Sigma + sum

[Morten Lerager]

Hi' I had read a lot of preprocessing with
Sigma + Sum.
But I can't find the function In the manual.
So how to do a sigma + sum?

[Iver]

In preprocessing, see attachment

[Morten Lerager]

I cant find it in the manual.
But then it must be Sigma for rejection and Sum for SNR

Thank You Iver

[Iver]

I thought you you couldn't find where to apply Sigma+sum. This is the only explanation I could find in the manual. In case you didn't see it.

"Sum, Average, Sigma or Median. Select how to build the final image and the master dark frame and flat field. "Sum" and "Average" are best methods in terms of S/N ratio, "Sum" should be preferred on 8 bits images. "Sigma" is the best method to reject hot pixels and other noise, if you don't correct the dark frame, but it's slower and it's not recommended for astrometry and photometry. "Median" is rarely used on images since it has a lower S/N ratio. "None" is useful when the option "Save each image" is enabled (see next page)."

[Morten Lerager]

Thanks again Iver
Yes that was what I found in the manual.
Nothing about sigma+sum.
I am using dithering so my gees is the best is sigma. Remove hot pix and trails.
Sum as I think about it should increase SNR.
But it says for 8bit files. I think my cr2 files are 16bit. My result with sigma + sum are good.
But is it better to drop sum. Yes I know, when I don't find it in the manual I have to make some test.
Regards Morten Lerager

[fabdev]

Hi, Average and Sum are almost the same thing, the only difference is the division by N at the end of the process.
This is true also for Sigma (which is just an average with rejection of bad values) and Sigma+Sum

Division by N :

Advantages: it keeps the dynamic of the original images, so the file size is smaller. The dynamic does not change if some images are removed. Small numbers are also easier to understand.

Disadvantages: it can cause quantization when stacking many many images: I would not use Average (so I would use Sum instead) if I need to stack more than 500 images with 16 bit range, or more than 100 images with 12 bit range, or more than 20 images with 8 bit range.

[Morten Lerager]

Thank you Fabio
I guess after a week I'll able to get it to work.
With Ditering I will use Sigma + sum
Found RGGB is the one to use with Canon 2000D
I had a try with Flats and Bias frame used in Darks place, that ditten work.
So I only use Lights 30x180 sec.
My set up give me 1" / pix so i tryed to bin 2x2 ...not good
But 1x1 worked fine and this came out.
No so bad after a week, as they say over in Jutland.
And no other processing tool than AA8

[Dunmunro]

Thank you Fabio
I guess after a week I'll able to get it to work.
With Ditering I will use Sigma + sum
Found RGGB is the one to use with Canon 2000D
I had a try with Flats and Bias frame used in Darks place, that ditten work.
So I only use Lights 30x180 sec.
My set up give me 1" / pix so i tryed to bin 2x2 ...not good
But 1x1 worked fine and this came out.
No so bad after a week, as they say over in Jutland.
And no other processing tool than AA8

M31-1x1.jpg

looks good!