PCL Win32 System Exception running ImageIntegration with LN and Driz Data

[Vassar1976]

Moderator:  please move to Bug Reports, should have posted there, sorry! --R

I'm running PixInsight verson 01.08.05.1353 Ripley (x64).

I used BPP to generate calibrated, cosmetic corrected, and registered monochrome frames with Generate Drizzle Data checked  (L, R, G, B).

I used LocalNormalization on the registered frames from BPP.

I then attempted to run ImageIntegration with both L.Norm and Drizzle files added.  They were added successfully (all images had <n> <d> prefixes in the Input Images window).

I clicked Apply Global in II and the Process Console showed the .xisf, .xnml and .xdrz files all being opened, but then I got a PCL Win32 System Exception.  See the attached image for the hex code (read from address zero).

I removed the Drizzle files and reran the ImageIntegration process with no errors.

Let me know if I can provide any other data to help.

Best,
Rowland

[Juan Conejero]

Hi Rowland,

This is a null pointer access, so it is obviously a bug. We cannot reproduce it with our data. Can you please upload a data set where this can be reproduced, including the registered light frames you are integrating, as well as the .xdrz and .xnml files? Thank you in advance.

[Vassar1976]

Hi Juan,

Well, wouldn't you know it, now it's working!  Must be Heisenberg's fault. 

I have restarted PI since getting the error and trying to reproduce it today, so looks like the restart changed initial conditions.

If it recurs I'll post some files.

Thanks,
Rowland

[johnpane]

I'm on a Mac and experiencing a problem that has some similarities to this. Here is excerpt from the Process console. I quit and restarted PixInsight and the problem persisted.

/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5028_c_d_a_r.xisf
/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5028_c_d_a_r.xnml
/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5028_c_d_a_r.xdrz
Loading image: w=5496 h=3670 n=3 RGB Float32
* Retrieved data from file cache.
45 FITS keyword(s) extracted.
Weight          :     0.72451    0.72451    0.72451

/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5029_c_d_a_r.xisf
/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5029_c_d_a_r.xnml
/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5029_c_d_a_r.xdrz
Loading image: w=5496 h=3670 n=3 RGB Float32
* Retrieved data from file cache.
45 FITS keyword(s) extracted.
Weight          :     0.71451    0.71451    0.71451

/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5030_c_d_a_r.xisf
/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5030_c_d_a_r.xnml
/Users/pane/NOT BACKED UP/2017-09-24 astrophotos/andromeda/JFP_5030_c_d_a_r.xdrz
Loading image: w=5496 h=3670 n=3 RGB Float32
* Retrieved data from file cache.
45 FITS keyword(s) extracted.
Weight          :     0.70796    0.70796    0.70796


Integration of 77 images:
Pixel combination .................. average
Output normalization ............... local
Weighting mode ..................... custom keyword: INTEGRATIONWEIGHT
Scale estimator .................... iterative k-sigma / BWMV
Range rejection .................... range_low=0.000000
Pixel rejection .................... linear fit clipping
Rejection normalization ............ local
Rejection clippings ................ low=yes high=yes
Rejection parameters ............... lfit_low=5.000 lfit_high=2.500
Large-scale rejection clippings .... low=yes high=yes
Large-scale rejection parameters ... lsr_layers_low=2 lsr_grow_low=2 lsr_layers_high=2 lsr_grow_high=2

* Using 1676 concurrent pixel stack(s) = 8187.22 MiB

* Integrating channel 1 of 3:
*** Error: Unknown exception
<* failed *>


If I turn off "Generate drizzle data" the issue does not occur.
Or, if I change the weights from using a FITS keyword to noise evaluation, it also works while allowing me to generate drizzle data.
All of this is reproducible, even after rebooting the computer.

[johnpane]

Edited: see below for earlier message, which was incorrect. Sorry for my incompetence.

By turning off pixel rejection I was able to reduce this to a minimal set of three images that reproduce the problem.

I placed the images, xdrz, xnml, and a copy of the ImageIntegration instance code in this zip file:

https://www.dropbox.com/s/iob9tcy2r8k9pz4/PixInsight%20ImageIntegration%20bug%20report.zip








--- earlier message (INCORRECT) ---
The problem, in my case, is that the keyword I specified for the weights does not exist in the image files (through my own fault). Instead of reporting this problem, PixInsight is reporting this "unknown exception".