Memory problem, and repeat capture bug #2
Description
Hello,
I am trying to use this software with a C9100-23B camera, and am having trouble with the basic example code. Running Python 3.8.8 on Windows.
I am able to take one set of images from the camera using the code below, which is heavily based on the example code provided:
def Snap(self,frames):
from hamamatsu.dcam import dcam, Stream, copy_frame, ECCDMode
returnvec=[]
with dcam:
camera = dcam[0]
with camera:
print(camera.info)
print(camera['image_width'].value, camera['image_height'].value)
# Simple acquisition example
nb_frames = 5
camera['ccd_mode']=ECCDMode.NORMALCCD
camera['readout_speed'] = 1
camera["exposure_time"] = 5
with Stream(camera, nb_frames) as stream:
camera.start()
for i, frame_buffer in enumerate(stream):
frame = copy_frame(frame_buffer)
returnvec.append(copy.deepcopy(frame))
print(f"acquired frame #%d/%d: %s", i+1, nb_frames)
print("finished acquisition")
return returnvec
However, if I run the function a second time; the camera takes the right number of images, but now uses its default settings for exposure time, EMCCD mode, rather than setting up the camera correctly with the ones provided in this function.
Trying to debug this problem I tried a few things, and one thing I noticed was that when I run the similar code block as a standalone python script from command line, it fails with a memory error at the end. The program and output are below. I wonder if cleanup of the objects at the the of the 'with' scope is not occurring properly and this is what is preventing the camera from being able to take more than one image set in a given program?
I am not sure what else to try to debug, but would be very grateful for any guidance.
#!/usr/bin/env python
from hamamatsu.dcam import dcam, Stream, copy_frame, ECCDMode
import numpy as np
returnvec=[]
with dcam:
camera = dcam[0]
with camera:
print(camera.info)
print(camera['image_width'].value, camera['image_height'].value)
# Simple acquisition example
camera['ccd_mode']=ECCDMode.NORMALCCD
camera['readout_speed'] = 1
camera["exposure_time"] = 5
with Stream(camera, 1) as stream:
camera.start()
for i, frame_buffer in enumerate(stream):
frame = copy_frame(frame_buffer)
np.savetxt('out_snap.txt',frame)
print(f"acquired frame #%d/%d: %s", i+1)
print("finished acquisition")
$ ./snap.py
{<EIDString.BUS: 67109121>: 'OHCI 1394', <EIDString.CAMERAID: 67109122>: 'S/N: 000342', <EIDString.VENDOR: 67109123>: 'Hamamatsu', <EIDString.MODEL: 67109124>: 'C9100-23B', <EIDString.CAMERAVERSION: 67109125>: '1.21.00.D', <EIDString.DRIVERVERSION: 67109126>: '9.28.14.5162', <EIDString.MODULEVERSION: 67109127>: '16.10.642.5162', <EIDString.DCAMAPIVERSION: 67109128>: '4.00'}
512 512
acquired frame #%d/%d: %s 1
finished acquisition
Exception ignored in: <function DCAM.del at 0x00000000035423A0>
Traceback (most recent call last):
File "C:\Users\jonesb\Anaconda3\lib\site-packages\hamamatsu\dcam.py", line 1797, in del
File "C:\Users\jonesb\Anaconda3\lib\site-packages\hamamatsu\dcam.py", line 1857, in close
OSError: exception: access violation reading 0xFFFFFFFFFFFFFFFF