Registration artefacts in binx2

Hello Adam.-

Every time I register images in binx2 (before normalizing and integrating) black artifacts appear around most of the stars. Each image features different artifact patterns.
I'm sure this problem produces bad results for me when integrating the images.
It is possible it is due to the registered algorithm but it is not clear to me. When I use WWBP I put "pixel interpolation" in auto mode, and by default it uses laczos-3 as the algorithm in my case.
I don't know if I should use other algorithms, or what consequences that might have.
That is why I ask for your advice.

I attach a jpg with three different images of a specific area where you can see how these artifacts are different in each shot.

image

Thank you so much

Comments

  • Yes indeed- Lanczos is the default because it minimizes artifacts and blurring for well sampled (and oversampled) images. BUT...for undersampled images it does produce exactly the artifacts you show. I see this in my own data in images from the astrograph I use. 

    Typically most (but not all) will go away when you integrate the images and all of this gets averaged together. 
    If you find it objectionable, you can try one of the bi-cubic interpolation algorithms. For undersampled images, it can be better.

    -the Blockhead
  • many thanks.
    I will try

Sign In or Register to comment.