Calibrating image in vision assistant: strange result and crash

[FrankMathijs]

I am trying to spatially calibrate images i have taken with DSLR camera in a fixed setup.

I successfully did this multiple times in the past. (creating a calibration image in vision assistant, that I use in a Labview script that further processes the images.)
We have a new camera now, which generates larger images so maybe this causes an issue?
Attached a snip of my resulting corrected image, with in the corner the original. Obviously the calibration does something overly complicated.
Info:

-I use Vision Assistant 2018 (32bit),  and would like to do a distortion calibration. In the 2nd attached image you can see my grid-images that I use. 

-The black edge is a 'hardware' masking border, in the bright/white area the samples that we image are supposed to be placed.

Any suggestions on why I get this result and on how to avoid it? A simpler calibration model might be a solution, but this worked previously, so I would like to just fix my system, instead of changing to something else.

[udka]

-The calibration changes when you have changed the camera and it cannot calibrate across different resolutions.
-You need to recalibrate with new camera and it should work with new setup.

[FrankMathijs]

No, I'm sorry that I wasn't clear.

These calibration-grid images are taken with the new camera.

Previously, I just loaded them in Vision Assistant to generate a new calibration model-image, that I could use in Labview.

I now repeated this proces with the new (higher resolution) camera:

1 take images of calibration grid

2 load them in Vision Assistant.

3 I should end up with a calibration model image, but the calibration fails as it seems to end up in an unstable result (or local minimum in the optimation?)

How can I avoid this situation? Is my grid not ideal? Does the black border mess up the calibration? Somethin else?

[FrankMathijs]

Update:
I have now succesfully created a calibration image.
By unselecting some of my grid-images in the training set. I have no idea why the images I excluded were messing up the calibration, I excluded them based on trial and error.

So, I'm still open to suggestions on what is happening, as I will need to repeat this process often in the future.

[udka]

-Refer this link to identify the minimum requirements for the Calibration Grid to be successful- https://knowledge.ni.com/KnowledgeArticleDetails?id=kA00Z0000019LXhSAM&l=en-IN
-Do you edit the previous calibration or Create new calibration?
-Instead of adding images to previous calibration, create the new calibration image with new images alone and see if still error persists.
-In the 3 step what error you got?

[FrankMathijs]

My grid adheres to the recommendation.

I always make a new calibration from scratch, so no old images are even loaded in Vision Assistant

The error can be two things: either I get this warped, folded inside out resulting image. Or Vision Assistant experiences a fatal crash. I think the crash might be something related to the higher resolution, although in taskmanager I'm not close to filling up my RAM.