pixel rejection levels when stacking

Software Siril
21

Tried same stacking with same files from a debian live where I compiled from source the latest siril version.
Result with 117 files:
12:53:40: Pixel rejection in channel #0: 0.040% - 5.473%
12:53:40: Pixel rejection in channel #1: 0.017% - 4.013%
12:53:40: Pixel rejection in channel #2: 0.039% - 5.104%
12:53:40: Rejection stacking complete. 117 images have been stacked.

Result with 292 files:
tarting stacking…
13:36:58: Pixel rejection in channel #0: 0.034% - 6.214%
13:36:58: Pixel rejection in channel #1: 0.019% - 5.177%
13:36:58: Pixel rejection in channel #2: 0.036% - 5.887%
13:36:58: Rejection stacking complete. 292 images have been stacked.
13:36:58: Integration of 292 images:
13:36:58: Pixel combination … average
13:36:58: Normalization … additive + scaling
13:36:58: Pixel rejection … Winsorized sigma clipping
13:36:58: Rejection parameters … low=4.000 high=3.000
13:36:58: Background noise value (channel: #0): 0.119 (1.809e-06)
13:36:58: Background noise value (channel: #1): 0.104 (1.580e-06)
13:36:58: Background noise value (channel: #2): 0.111 (1.697e-06)

So indeed it has something to do with the windows version of siril. I tried on windows 10 the same amount of files and beginning at around 100 lights rejection values turned negative. Do you still would need that I upload the files? they are about 4.5 gb and would indeed take a while, about 14 hours :confused:

22

Thanks for the test, that’s useful. No don’t worry for the files then, it’s too much. We’ll try to find the problem with other files.

23

@nightspace : please, try this version and give me the log:
https://gitlab.com/free-astro/siril/-/jobs/814539090/artifacts/download

24

16:48:49: debug: nb_tot=7068855136.000000 and irej[channel][0]=2029235 and irej[channel][1]=332744139
16:48:49: Pixel rejection in channel #0: 0.029% - 4.707%
16:48:49: debug: nb_tot=7068855136.000000 and irej[channel][0]=1097931 and irej[channel][1]=287581987
16:48:49: Pixel rejection in channel #1: 0.016% - 4.068%
16:48:49: debug: nb_tot=7068855136.000000 and irej[channel][0]=2095879 and irej[channel][1]=319979694
16:48:49: Pixel rejection in channel #2: 0.030% - 4.527%
16:48:50: Rejection stacking complete. 292 images have been stacked.
16:48:50: Integration of 292 images:
16:48:50: Pixel combination … average
16:48:50: Normalization … additive + scaling
16:48:50: Pixel rejection … Winsorized sigma clipping
16:48:50: Rejection parameters … low=4.000 high=3.300
16:48:51: Background noise value (channel: #0): 0.109 (1.661e-06)
16:48:51: Background noise value (channel: #1): 0.099 (1.504e-06)
16:48:51: Background noise value (channel: #2): 0.105 (1.602e-06)

Looks good! Thanks!