Author Topic: What do you think - fixing flat field offset with pixelmath  (Read 527 times)

Offline bvalente

  • Member
  • *
  • Posts: 55
    • View Profile
hi folks

i sometimes run into this issue of "3D dust donut halos".

I believe they happen when my flats have somehow become offset to my lights .Assuming that's true, i'm thinking of using pixelmath to create a new flat that is X pixels offset from the original. Through trial and error i should be able to figure out the correct offset

One thing I'm not sure of within pixelmath is how to reference a different pixel than the current one?

roughly i want to say: this pixel value should be the pixel value of one that is +10x, +1y offset kind of thing

looking for any feedback or suggestions?



here's an example (please ignore the poor collimation of my scope, i'm first light on this)

 

Offline pfile

  • PTeam Member
  • PixInsight Jedi Grand Master
  • ********
  • Posts: 4466
    • View Profile
Re: What do you think - fixing flat field offset with pixelmath
« Reply #1 on: 2019 September 06 12:30:55 »
i suppose the only problem with this idea is that it will shift the vignetting as well, so even if the dust motes are taken care of the flattening will be wrong in the corners.

the pixelmath functions x() and y() return the current x and y position being processed. the function pixel(target, x, y) should return the value of the pixel in the target image at x,y, so i think an expression like pixel($T,x()+10,y()+1) evaluated against the flat would do what you want, though i'm not sure what happens when you start reading past the end of the image (meaning at pixel xmax-5 you're asking for pixel xmax+5 which does not exist in the target image.) at best you'll have some black borders in the new flat which need to be overwritten with something...

i'd have to mess around with it to see.

rob

Offline bvalente

  • Member
  • *
  • Posts: 55
    • View Profile
Re: What do you think - fixing flat field offset with pixelmath
« Reply #2 on: 2019 September 06 12:38:29 »
thanks Rob

regarding start and end, i'll just have to test for that and either duplicate the last pixel row or do nothing.

i'll give it a shot and see what appens

thx

Brian