search

tjlane / pypad

Intuitive, high-level interface to the CSPAD
Other
2 stars 1 forks source link

view raw data in masking with each quadrant corresponding to a column instead of a row of 2x1's #21

Closed sellberg closed 11 years ago

sellberg commented 11 years ago

doesn't really matter but it makes it more convenient to view (at least for me, since I've used it as a convention with the origin in the lower left corner)

tjlane commented 11 years ago

Posting this image as a reference of what things should look like:

@sellberg correct me if this looks wrong

cheetah_refimg

sellberg commented 11 years ago

@tjlane yep, that looks what I'm used to seeing, each 2x1 is stacked lying down from top to bottom and each 2x1 column represents a quad

tjlane commented 11 years ago

@sellberg I think this is fixed. Can you confirm and if so close the issue?

sellberg commented 11 years ago

looks good when opening genmask, but the script is sneaky and rotates the image back after pressing a button, e.g. borders

reopening this issue...

tjlane commented 11 years ago

@sellberg this looks fixed from my point of view -- can you check it out to confirm?

Also: I think we did this before, but is there a way to test that the pixel mask we are writing in cheetah format is correct?

sellberg commented 11 years ago

looks good, here is an image of the cheetah output using the pypad mask as input, which looks consistent with what I made in pypad. maybe there's a better way of testing this..

screen shot 2013-07-24 at 18 04 47

closing this issue.

sellberg commented 11 years ago

I was too hasty in my statement.. So the image now looks good in the sense that a quadrant is a column of 2x1's, but now when I mask manually the masking is out of position:

MARKED FOR MASKING screen shot 2013-07-25 at 15 35 27

MASKED AREA screen shot 2013-07-25 at 15 35 42

I'm pretty sure it wasn't like this during the beamtime, and must have happened when you did the switch of the direction, so it's probably a transpose issue.

opening this again...

tjlane commented 11 years ago

working on this -- I am baffled currently

tjlane commented 11 years ago

@sellberg try once more :). That was a sneaky bug!

sellberg commented 11 years ago

@tjlane tried it again, and the masking looks good, but the whole array has the wrong size, where the array is 1552x1480 instead of 1480x1552 (see image and you'll see what I mean)

ARRAY SIZE screen shot 2013-07-31 at 15 10 07

MASKING screen shot 2013-07-31 at 15 21 12

screen shot 2013-07-31 at 15 21 28

tjlane commented 11 years ago

@sellberg can you confirm, once more, that everything looks good and (most importantly) works in cheetah!

sellberg commented 11 years ago

@tjlane OK, so I've checked the mask in pypad and it seems to work fine now. I also ran it as a test in the cheetah (r0013 - 150 hits) and the written mask seems to mask the right area. However, when I looked closer at the saved masked image I saw that it seems like the automatic masking is ignored in the written pixel mask, since it's not visible in the cheetah image... I'm tempted to close this issue, but will open a separate one about the automask. ;-P

PYPAD - ONLY AUTOMASK screen shot 2013-08-01 at 11 43 55

PYPAD - WITH MANUAL MASKING screen shot 2013-08-01 at 11 44 22

CHEETAH - NO MASK screen shot 2013-08-01 at 12 00 50

screen shot 2013-08-01 at 12 01 00

CHEETAH - MASK screen shot 2013-08-01 at 12 03 41

screen shot 2013-08-01 at 12 03 48