On 08.07.2012 09:22, Eric Ding wrote:
On 07/08/2012 02:52 PM, Oleksij Rempel (fishor) wrote:
Hi, can you please do one more test.
Try to reproduce it with MJPEG stream. Use this command:
luvcview -f jpg -C
it will force jpeg (i think your cam support it) and store each frame
separately. If you can reproduce it, please send me one of this frames.
Sure, I reproduced it fairly easily (with the kernel corresponding to
commit e1620d5). I've attached the first frame dumped by luvcview (I
already converted it from raw MJPEG to JPEG w/ Huffman tables to save
you the trouble).
Eric
P.S. I actually attached a video showing another example of this problem
to my original bug report on Launchpad:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1017072/+attachment/3202254/+files/SecondTime.avi
Ok,
i have no more doubt about quirk.
The image should have fallowing path inside of the cam:
- sensor
- demosaicking processor
- jpeg compressor
- uvc packer
since correct jpeg and uvc headers are produced, then the problem is on
first two steps.
I also tested my cam with USB_QUIRK_RESET_RESUME without any
regressions, in case you wont all_logitech_quirk. My cam is:
+ /* Logitech Quickcam Notebook Pro */
+ { USB_DEVICE(0x046d, 0x0991), .driver_info =
USB_QUIRK_RESET_RESUME },
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
