Years ago (I think it was with LV 7.1), I created some picture constants for a clockwise and counterclockwise arrow. I would wire these into a picture indicator with a transparent background border and overlay this on an image of my test stand. I remember playing with the mask array of the picture information to get the non-arrow parts of the images to be transparent.
Now (with LV 2011), I am playing with these images. I want to take a mirror image of them. But I'm finding that translating the picture to the pixmap data, then unflattening that back to the picture (with no other manipulation in between) is not giving me the same results.
It appears that the transparency information in the mask array is lost when the original picture is translated to the pixmap data as can be seen by the empty mask array. So when that pixmap data is put back into a picture image, the background is not transparent like it originally was.
Is there a bug in the translation? Am I missing something?
Please see the attach VI in LV2011 that shows the problem in its simplest form.
I took it back to LabVIEW 2009 and saw the same behavior.
I started playing around some more and got it to work by adding the create mask function and defining my mask as the white background. So now it works.
But I don't see why the initial function going from picture to image data didn't return the mask information.
It has always been that way. I usually specify a color like hot pink which is not in the picture as both the background of Flatten to Pixmap and the mask color of Create Mask.