Increase the polynomial order of gradient removal.

Got an idea for something that SharpCap should do? Share it here.
Forum rules
'+1' posts are welcome in this area of the forums to indicate your support for a particular feature suggestion. Suggestions that get the most +1's will be seriously considered for inclusion in future versions of SharpCap.
Borodog
Posts: 330
Joined: Fri Jan 01, 2021 7:25 pm

Increase the polynomial order of gradient removal.

#1

Post by Borodog »

First of all, gradient removal and hot pixel rejection have MASSIVELY increased my enjoyment of using SharpCap for live stacking. The two together have been game changing. Can't thank you enough for implementing these, Dr. Glover. Now, on to the thing.

If there's one thing I dislike almost as much as taking darks, it's taking flats. Hence, I would like to request that the polynomial order of the gradient removal be increased from linear to at least quadratic, if not fourth order, which I believe is as high as most astro post-processing software goes. SiriL, APP, StarTools, PixInsight; I believe all will do up to 4th order background extraction. Obviously this will not address dust shadows, it may not always be appropriate, it may not be perfect, you may have to increase the density of sample points for higher order fits, you may have to reject sample points that fall on nebulosity, etc. Still, I feel like it would be a massively useful feature.

M51 live stacked in SharpCap with HPR and gradient removal but no flats:

Image

The exact same image with only 4th order background extraction in SiriL (note that this automatically neutralizes the background as well):

Image

Thanks for your time and consideration.
monsonmjm@gmail.com
Posts: 1
Joined: Thu Nov 11, 2021 5:12 pm

Re: Increase the polynomial order of gradient removal.

#2

Post by monsonmjm@gmail.com »

I was happy to buy a Pro Lifetime Membership to Sharpcap. It is not only excellent, but wonderfully supported. And, anything that makes Darks or Flats easier would be greatly appreciated. Not just +1, but +3 because each New Feature has been hugely beneficial for my hobby's enjoyment!
rwsmythe
Posts: 10
Joined: Sat Jan 15, 2022 10:40 pm

Re: Increase the polynomial order of gradient removal.

#3

Post by rwsmythe »

+1
User avatar
admin
Site Admin
Posts: 13173
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Increase the polynomial order of gradient removal.

#4

Post by admin »

Hi,

thanks for the suggestion - if people want to share a few frames that would benefit from higher order background subtraction then that would help me be in a position to investigate this.

One thing will be to see how numerically stable the calculations are with a higher order polynomial. Least absolute deviation (which we use to avoid being affected by outliers) needs an iterative solution, which is not always stable. For linear removal there are only 3 unknowns to be calculated. Quadradic has 6 and quartic has 15 if I have counted correctly.

cheers,

Robin
Borodog
Posts: 330
Joined: Fri Jan 01, 2021 7:25 pm

Re: Increase the polynomial order of gradient removal.

#5

Post by Borodog »

Thank you very much for looking at this, Dr. Glover. Here are a selection of lights.

< links removed because they were in error >

https://drive.google.com/file/d/1WnrZrd ... sp=sharing

I've included lights from a couple of M101s, different nights and sky conditions, an M27, and an M51. All shot with a reduced C8 with the ASI533MC. I have some lights from my 70mm refractor, but honestly the sensor is so well illuminated there that there is practically no vignetting.

Edit: My apologies but I realized that 3 of those files were inadvertently from the end of the sessions and not the beginning, meaning they were not usable lights (dawn's early light or tracking limit reached). I will replace them shortly.
Last edited by Borodog on Mon Jun 27, 2022 11:29 am, edited 1 time in total.
paul07834
Posts: 13
Joined: Mon Feb 27, 2017 1:11 am

Re: Increase the polynomial order of gradient removal.

#6

Post by paul07834 »

+1 for requested feature (if doable)

Paul
User avatar
Tiago Ferreira
Posts: 143
Joined: Sun Mar 21, 2021 10:50 pm

Re: Increase the polynomial order of gradient removal.

#7

Post by Tiago Ferreira »

What an excellent idea. I never take flats and would be so useful this feature
Youtube channel: https://youtube.com/@DobsonianPower

12" f/5 Dobsonian GSO
8" f/6 Dobsonian GSO
6" f/5 Dobsonian Go-To Virtuoso Gti
Dwarf II Smart Telescope
Geoptik Eartha Eq Platform
ZWO ASI224MC + ZWO ASI294MC
kite55surfer
Posts: 8
Joined: Mon Sep 27, 2021 7:15 pm

Re: Increase the polynomial order of gradient removal.

#8

Post by kite55surfer »

+1 providing it doesn't slow the process much
Xio1996
Posts: 41
Joined: Fri Feb 21, 2020 10:15 am
Location: Isle of Wight, UK

Re: Increase the polynomial order of gradient removal.

#9

Post by Xio1996 »

I would like to +1 this feature.

The current Hot Pixel Removal and background subtraction features of SharpCap are excellent. This feature (if possible) would be a great addition.
Borodog
Posts: 330
Joined: Fri Jan 01, 2021 7:25 pm

Re: Increase the polynomial order of gradient removal.

#10

Post by Borodog »

Dr. Glover,

As I mentioned in my edit above, 3/4 of the files I linked above were incorrect. Here is a link to the correct files; all of them should be usable lights. My apologies for the inconvenience.

I've also added an M13, which I expect to be a tough case, and a Crescent Nebula, also perhaps tough (I tried SiriL on it and while it was greatly improved it definitely was not perfect).

https://drive.google.com/drive/folders/ ... sp=sharing
Post Reply