Error message in Weighted Batch Pre processing

dkahraman

Member
Up to now I have had this script working quite well. Lately I am getting this message:

Global.format(): invalid argument type: numerical value expected

I have no idea what it means nor how to correct it

Dan Kahraman
 
Not after that...I don't know about before. I know it isn't giving a value for FWHM

I have two more error screenshots which are attached. Never seen anything like these before
 

Attachments

  • PI WBP screen_2.png
    PI WBP screen_2.png
    72.2 KB · Views: 67
  • PI WBP screen1.png
    PI WBP screen1.png
    31.2 KB · Views: 65
It isn't one file. It is all of them. Here are all error screenshots. Maybe something is wrong or has gone wrong with WBPP?
 

Attachments

  • PI error messahet_1.png
    PI error messahet_1.png
    73.8 KB · Views: 52
  • PI WBP screen_2.png
    PI WBP screen_2.png
    72.2 KB · Views: 51
  • PI WBP screen_3.png
    PI WBP screen_3.png
    76.6 KB · Views: 60
  • PI WBP screen1.png
    PI WBP screen1.png
    31.2 KB · Views: 54
The exposures were very short (5 seconds) at Gain 8000 with a nearby moon. Maybe the exposures are too short to fill the wells.
 
If you could post a link to one of the the failing files (in full resolution .xisf format) it might be possible to offer some suggestions. The errror reports are not enough to help diagnose the problem.
 
We really need to see one of the calibrated, debayered ..._c_d.xisf files that are causing the crash (it is not the raw .fits file that is being processed).
 
The link has all 19 subs debayered. My Pixinsight stopped working so I have uninstalled it and now I am restarting the computer. Something obviously has gone wrong not only with the latest version of my installed PI but the script Weighted Batch Pre- Processing.
 
Because you selected such a high gain, your images consist mostly of highly amplified noise and saturated stars. This is upsetting the subframe weighting analysis, which fails to weight the frames. Without weights, alignment can't auto-select a reference frame, and fails. If you unselect subframe weighting and provide a manual registration reference image (I used the uncalibrated debayered image of the first raw frame), then you can persuade the WBPP script to run to completion (see image below - with a bit of polishing). There is no problem here with PI or WBPP. The error messages really pointed out the problems with the data.
The main problem, however, is that the exposures are too short. It is tempting to try and use increased gain to compensate for this, but since you are not increasing the sensitivity, you are just amplifying the noise. If you had used lower gain you would not have saturated the few objects that are bright enough to be detected!
1614263927053.jpeg
 
Last edited:
It's actually "0'dan kahramana", which is "zero to hero" in Turkish. "Zero-to-hero" is just the anime-fan name for the genre of anime plots where an initially weak character becomes superhuman. (No I'm not an anime fan!).
 
Back
Top