged In: YES
user_id=67848

I have exactly same problems. I tried kernel 2.4.2 (default
from RH 7.1), but it sound like it plays music two times at
once, one "track" seems normal, and one sounds very bad =>
result is _extremely_ bad. My motherboard is MicroStar
MS-6340. When I'm playing first MP3 (in mpg123), everything
is ok, except crash-like sound when playing starts. Every
next sound played is damaged. I tried to update to newer
kernel, but newest 2.4.4-ac14 has same bug.

When I send SIGINT (^C) to mpg123, bad "track" seems to
disappear, I can hear correct sound for a few seconds, then
mpg123 finally dies.

Logged In: NO

ESD is likely to blame. I had the same problem as you
(mpg123 sounds terrible), so I tried compiling without ESD
support and everything is okay now.

Logged In: YES
user_id=17443

This problem is fixed in the latest version of the driver. You
can download the driver from the download section on this
SourceForge web page.