Bad Frame Rejection in WBPP

Bruce Donzanti

I am using a set of darks and dark flats from a permanent setup in WBPP (default settings) to align and integrate Ha light frames of a closeup of the Horsehead Nebula.  I've been doing this for many images over many months with no issues.  However, I am getting about a 20% Bad Frame Rejection rate with this object.  There are not a ton of stars in the subframes but when I blink each sub, they really do not look that bad.  I took some subs at 300s and some at 600s.  While most of the rejected frames were from the 300s group, there were a couple from the 600s group.  I am perplexed by this as I have never had this happen before.   The only difference between this object and the others I have done is the lack of a significant star population and much closer to the full moon.   

Any thoughts as to where to start looking for the problem?

BlockHead

No zeros in the calibrated data correct?

-the Blockhead

Bruce Donzanti

Correct- I did check 

BlockHead

You are going to need to identify the frames that are considered "bad" and compare with "good" ones in your calibrated data. There will be a difference and it should be visible. Remember to zoom in.

-the Blockhead

Bruce Donzanti

ok- I identified the "bad" frames from the good frames that calibrated ones after running WBPP (pedestal was set to automatic).  I zoomed in on the bad ones after subtracting out the master dark and saw a number of zeros that were not in the good, calibrated frames.  I then opened image calibration, put in a bad frame and added a pedestal of 100, ran it and compared it to the same frame before adding the pedestal.  The zeros were now a positive integer (e.g., k = .000 to like K=.0008) after calibration.  So, I went back to WBPP and re-ran the data using a pedestal of 100 but I still got a few rejectio0ns as bad frames.  I ran it twice more, once with a pedestal of 200 and once with a pedestal of 500.....same result every time. That is, 6 rejections out of 37 frames.  

BlockHead

I wonder... you mention that you have two different exposure times. 

Are you certain you are not subtracting a 600 dark from a 300 second light frame?

This would cause large negative numbers that a normal pedestal would not be able to deal with.

But... if I understand you, it seems you have your answer. Either your dark frames are not good... if they are not be subtracted correctly...or there is an issue with the pedestal which needs to be larger. Zeros in your data... too many of them anyway, will result in WBPP saying they are bad frames because the analysis of the noise and background levels give unusual values. 

-the Blockhead

Bruce Donzanti

To address your question of mixing darks with the wrong light frames, I did run WBPP 3 times: 1) only the 300s subs with the 300s darks; 2) only the 600s with the 600s lights and 3) adding all of the data in WBPP at the same time, both 300s & 600s data.  Under scenario 1, I get 6 rejections out of 37 as stated above.  Under scenario 2, I get no rejections out of the 19-600s frames.  Under scenario 3, I get 6 rejections from the 300s frames again but I now got 3 rejections from the 600s frames.     

Couple of other points to mention:

1- I have done many NB images over the past several years with WBPP and this problem is new.  I always added a 100 pedestal just as insurance on every NB image.  

2- I can make new darks but they are not that old and I have used them recently on a number of other objects with no issues.  

What still confuses me is that the 100 pedestal and master dark works on the frames when I use Image Calibration outside of WBPP but not when the pedestal is applied within WBPP. 

I guess I can make new darks and see what happens as they are easy enough to make.