Is it just me or is there a problem in this version with Photometric calibration?
I have even tried it with the sample images found in the Brutes_180s.zip. I have tried on Linux & MacOS.
Here is what I get as a typical output.
14:50:52: There is no valid WCS information in the header. Let’s make a plate solving.
14:51:05: Catalog NOMAD size: 1117 objects
14:51:05: Findstar: processing for channel 1…
14:51:05: Saved focal length 388.25 and pixel size 3.76 as default values
14:51:05: Image center: alpha: 18h03m56s, delta: -23°29’57"
14:51:05: Normalizing on blue channel.
14:51:05: Photometry radii set to 8.9 for inner and 18.9 for outer
14:51:05: Applying aperture photometry to 655 stars.
14:51:05: 609 stars excluded from the calculation
14:51:05: Distribution of errors: 228 no error, 6 not in area, 75 inner radius too small, 528 pixel out of range
14:51:05: Found a solution for color calibration using 46 stars. Factors:
14:51:05: K0: 1.212 (deviation: 0.217)
14:51:05: K1: 0.889 (deviation: 0.111)
14:51:05: K2: 1.000 (deviation: 0.164)
14:51:05: The photometric color correction seems to have found an imprecise solution, consider correcting the image gradient first
14:51:05: After renormalization, the following coefficients are applied
14:51:05: White balance factors:
14:51:05: K0: 1.000
14:51:05: K1: 0.734
14:51:05: K2: 0.825
14:51:05: Background reference:
14:51:05: B0: +0.00000e+00
14:51:05: B1: +2.38557e+02
14:51:05: B2: +1.57228e+02
But what does it do to the image? Was it on a raw image, calibrated image, stack?
The log indicates that many stars were saturated and could not be used for photometric analysis, leaving only 46 to do the PCC process.
The fact that a star is considered out of range depends on the configured low and high thresholds. In 1.2 the low threshold should be negative to allow for noise, but depending on how settings were imported from 1.0, it may not be the case.
Its a RAW image before stacking from the tutorial downloaded from Index of /colmic/Traitement_SiriL/brutes
I just used Light_M20_180s_1x1_gain100_20200818-222530_-9.8C_0001.fit as a way to show the issue.
The min pixel value shows -1500 and max is 60000 under Preferences/Photometry.
If I do photometric calibration with older 1.0.6 it works fine and goes about applying photometry to the stars
Yes because old version do the (bad) job of taking some bad stars.
Now siril is able to better remove outliers.
I encounter the same problem. Photometric calibration started and the progress bar moved but then stopped at about 2/3 and stayed there. I thought the program hanged but it accepted other operations. But the progress bar just still stuck there.
I just am not sure if the calibration fails or what. If it fails, then why it stuck there without saying that it fails (like in 1.0.6). I accpet that you say you are now doing a better job now but a clearer indication is needed.
This is fixed in next beta.
And now it seems to reject most stars 
14:51:05: Applying aperture photometry to 655 stars.
14:51:05: 609 stars excluded from the calculation
If the stars are rejected this is because they are not good for photometry. We have decided it was better to reject bad stars instead of doing wrong photometry analysis.
However, you can extend the range of the photometry.
First check the stats of your images and see what is going on.
I am no expert at AP @lock042, hence tried using an image by Colmic in the tutorial.
Would be good to know what should be a pragmatic setting for it to select atleast a fair few stars. Findstar obviously seems to find a lot of them and then photometry says it rejects most.
11:10:27: Running command: findstar
11:10:27: Findstar: processing for channel 1…
11:10:29: Execution time: 2.04 s
11:10:29: Found 18711 Gaussian profile stars in image, channel #1 (FWHM 2.779785)
It wont let me upload the fits file hence uploading a jpeg version
You can share FITS files with WeTransfer. Jpeg will not help.
Its taken from here http://www.astrosurf.com/colmic/Traitement_SiriL/brutes/Brutes_180s.zip
For photometry you don’t have the stacking results? I think it is on it you tried. No?
Tried it with one of my own stacked image this time…and here is what I got:
12:26:45: Existing plate solve (WCS information) will be resused for image
12:26:45: Findstar: processing for channel 0…
12:26:45: Findstar: processing for channel 2…
12:26:45: Findstar: processing for channel 1…
12:26:45: Image has a field of view of 1.41 degrees, using a limit magnitude of 13.63
12:26:46: The NOMAD catalog has been successfully downloaded.
12:26:46: Normalizing on red channel.
12:26:46: Photometry radii set to 14.8 for inner and 24.8 for outer
12:26:46: Applying aperture photometry to 283 stars.
12:26:47: 114 stars excluded from the calculation
12:26:47: Distribution of errors: 577 no error, 66 not in area, 30 inner radius too small, 1 invalid measurement error, 11 pixel out of range
12:26:47: Found a solution for color calibration using 169 stars. Factors:
12:26:47: K0: 1.000 (deviation: 0.030)
12:26:47: K1: 0.628 (deviation: 0.092)
12:26:47: K2: 0.551 (deviation: 0.155)
12:26:47: The photometric color correction seems to have found an imprecise solution, consider correcting the image gradient first
12:26:47: After renormalization, the following coefficients are applied
12:26:47: White balance factors:
12:26:47: K0: 1.201
12:26:47: K1: 0.754
12:26:47: K2: 0.661
12:26:47: Background reference:
12:26:47: B0: -1.31697e+04
12:26:47: B1: -8.06828e+03
12:26:47: B2: -7.00772e+03
Does this mean its applied to 169 stars? The comment “The photometric color correction seems to have found an imprecise solution, consider correcting the image gradient first” gives me a doubt as to what its done
I also noticed something strange with the output of Search objects. Not sure if its because the photometric calibration is NOT flipping the image or its something else.
I need the stacked image to check. Without I can’t say something.
Here is the file which has been cropped and background extracted. Saved as TIFF as it wont let me upload FITS. I dont know what or how to use Wetransfer.
NGC891_RGB.tif (42.6 MB)
For me there are no issues with PCC.
He does the job and just warns you that the result could be not very accurate.
About the astrometry you show. This is done before Siril. So I don’t know. Because After astrometry in Siril, the orientation is ok.
Sorry for dropping in. Need to understand your statement.
When the program says how many starts detected and then a number of them are rejected for PCC and the progress bar kind of “stucked”, the PCC DOES apply already, it just warns us that the result is not accurate. And we can proceed to the next step is we are OK with this inaccuracy. The “stucked” bar will be addressed in next release.
Have I understood it right?
Indeed. ANd yes, the only bug here is the “stucked” bar. But of course, it is really annoying and will be fixed, for sure, in the next release.
Do you mean after I do a photometric calbration? If so it doesnt flip my image at all! I am using the Mac version.