SiriL 1.2.3/ 1.2.4 crashes upon registration lights

Hi all,
I’m having an issue with SiriL 1.2.3 and today 1.2.4 on MacOS 14.6.1 (23G93) (M3 Max, 36GB RAM, swap drive is a 8 TB SSD system. The data is on a SSD, too. Available storage on both SSDs is > 4 TB). Always as package download from the official site.

Data used:
205 lights from a Fuji GFX 100 II (102 MP per file)

What did I do:

  1. Script OSC_preprocessing → processing of biases, darks, flats works as anticipated and masters are made from them. Preprocessing of lights, too. But as soon as the registration process starts SiriL crashes. Therefore, I don’t have a log.
  2. Manuel processing to find the issue. It seems that only the part of the registration of the lights is affected.

This is the log I obtained from the terminal:

Siril is started as MacOS application
FFTW-Multithreading-Unterstützung wird initialisiert…
log: Wilkommen zu siril v1.2.4
log: Unterstützte Dateitypen: BMP Bilder, PIC Bilder (IRIS), Binäre PGM und PPM Bilder, RAW Bilder, FITS-CFA Bilder, Filme/Videos, SER Sequenzen, TIFF Bilder, JPG Bilder, PNG Bilder, HEIF Bilder.
log: Setting CWD (Current Working Directory) to ‘/Volumes/Data/Astro’
OMP: Info #276: omp_set_nested routine deprecated, please use omp_set_max_active_levels instead.
OMP: Info #276: omp_get_nested routine deprecated, please use omp_get_max_active_levels instead.
log: Parallel processing enabled: using 14 logical processors.
Successfully loaded ‘/Applications/Siril.app/Contents/MacOS/…/Resources/share/siril/siril.css’
Successfully loaded ‘/Applications/Siril.app/Contents/MacOS/…/Resources/share/siril/siril3.glade’
fetch_url(): https://siril.org/siril_versions.json
initializing CURL
log: Searching scripts in: “/Applications/Siril.app/Contents/MacOS/…/Resources/share/siril/scripts”…
log: Loading script: Mono_Preprocessing
log: Loading script: OSC_Extract_Ha
log: Loading script: OSC_Extract_HaOIII
log: Loading script: OSC_Preprocessing
log: Loading script: OSC_Preprocessing_WithDrizzle
log: Loading script: RGB_Composition
scripts: Error opening directory “/Users/bjoernlanglotz/Users/bjoernlanglotz/Library/Application Support/org.free-astro.Siril/scripts”: No such file or directory
scripts: Error opening directory “/Users/BKL24/.siril/scripts”: No such file or directory
scripts: Error opening directory “/Users/BKL24/siril/scripts”: No such file or directory
log: Loading registration method: 1-2-3 Stars Registration (deep-sky)
log: Loading registration method: Global Star Alignment (deep-sky)
log: Loading registration method: Two-Pass Global Star Alignment (deep-sky)
log: Loading registration method: Image Pattern Alignment (planetary - full disk)
log: Loading registration method: KOMBAT (planetary surfaces or full disk)
log: Loading registration method: Comet/Asteroid Registration
log: Loading registration method: Apply Existing Registration
log: Default FITS extension is set to .fit
Fetch succeeded with code 200 for URL https://siril.org/siril_versions.json
Last available version: 1.2.4
The URI points to a local file.
log: Setting CWD (Current Working Directory) to ‘/Volumes/Data/Astro/process’
log: Checking sequences in the directory: /Volumes/Data/Astro/process.

(siril:77303): Gtk-WARNING **: 09:36:56.999: no trigger event for menu popup
seqfile ‘dark_.seq’ already exists, not overwriting
seqfile ‘bias_.seq’ already exists, not overwriting
seqfile ‘pp_flat_.seq’ already exists, not overwriting
seqfile ‘pp_light_.seq’ already exists, not overwriting
seqfile ‘light_.seq’ already exists, not overwriting
seqfile ‘flat_.seq’ already exists, not overwriting
Reading sequence file bias_.seq'. Reading sequence file dark_.seq’.
Reading sequence file flat_.seq'. Reading sequence file light_.seq’.
Reading sequence file pp_flat_.seq'. Reading sequence file pp_light_.seq’.
Loaded 6 sequences
Reading sequence file `pp_light_.seq’.
Purging previously saved reference frame data.
log: Reading FITS: file pp_light_00001.fit, 3 layer(s), 11662x8752 pixels, 32 bits
bitpix for the sequence is set as -32
log: Sequence loaded: pp_light_ (1->205)
Purging previously saved reference frame data.
number of filtered-in images: 205
number of filtered-in images: 205
number of filtered-in images: 205
number of filtered-in images: 205
Saved the reference frame buffer for alignment preview.
log: Registration: processing using method: Global Star Alignment (deep-sky)
log: Interpolation clamping active
Memory per image: 1168 MB. Max memory: 25938 MB
log: Global star registration: with the current memory and thread limits, up to 6 thread(s) can be used
log: Reading FITS: file pp_light_00001.fit, 3 layer(s), 11662x8752 pixels, 32 bits
log: Reference Image:
log: Findstar: processing for channel 1…
0 Siril 0x0000000102455ca8 signal_handled + 168
1 libsystem_platform.dylib 0x0000000185522584 _sigtramp + 56
2 Siril 0x000000010254b2c0 _ZN12_GLOBAL__N_113gaussVerticalIfEEvPPT_S3_iid + 240
3 Siril 0x0000000102545730 .omp_outlined. + 48
4 libomp.dylib 0x0000000102ccdc7c __kmp_invoke_microtask + 156
5 ??? 0x000000000000000e 0x0 + 14
Error, signal 10:

Thanks,

Bjoern

It is difficult to say.
As you say, it is not a 1.2.4 specific issue. However, we never heard about crashes in registration from other users. So either it is specific to your system or to the images.

Please, load the pp_light_00001.fit image. Then, type findstar in the command line. If it crashes, share the image, please.

Thanks for the quick tipp! SiriL crashes. on using the findstar command. Can I send the fit image via retransfer? In that case I need a email address to send it to.

Thanks again,

Bjoern

You can send it via WeTransfer. It will produce a direct link.

OK, I’ll check later today and come back with the link.
I just check SiriL with a file (converted to fit) from a 50 MP Sony a1: With this file the findstar command works. With the 102 MP GFX 100 II files it crashes.

The link is:

There’re two files: pp_light_00001.fit and the corresponding raw-file from the camera in case you’d like to check with it what’s going on.

Thanks again!

Bjoern

I tried your images. No crash. But your fit image is very strange. I don’t understand how you got something like that :

If I debayer your RAF image I get:

Thanks,
I just found out that I accidentally used flats from a different lens combination to create the pp_light_00001.fit.
Do you have any tipp for me what I can do to get my installation running? Deleting / reinstalling resources?
It’s running on MacOS 14.6.1 (23G93), processor is a M3 Max

A short update: converting the RAF files to dng-files doesn’t help.
Applying the findstar command to the RAF-file itself leads to a crash, too.
Is there a possibility for a memory issue or something specific for MacOS?

On more thing I found out:
I loaded a tiff-file into SiriL and “tried” to use the editing functions. I started with a background extraction to remove a gradient. SiriL immediately crashed. It first changed when I reduced the file size of the original 102 MP. Though, it seems that one “option for the problem” with my configuration might be in the memory handling of SiriL. I have no idea, whether this helps. Obviously with other machines it’s working. It was a surprise to learn that a quite powerful machine like a M3Max equipped with plenty of very fast memory runs into such an issue.

Nevertheless, is there any suggestion how one could work around this issue?

Hello. We are working on it. Yes it looks like a memory issue, probably specific to macOS because I really can’t spot any error on my Linux machine.

If you need any help i.e. with testing, don’t hesitate to get in touch. I’m happy to support!

1 Like

You could test the master version, to see if the bug always occurs:
Intel version: https://gitlab.com/free-astro/siril/-/jobs/7779758359/artifacts/download
Apple Silicon version: https://gitlab.com/free-astro/siril/-/jobs/7779758357/artifacts/download

@BKL24 : any update?

Unfortunately,
it’s not working at all. Upon starting MacOS brings a popup stating that the program is broken.

The line you have to write in the command line is written in the dmg

Sorry,
haven’t seen it. It’s working! At least I could load the 102 MP GFX 100 II raw file and use the findstar command. I’ll check tomorrow in more “depth”.

Thanks very much!!!

Can’t reproduce the background extraction crash. Could you,

Hi,
it’s working. Great thanks very much!
All processing works including registration, stacking, background extraction etc. At least I can stack now all the images with this camera. Once this version is final and all functions are activated (some are grey, i.e. phometric color calibration) I finalize the images.

All function are available.
The behavior of dev version is quite different from 1.2 for color calibration. You need to plate solve first.

1 Like