Most users on windows and Mac os have no issue though.
This is why I don’t get it.
Hi, I’m using SiriL on macOS and windows. Recently I started to have problems with the PCC, but only when using NOMAD catalogue. If you change it to APASS one it works as expected.
One thing that I’ve discovered is that wen trying to use NOMAD catalog, s small number of stars was downloaded
log: Reading FITS: file r_pp_ngc7023_stacked.fits, 3 layer(s), 3008x3008 pixels
OMP: Info #270: omp_set_nested routine deprecated, please use omp_set_max_active_levels instead.
OMP: Info #270: omp_get_nested routine deprecated, please use omp_get_max_active_levels instead.
log: Existing plate solve (WCS information) will be resused for image
log: Findstar: processing for channel 0...
log: Findstar: processing for channel 2...
log: Findstar: processing for channel 1...
log: Image has a field of view of 0.00 degrees, using a limit magnitude of 18.00
log: Using already downloaded star catalogue
log: The NOMAD catalog has been successfully downloaded.
log: Normalizing on red channel.
log: Applying aperture photometry to 256 stars.
log: 256 stars excluded from the calculation
log: Distribution of errors: 512 no error, 256 not in area
log: No valid stars found.
log: Existing plate solve (WCS information) will be resused for image
log: Findstar: processing for channel 0...
log: Findstar: processing for channel 2...
log: Findstar: processing for channel 1...
log: Image has a field of view of 0.00 degrees, using a limit magnitude of 17.00
log: Using already downloaded star catalogue
log: The APASS catalog has been successfully downloaded.
log: Normalizing on red channel.
log: Applying aperture photometry to 3961 stars.
log: 3055 stars excluded from the calculation
log: Distribution of errors: 3875 no error, 99 Gaussian fit failed, 2245 not in area, 652 inner radius too small, 46 aperture too small, 13 pixel out of range
log: Found a solution for color calibration using 906 stars. Factors:
log: K0: 1.000 (deviation: 0.000)
log: K1: 0.868 (deviation: 0.079)
log: K2: 2.569 (deviation: 0.649)
log: The photometric color correction seem to have found an imprecise solution, consider correcting the image gradient first
log: Background reference:
log: B0: 2.38041e-03
log: B1: 2.97551e-03
log: B2: 3.79950e-03
Hope this help
Those whole number of stars that are powers of two looks suspicious to me (256 stars, 512 no error, 256 not in area)
Rafa, I’m quite puzzled by the log: Image has a field of view of 0.00 degrees line. Did you use the GUI or command to get there? Did you set pixel size and focal length correctly?
Glups! I’m afraid now 
Rafa
Any Windows clients reading this post… Are any of you using 1.0.3 that don’t have this problem with Photometric Color?
Hi,
Thanks for investigating and for the report. I’m running Windows as well. I don’t exactly understand which url with colons you are refering to.
The 2 urls we pass, in this order, are (for m42):
http://cdsweb.u-strasbg.fr/cgi-bin/nph-sesame/-oI/A?m42
and if we get no answer, we query simbad at:
http://simbad.u-strasbg.fr/simbad/sim-tap/sync?request=doQuery&lang=adql&format=TSV&query=SELECT basic.OID, ra, dec, main_id FROM basic JOIN ident ON ident.oidref = oid WHERE id ='m42';
There is no colon anywhere (see there: src/algos/astrometry_solver.c · 1.0.3 · FA / Siril · GitLab)… and both queries are successful (I’ve just tried).
So I don’t get how a colon could be the source of your error…
1 Like
yes, it used to work with 1.0.1, no longer with 1.0.3
But there is no change in the code in this area. Especially in the 1.0 branch since 1.0.1.
I can tell you that the majority of users can use astrometry with Siril on Windows.
So I can’t tell you why it doesn’t work. I would say it’s something like bitdefender or AVAST because 1.0.3 is recognized as a different software by the AV databases.
Thanks for the reply 
There is also no change in my system, besides of course whatever windows updates happened in the meanwhile and the Siril update.
To put it in context and hopefully find a solution, we are talking about an Windows 11 system with the default Windows defender security solution.
I enabled the ‘siril.exe’ and ‘siril-cli.exe’ as apps in the firewall and also created explicit inbound and outbound rules for them to be sure they will not be blocked.
The result did not changed, when I try to search for an object in Photometric calibration, I always see the ‘No catalog’ red error message in the Console.
I also disabled the Firewall and Defender completely and tried, with no change in the result or the error message…
Is there a log file somewhere that can give more details?
Also, what mechanism does Siril use to make the connection to the catalog? maybe there are also other binaries that have to be explicitly allowed through the firewall?!?
Oh. Give me an example of what you search for example
In one of the recent photos I made I search for the Trifid Nebula. I can see it exactly in the middle of the photo
But this is just an example… I can search whatever there, the only result I get with the last version is always ‘No catalog’. It also does not matter which catalog I choose or if the ‘Use Cache’ tick is checked or not.
Ok. This kind of name cannot be search.
Try M20.
Ok, this went further, I do not get the ‘No catalog’ error any longer (so maybe an idea is to change it as it was kind of misleading in this case…).
The plate solving still fails with ‘08:37:14: There are not enough stars picked in the image. At least 6 stars are needed.’.
But I guess this is just a thing of playing with the various options there. Is it better to do the calibration after or before stretching?
Always before.
Ok, thanks for the tip.
I was just reading the ‘no catalog’ error and thought that there is no point trying to search something else if there is no catalog available . Maybe you can consider changing it for something more relevant for this case
Is there somewhere a list of what names is better to use?
It is a very bad log indeed and it is already changed in the dev version to be more helpful.
I don’t think, but every name like M…, NGC…, IC…, HD… that are recognized by SIMBAD query.
1 Like