Large Vertical Band

Alan Karty

Attached is a master H light resulting from WBPP v.2.3.1.  As you will see, there is a large bright band right down the middle of the image and the DSO.  This band is less visible in the R master and even less so in all other light masters.

What is this and how can I fix it?  No other processing was done.

The band is slightly visible in the registered images but really blooms in the integrated master. 

Thank you.

BlockHead

I do not know what is causing that fine structured band!

It sure looks electronic- but the fact it shows up in Red (Ha... Red...etc) makes me suspicious of IR light leaks.

You perhaps could do some pixel substitution. But it would likely be more work than just solving the issue.

LPS (Linear Pattern Subtraction) might also be something to try.

-the Blockhead

Alan Karty

Well, I reran WBPP without LPS and the vertical lines all went away.  Any comment?

BlockHead

Oooh...

Sure... LPS could *cause* them. 

My comment? Don't use LPS! lol

Actually there is a way to optimize LPS if you want to dig deeper- but the version implemented in WBPP has fixed (usually good) parameters.

-the Blockhead

Alan Karty

OK.  So, without the benefit of LPS, how do I get rid of defective columns if I don't have access to the camera's defect map?

BlockHead

You can use LPS- but you will need to do it "manually" and not use the automated method. I suggest you read Vicent's description of this method and see if it interests you. 

Regarding a defect map- you can identify the bad columns and use pixel math to create a defect map. 

I think I demonstrate this. (I need to look back).

-the BLockhead