My understanding is that you're manipulating the mirror array in the DMD and then taking an image. It's not clear why you're repeatedly setting the mirror to different states and then taking the image with the last state. The documentation for alpdevice.put is as useless as I've come to expect:
% Sets pixel to given frame
Which makes it sound like the device needs to be updated elementwise. The usage of obj.put in the other methods (e.g. on and off) suggest that the entire frame is updated at once.
I'm inclined to think that you're trying to do it elementwise, but that's not how the loop is written.
% this loop
for i = 1:512, j = 1:384;
% ...
end
% is the same as this
for i = 1:512
j = 1:384;
% ...
end
% in other words, j is a literal vector and is not part of the iteration
If the entire frame can be updated at once, and the desired area is {1:384,1:512} (a quarter frame), then no loop is needed:
% do device setup ...
dmd = zeros(768,1024,'uint8');
dmd(1:384,1:512) = 255;
mydev.put(myframe);
% turn on camera, take image
Image = captureimage();
% do device release ...
Bear in mind that you're preallocating the dmd array using ones(). That means that the background element value is 1 instead of zero. You might want to be using zeros() instead of ones().
As to whether it's possible to interlock the process by verifying the device state, I don't know. I don't immediately see an obvious way to query the mirror state. There might be, but I don't see it yet.
