CaLIGHTs v3.1.0.91 Released

4 views
Skip to first unread message

wols...@gmail.com

unread,
Dec 16, 2020, 10:02:22 PM12/16/20
to CaLIGHTs
I have released an update that contains the fix to the issue that John M reported.
The standard deviation calculation had a software bug that would only appear with images that had odd valued widths.  The calculation would end up trying to figure out the square root of a negative number.  This situation has been fixed but it has ramifications.

John needs to regenerate his masterBIAS and masterFLAT files.  The masterDARKs look ok but I would strongly suggest that he regenerates his masterDARKs as well.  I suspect that he should also regenerate his masterDARKFLATs. I hope he still has his originals.  I've gotten into the habit of buying large external USB drives and backing up all my photos.  A month ago I bought a 4TB USB drive.  This year I took over 150Gb of astrophotos.

This guidance to regenerate your masters also applies to any other user whose camera's full size image have a width that is an odd number. 
 
The reason for regenerating your masters is that the standard deviations for master files is only calculated when the master is written. The standard deviation and means values are written to the FITS header in these master files.  This allows me to quickly read back these important statistics without performing the entire calculation again. The standard deviation calculation is primarily used to identify hot/cold pixels.  When John specified a masterBIAS frame for his masterDARK the faulty standard deviations in the masterBIAS caused the bad pixel logic to decided that over 50% of all the pixels were bad. The only method for fixing this issue is to regenerate the masters using v3.1.0.91. There has been some logic changes to the bad pixel logic in v3.1.0.70 which might have contributed to this issue.

A general guidance for astrophotographers is to regenerate their entire BIAS and DARK libraries each year because imaging chips have a history of developing addition bad pixels as time goes by.  If you do this then you should also use unique folder names so that you can keep track of masters from different years.

Finding an error in my software is never fun...but it's nice to know it's been found and, hopefully, fixed.

Peter
Reply all
Reply to author
Forward
0 new messages