Author Topic: DBE rejection problem  (Read 180 times)

Offline libraryman2

  • Newcomer
  • Posts: 5
    • View Profile
DBE rejection problem
« on: 2019 October 17 03:33:06 »
Hi all, i need some advice on DBE and how to apply it in my image!
i have 3 stacks...HA, O111, S11 on the Cave nebula wide field..

i assume that i need to apply DBE on each stack before i combine them...
here is my issue, the HA is relitivly easy, i think the rejection map looks o.k. (to me)

The O111 noisy and i find it dificult to choose between various rejection maps!

i guess my question in part is: what makes up a good rejection map?

attached are a couple of examples..i realise that this is a subjective assessment but it would help me greatly if i could get a sense of what is bad in a rejection map...

thanks Ray

Offline libraryman2

  • Newcomer
  • Posts: 5
    • View Profile
Re: DBE rejection problem
« Reply #1 on: 2019 October 17 04:30:21 »
i have attached the O111 image that i am trying to DBE.....any recomendations would help..

Ray

Offline ngc1535

  • PixInsight Old Hand
  • ****
  • Posts: 304
    • View Profile
Re: DBE rejection problem
« Reply #2 on: 2019 October 17 09:27:08 »
Hi Ray,

It isn't a "rejection" map- instead it is a compensation model that is constructed from the samples that you choose.
I would argue that there isn't a way to look at a model in isolation and ask if the map looks "OK" or not. You need to compare the model to your image/data can see if the relationship and correlations make sense. If you clearly see a horizontal gradient in your image and the model shows a circular result- it is very likely *not* OK (or not good enough).

The key to DBE is choosing the samples and tolerances wisely. I do have a video that gives one approach to DBE- but it for a specific kind of typical flat error.
You might get something out of it see https://www.youtube.com/watch?v=f6IN9cXVh-4&feature=youtu.be (it is from working on a galaxy image).

Sometimes people think the model will do more than you can see. That is, if you can't see anything wrong with your data- still run DBE and it will do something magical and improve the image. Assuming this is not an eyesight limited operation- I do not think this is the way to approach DBE. Instead I think it is better to look closely at the data and see the need for applying DBE to correct things that are evidenced in the analysis of the image. This is especially true for fields of nebulosity that fill the field. If sky should otherwise be uniform- applying DBE doesn't hurt an already uniform image... but it also will not necessarily help it either.

-adam

Offline libraryman2

  • Newcomer
  • Posts: 5
    • View Profile
Re: DBE rejection problem
« Reply #3 on: 2019 October 18 04:36:00 »
Thank you Adam, i have already watched it, ive not had time to practise it yet..
its quite a departure from what i would expect, in terme of your settings, but hey..if it works     all's good.

Ray