Issue with Defect Map

[David Raphael]

I'm trying to execute a defect map on an image...

I generated an image with 1 values for hot and dead pixels and 0 values for everything else.

When I go to execute the defect map against an image, it just sits at 0% progress on the morphological transformation.  I've tried it on 64bit and 32bit Windows. 

I may go try it on my Mac next...

If anyone else has had any trouble - let me know!

Cheers,
Dave

[David Raphael]

It works fine on my OS X install.

Any ideas?

[Carlos Milovic]

Hmmm... it may be something related to the parallel processing. Try disabling it. Execute this on the console: parallel --disable

[David Raphael]

I'm on vacation away from the horror of Windows machines :-P

I'll try this when I get back.

Thanks for the thoughts.

Cheers,
Dave

[David Raphael]

I disabled parallel processing, but this didn't eliminate the issue.

Thoughts?

[David Raphael]

Just for grins - I reinstalled PixInsight ...

Now I am getting a nasty exception.  I'm gonna reboot since I've had the VS debugger attached etc...maybe there is some issue with the library loading...

[David Raphael]

Ok - the problem is still there. 

Can anyone else confirm that defect_map is working on Windows 7 x64?

[Enzo De Bernardini]

Can anyone else confirm that defect_map is working on Windows 7 x64?

I can confirm a similar problem on Windows XP 32b, only with structure settings Minimum, Maximum and Median (Mean and Gaussian works perfectly)

Greetings,

Enzo.

[Carlos Milovic]

I'm going to check the code...

[David Raphael]

Is the source that comes with PCL 100% the same as what is distributed?  Because I compiled it from source, and it behaved differently...

Also - I don't mind looking at the source either when it's available...

Cheers,
dave

[Carlos Milovic]

It should be the same... The code I sent to Juan at first was different, since he optimized it and rewrote some functions, but basically is the same thing. I don't see any problem there. Maybe it was a bug in the windows compilation... old code that went through.

[David Raphael]

Juan - can you weigh in on this?

I compiled the code for the defect map from source, but it behaves differently than the library distributed.  Both crash, but at different points.  I'm happy to investigate, but I would need to make sure I am working with the authoritative source.

Thanks!
Dave

[David Raphael]

FWIW - I am still having a problem with the defect map.  Is anyone actually using this tool?

[Carlos Milovic]

I used it a few weeks ago, without problems... Could you upload a screenshot with the parameters, and the defect map image?

[David Raphael]

I thought I was having a problem still - but it is working.  I was using an inverted defect map which takes a "bit" longer to process :-)

Thanks,
Dave