On Thu, 2011-05-12 at 14:16 -0400, Jarod Wilson wrote:
> On May 12, 2011, at 12:49 AM, Greg Fruth wrote:
>
> > Thanks, Gary, the suggestion to use dd=/dev/video0 is brilliant! I had forgotten about low-level
> > troubleshooting, since the card had been working properly for 5+ years.
> >
> > You're absolutely right, the audio has the same problem when using dd=/dev/video0, so it's not a
> > MythTV problem. The dmesg errors look like:
> >
> > [ 1911.814182] ivtv0: Audio has died (Encoder OK) : ivtv_serialized_open
> > [ 1911.919083] ivtv0: Decoder has died : ivtv_serialized_open
> > [ 1911.919090] ivtv0: Detected in ivtv_serialized_open that firmware had failed - Reloading
> > [ 1912.686864] ivtv0: Loaded v4l-cx2341x-enc.fw firmware (376836 bytes)
> > [ 1912.711139] ivtv0: Loaded v4l-cx2341x-dec.fw firmware (262144 bytes)
> > [ 1913.001059] ivtv0: Loaded v4l-cx2341x-init.mpg firmware (155648 bytes)
> > [ 1913.111809] ivtv0: Firmware restart okay
> >
> > These messages appear to be associated with the time I restart dd=/dev/video, not with the time that
> > the audio dies. There's doesn't appear to be a message printed when the audio dies. Googling these
> > errors
>
> I'd ask the ivtv driver maintainer, he may know for sure whether
> or not this is an indication the card is toast. Andy? :)
The CX23415 chip does have multiple processing units internal to it, one
of which is for processing and compressing audio. It is entirely
possible that this unit in the chip is failing while the other are still
OK.
$ cat /dev/video0 > /dev/null &
$ aplay -f dat < /dev/video24
Will let you hear the uncompressed PCM audio that the CX23415 is
receiving from the audio digitizer chip. (Assuming the card is still
set for 48 ksps.)
> (Something this low-level is really no longer mythtv-related, it
> v4l/dvb driver-related, and better discussed on the linux-media mailing
> list, but hey).
For ivtv and cx18 card specific issues, you'll find a focused group of
users on the ivtv-users list:
http://ivtvdriver.org/mailman/listinfo/ivtv-users
[.Crud, I just noticed the main ivtvdriver.org wiki site is down. :( ]
After reading this thread:
http://www.gossamer-threads.com/lists/mythtv/users/479775?page=last
I'll speculate you have one of these problems:
1. Dust in the PCI slots
2. A failing PCI card in the system (PVR-350 or some other card)
3. A failing motherboard PCI chipset
4. A bad disk image (firmware images, or kernel)
The standard PCI bus is a _shared_ I/O bus, which uses reflected wave
switching for signal levels. A problem with one card/slot can cause a
problem for all other cards/slots.
Corrective actions you can take:
1. Remove all PCI cards, blow the dust out of *all* the slots, reinstall
all the cards, and retest.
2. Remove all the PCI cards other than the PVR-350, including any plug
in video graphics adapters, network adapters, etc. Bring the system up
to single user mode (so no X windows) and perform some test captures to
a few files:
$ ivtv-tune -t us-cable -c 5 (channel 5 is just an example)
$ cat /dev/video0 > foo1.mpg
^C
$ cat /dev/video0 > foo2.mpg
^C
$ aplay -f dat < /dev/video24
^C
If you can check the files by moving them to another machine via USB
drive, do so.
If the files are fine. Add back one other PCI card and test again.
Repeat.
3. Test the card in other system or motherboards.
4. Now that you've wasted 10 hours, and assuming you gross a piddling
$10/hour for time that you work, and assuming your leisure time is much
more valuable, you can give up after realizing that a new HVR-1600 can
be purchased for about $100. ;)
Fedora Core 4? That means your card is about 8 years old. A dyning
PVR-350 is likely. Probably time for a new card.
Regards,
Andy
_______________________________________________
ivtv-users mailing list
ivtv-users@ivtvdriver.org
http://ivtvdriver.org/mailman/listinfo/ivtv-users
> On May 12, 2011, at 12:49 AM, Greg Fruth wrote:
>
> > Thanks, Gary, the suggestion to use dd=/dev/video0 is brilliant! I had forgotten about low-level
> > troubleshooting, since the card had been working properly for 5+ years.
> >
> > You're absolutely right, the audio has the same problem when using dd=/dev/video0, so it's not a
> > MythTV problem. The dmesg errors look like:
> >
> > [ 1911.814182] ivtv0: Audio has died (Encoder OK) : ivtv_serialized_open
> > [ 1911.919083] ivtv0: Decoder has died : ivtv_serialized_open
> > [ 1911.919090] ivtv0: Detected in ivtv_serialized_open that firmware had failed - Reloading
> > [ 1912.686864] ivtv0: Loaded v4l-cx2341x-enc.fw firmware (376836 bytes)
> > [ 1912.711139] ivtv0: Loaded v4l-cx2341x-dec.fw firmware (262144 bytes)
> > [ 1913.001059] ivtv0: Loaded v4l-cx2341x-init.mpg firmware (155648 bytes)
> > [ 1913.111809] ivtv0: Firmware restart okay
> >
> > These messages appear to be associated with the time I restart dd=/dev/video, not with the time that
> > the audio dies. There's doesn't appear to be a message printed when the audio dies. Googling these
> > errors
>
> I'd ask the ivtv driver maintainer, he may know for sure whether
> or not this is an indication the card is toast. Andy? :)
The CX23415 chip does have multiple processing units internal to it, one
of which is for processing and compressing audio. It is entirely
possible that this unit in the chip is failing while the other are still
OK.
$ cat /dev/video0 > /dev/null &
$ aplay -f dat < /dev/video24
Will let you hear the uncompressed PCM audio that the CX23415 is
receiving from the audio digitizer chip. (Assuming the card is still
set for 48 ksps.)
> (Something this low-level is really no longer mythtv-related, it
> v4l/dvb driver-related, and better discussed on the linux-media mailing
> list, but hey).
For ivtv and cx18 card specific issues, you'll find a focused group of
users on the ivtv-users list:
http://ivtvdriver.org/mailman/listinfo/ivtv-users
[.Crud, I just noticed the main ivtvdriver.org wiki site is down. :( ]
After reading this thread:
http://www.gossamer-threads.com/lists/mythtv/users/479775?page=last
I'll speculate you have one of these problems:
1. Dust in the PCI slots
2. A failing PCI card in the system (PVR-350 or some other card)
3. A failing motherboard PCI chipset
4. A bad disk image (firmware images, or kernel)
The standard PCI bus is a _shared_ I/O bus, which uses reflected wave
switching for signal levels. A problem with one card/slot can cause a
problem for all other cards/slots.
Corrective actions you can take:
1. Remove all PCI cards, blow the dust out of *all* the slots, reinstall
all the cards, and retest.
2. Remove all the PCI cards other than the PVR-350, including any plug
in video graphics adapters, network adapters, etc. Bring the system up
to single user mode (so no X windows) and perform some test captures to
a few files:
$ ivtv-tune -t us-cable -c 5 (channel 5 is just an example)
$ cat /dev/video0 > foo1.mpg
^C
$ cat /dev/video0 > foo2.mpg
^C
$ aplay -f dat < /dev/video24
^C
If you can check the files by moving them to another machine via USB
drive, do so.
If the files are fine. Add back one other PCI card and test again.
Repeat.
3. Test the card in other system or motherboards.
4. Now that you've wasted 10 hours, and assuming you gross a piddling
$10/hour for time that you work, and assuming your leisure time is much
more valuable, you can give up after realizing that a new HVR-1600 can
be purchased for about $100. ;)
Fedora Core 4? That means your card is about 8 years old. A dyning
PVR-350 is likely. Probably time for a new card.
Regards,
Andy
_______________________________________________
ivtv-users mailing list
ivtv-users@ivtvdriver.org
http://ivtvdriver.org/mailman/listinfo/ivtv-users