Sound Crackling When Reading From Hard Drives
Mark Kane
mark at mkproductions.org
Sat Mar 18 09:18:29 UTC 2006
Ariff Abdullah wrote:
> On Thu, 16 Mar 2006 16:50:17 -0600
> Mark Kane <mark at mkproductions.org> wrote:
>> My sound is compiled into the kernel so I'm not sure how to reload
>> the driver without rebooting. I did try the above suggestions
>> though. When trying 32768 and 65536 the crackling in the audio got
>> worse/louder than before when copying data.
>>
> Maybe you're experiencing resource contention caused by irq sharing.
> Take a look at the output of vmstat -i. Try to move the soundcard
> elsewhere (other pci slot). Another thing to consider is PCI Latency
> Timer settings. If your BIOS has the option to toggle the value, try
> to increase it. If not, use this:
>
> http://people.freebsd.org/~ariff/utils/pcilattimer
Thanks again for the suggestions. So far here is what I have:
[mixx941 at amd64:~]% vmstat -i
interrupt total rate
irq1: atkbd0 200599 0
irq3: sio1 1 0
irq4: sio0 1 0
irq6: fdc0 3 0
irq12: psm0 5873314 26
irq14: ata0 1285014 5
irq15: ata1 159907 0
irq16: atapci3 1100937 5
irq17: pcm0 6193491 28
irq18: fwohci0+ 2 0
irq19: skc0+ 19711298 90
cpu0: timer 435955564 2000
Total 470480131 2158
Nothing is sharing the same irq from what I can see. I have not been
able to shut this machine down yet so I have not tried moving the sound
card, but if nothing else is on the same irq will that even make a
difference?
I tried your utility to change the PCI timings as I have not been able
to check in BIOS yet (I don't believe it has those settings anyway). I
performed various tests while increasing the value for the sound card to
several different values. I have never had to do this before so I just
guessed at a bunch of numbers, and then returned it back to 32 which is
where it was originally. Some numbers I tried were 64, 96, 128, 192, 256
(which auto set it to 200), etc. None of the changes seemed to change
the crackling of the sound at all. I stopped and restarted the song each
time after changing the settings and before accessing the hard drive to
test, but still no change. I also tried increasing the timing on atapci3
which is the controller where the test HD for this round was.
Here is the listing of the unmodified times with -l:
PCI Latency Timer agp0 at pci0:0:0: : 0
PCI Latency Timer isab0 at pci0:1:0: : 0
PCI Latency Timer none0 at pci0:1:1: : 0
PCI Latency Timer ohci0 at pci0:2:0: : 0
PCI Latency Timer ohci1 at pci0:2:1: : 0
PCI Latency Timer ehci0 at pci0:2:2: : 0
PCI Latency Timer atapci0 at pci0:8:0: : 0
PCI Latency Timer atapci1 at pci0:10:0: : 0
PCI Latency Timer pcib1 at pci0:11:0: : 16
PCI Latency Timer pcib2 at pci0:14:0: : 0
PCI Latency Timer hostb0 at pci0:24:0: : 0
PCI Latency Timer hostb1 at pci0:24:1: : 0
PCI Latency Timer hostb2 at pci0:24:2: : 0
PCI Latency Timer hostb3 at pci0:24:3: : 0
PCI Latency Timer none1 at pci1:0:0: : 32
PCI Latency Timer atapci2 at pci2:7:0: : 32
PCI Latency Timer pcm0 at pci2:9:0: : 32
PCI Latency Timer emujoy0 at pci2:9:1: : 32
PCI Latency Timer fwohci0 at pci2:9:2: : 32
PCI Latency Timer skc0 at pci2:11:0: : 64
PCI Latency Timer atapci3 at pci2:12:0: : 0
PCI Latency Timer fwohci1 at pci2:14:0: : 32
-Mark
--
GnuPG Public Key:
http://www.mkproductions.org/mk_pubkey.asc
Internet Radio:
Party107 (Trance/Electronic) - http://www.party107.com
Rock 101.9 The Edge (Rock) - http://www.rock1019.net
IRC:
MIXXnet IRC Network - irc.mixxnet.net (Nick: MIXX941)
More information about the freebsd-questions
mailing list