MarkRivers
commented
3 years ago Because this camera cannot set the frame rate independently of the exposure time, I have built a work around ...
Are you sure about that? Can you post screen shots of the main ADSpinnaker screen and the camera-specific screens?
the memory consumption of the ioc process starts to continously increase with a rate of around 7 mByte/hour at 1s Scan interval
That works out to about 2 KB per image, which is quite small.
What happens if you use a normal internal trigger and acquire at 10 frames/s. Does the memory usage also increase at 2 KB per image?
rettigl
This is the main screen of the ADSpinnaker for my camera. If I set ExposureAuto to Off, and Frame Rate Enable to On, and then set a FrameRate, I get an error in the Epics shell:
This indeed allows some control, but still not what I need. Here is what I observe:
These settings do not produce errors in the ioc shell, the numbers are just capped.
As I need this range of expsoures, this is the mode I have been running before, and used the trigger to control the frame rate to around 1 Hz, independent of acquisition time.
Hi,
I'm using a FLIR Blackfly BFLY-PGE-50A2C-CS GigE camera with areadetector and ADSpinnaker on linux Ubuntu 18.04 and epics 7 and quite recent github versions of support modules, which mostly works fine. However, I am experiencing the following strange behavior:
Because this camera cannot set the frame rate independently of the exposure time, I have built a work around using its software trigger, with the idea to set it to software trigger and use a bo soft channel with periodic scan to write to the "TriggerSoftware" PV of the camera (I only need low frame rates for monitoring, but small exposure times). This functionality all works well, however, with the problem that whenever I have the bo record repeatedly trigger the "TriggerSoftware" PV, the memory consumption of the ioc process starts to continously increase with a rate of around 7 mByte/hour at 1s Scan interval (the rate scales with the trigger frequency, the example below has a trigger frequency of 0.5 seconds). This does not happen if I stop the scanning of the softchannel, but it is independent of whether the camera is actually triggered or not "TriggerMode=Off or On". No plugins apart from an Image plugin are loaded. When I stop the acquisition of the camera, the memory consumption drops to a baseline and stays there constant, but if I resume acquisition later, it resumes at a higher memory consumtion, as if it was running all the time:
The definition of the softchannel PV is the following:
record(bo,"$(P)trigger") { field(DESC,"Camera trigger") field(VAL,1) field(OUT,"$(P)cam1:TriggerSoftware PP MS") field(SCAN,"1 second") }
Is this a bug in ADSpinnaker or the Spinnaker SDK, or am I doing something silly here?
Best, Laurenz