problem with one channel on a 8-port card, [screenshot]

gregorsk · Post by **gregorsk** » Wed Oct 08, 2008 9:55 pm

Hi everyone,

I've dug into this problem and haven't found anything wrong, no hints in dmesg, or messages. The screen shot, is of the camera to the back_door, the image is present but its like it is overlaped somehow.
http://www.lsc-group.phys.uwm.edu/~gske ... oblem.jpeg

I've switched cameras on that particular port /dev/video3(1) and the screwed up image stays. The camera still records events, but the images it records are in the screwed up state.

Does anyone know where I should start looking?

Thanks in advance,

PS I tried googling the problem and looking in the forums , but didn't find anything remotely close

[code]

Oct 8 16:41:24 surveillance zmc_dvideo0[15803]: INF [back_door: 36000 - Capturing at 5.88 fps]
Oct 8 16:41:54 surveillance zma_m2[29029]: INF [back_door: 297000 - Processing at 6.02 fps]
Oct 8 16:42:08 surveillance zma_m2[29029]: INF [back_door: 297074 - Gone into alarm state]
Oct 8 16:42:08 surveillance zma_m2[29029]: INF [back_door: 297074 - Creating new event 429649]
Oct 8 16:42:10 surveillance zma_m2[29029]: INF [back_door: 297089 - Gone into alert state]
Oct 8 16:42:11 surveillance zma_m2[29029]: INF [back_door: 297093 - Gone back into alarm state]
Oct 8 16:42:11 surveillance zma_m2[29029]: INF [back_door: 297096 - Gone into alert state]
Oct 8 16:42:13 surveillance zma_m2[29029]: INF [back_door: 297106 - Left alarm state (429649) - 42(18) images]
Oct 8 16:42:15 surveillance zma_m7[29051]: INF [office_door: 408922 - Gone into alarm state]
Oct 8 16:42:15 surveillance zma_m7[29051]: INF [office_door: 408922 - Creating new event 429650]
Oct 8 16:42:17 surveillance zma_m7[29051]: INF [office_door: 408931 - Gone into alert state]
Oct 8 16:42:18 surveillance zma_m7[29051]: INF [office_door: 408941 - Left alarm state (429650) - 29(9) images]

[/code]

Post by **cordel** » Wed Oct 08, 2008 10:48 pm

Are all your inputs on /dev/video3 set to the same format, size and palette?
Have you restarted ZM since making any changes to the monitors?

gregorsk · Post by **gregorsk** » Thu Oct 09, 2008 1:35 am

>Are all your inputs on /dev/video3 set to the same format, size and palette?
Yes, all 8 camera's are all set to NTSC, RGB24, and 480x320

>Have you restarted ZM since making any changes to the monitors?
Yes, I've restarted ZM a couple times and I still have this problem.

I probably should note, I saw this problem when I was running BlueCherry Ubuntu a couple months ago, before I switched to Centos 5.2 and haven't seen it for like 6 months, and now its cropped up again.

Thanks for the questions,
Greg

Post by **cordel** » Thu Oct 09, 2008 2:25 am

Any errors in dmesg?
What are your interrupt assignments?

Lee Sharp · Post by **Lee Sharp** » Thu Oct 09, 2008 2:46 am

Put a heat sink on those bt878 chips.

Seriously, an overheating chip can cause this. As can a shared IRQ... Any new hardware?

gregorsk · Post by **gregorsk** » Fri Oct 10, 2008 7:26 pm

Any errors in dmesg?

There's nothing out of the ordinary in dmesgs, not that it would point to /dev/video3 or anything.

What are your interrupt assignments?
<snip>
CPU0
0: 279855899 IO-APIC-edge timer
1: 8 IO-APIC-edge i8042
6: 6 IO-APIC-edge floppy
7: 0 IO-APIC-edge parport0
8: 1 IO-APIC-edge rtc
9: 0 IO-APIC-level acpi
12: 4 IO-APIC-edge i8042
14: 2481579 IO-APIC-edge ide0
169: 24990748 IO-APIC-level bttv2
177: 0 IO-APIC-level ohci_hcd:usb1
185: 0 IO-APIC-level ohci_hcd:usb2
193: 0 IO-APIC-level ohci_hcd:usb3
201: 0 IO-APIC-level ehci_hcd:usb4
209: 30965688 IO-APIC-level sata_sis, bttv3
217: 24067972 IO-APIC-level bttv0, SiS SI7012
225: 35708088 IO-APIC-level bttv1, eth0
NMI: 0
LOC: 279870176
ERR: 0
MIS: 0
</snip>

Post by **cordel** » Fri Oct 10, 2008 8:17 pm

For dmesg, are you looking at /var/log/dmesg or are you running dmesg from the command line.

The two are not the same. dmesg in your log folder is only at time of bootup. If you do a dmesg from the command line it will show everything from bootup till current.
The one you should be looking at is the dmesg from the command line.

gregorsk · Post by **gregorsk** » Fri Oct 10, 2008 9:26 pm

Yep, I always check that one.

Have anyone seen this problem before?
I can't say for sure, but I think a guy that guy that posted about putting a heat sink on the card was probably right. The guy who brought the box, put it in a wooden box for security reasons; its gotta be at least 100+ degree's in there all the time.

Though, does anyone have any other thoughts or ideas?

Cheers,
Greg

[quote="cordel"]For dmesg, are you looking at /var/log/dmesg or are you running dmesg from the command line.

The two are not the same. dmesg in your log folder is only at time of bootup. If you do a dmesg from the command line it will show everything from bootup till current.
The one you should be looking at is the dmesg from the command line.[/quote]

xyz · Post by **xyz** » Fri Dec 05, 2008 10:13 pm

What will xawtv expose when used at the same source? Is the card itself OK? In case you have X-less computer, it's still possible to record images from /dev/video*, use something like: ffmpeg -s 320x240 -vd /dev/video0 test.mpg.

I have an old Boeder capture card (without tuner) which worked fine 8 yrs ago on a 200MHz computer but which shows approx the same unsync pic while used with new motherboards. Have tried several video sources - no sync anyway. I suspect the card needs recapping (changing electrolytic capacitors on the board).

Other ideas - feeding your video via incorrect S-VHS connector, having exchanged signal and ground wires, 50/60Hz will somehow find it's way into the signal path (via PSU)...