On 11/2/10, Ákos Maróy <a...@maroy.hu> wrote: > but, the image I get from the camera itself seems to be broken - there > are vertical stripes on the image. I wonder if your received the image I > sent you?
Akos, the image looks really bad - I'm afraid this time the board is really broken. It is our fault - we suggested you to take the board out and twist it (something wrong as I wrote before). As the result of such twisting there are either the broken contacts between the pads and the traces - either on the FPGA or on the video memory. ESD damage can also be the cause. The only thing you can try while waiting for the replacement board - try to carefully reinsert the cable that connects the sensor board and the system one, at least at the end where you already disconnected it. But the connection is rather fragile so you might accidentally pull the cable on the other end too while disassembling the camera. > also, the /usr/local/bin/str seems to be taking 108% of CPU time when > running, with 3 such processes reported. I wonder if this is normal. You probably meant 108% of memory, not the CPU - yes, that is normal. The program uses huge (for the camera) 19MB circular buffer with mmap, doing that twice to simplify processing of the compressed image files that roll over - with such mmap usage all the image files can be processed as continuous ones. And 3 instances - well it was designed so, that the first one launches the RTSP server - it wasn't me who wrote that code, so I'm not ready to describe all the details of the str operation. Multiple (3) instances of PHP - that is my responsibility, I configured it lighttpd as a compromise between memory and performance, these instances are waiting to process PHP scripts using FastCGI. Andrey > _______________________________________________ Support-list mailing list Support-list@support.elphel.com http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com