I have encountered an error with Siril and Starnet. I start Siri with the.bat file first, as instructed in the release note. Siri crashed after 1-2 minutes of treatment, and I can see briefly in the console: “Cannot open the TIFF file” (or something similar).
Also, while Starnet is running, I have a prompt command opened with this:
“I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: AVX2
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.”
Hello, it means your processor does not appear to support the required instruction set (AVX2) to run Starnet. It is probably too old or incompatible with this program.
Thank you for your response.
I still think it is weird because my computer is two years old and I can run Starnet outside of Siril with no issues at all.
Hi,yes this is “normal” to have this window opening (this is a side effect of the workaround). Its output should not be any different than when you run Starnet outside of Siril. So all the stuff about AVX2 is info about how Starnet runs on your computer, independantly of Siril.
Now, to get more info on the crash (it should not crash…), could you pls do the following:
Edit the bat file (any text editor would do) that we have published and instead of >nul, replace with >log.txt, save, close and launch siril with this modified bat.
Run starnet on the image which caused the crash. Once it has crashed, there will be a log.txt file on your desktop. Just send it here and I will have a look.
I have the exact same problem with the AVX2 instruction set.
I use it on a gaming laptop which is only 1 year old; so that can’t be the problem.
I think this is a starnet problem; siril does the star removal but gives this error without crashing.
I never used the cmd of starnet only the gui so i can’t tell if this always occurs.
The log file on the desktop doesn’t mention this error.
Here is the error; hope this helps!
Btw, been giving the beta a testrun and i must see; great work!
I have the same problems. With modified bat it runs for a while before it crashes.
log: StarNet: running. Please wait…
log: StarNet: Pixel values exceed 1.0. Rescaling to avoid clipping peaks.
log: StarNet: linear mode. Applying Midtone Transfer Function (MTF) pre-stretch to image.
log: Applying MTF with values 0.028223, 0.009455, 1.000000
log: Saving TIFF: 16-bit file I:\0 Astro\working~directory\starnet_result.tif, 3 layer(s), 2781x2802 pixels
log: Could not open the TIFF file I:\0 Astro\working~directory\starless_result.tif
log: Error: unable to read starless image from TIFF…
log: Execution time: 11.50 s
Error, signal 11:
e[1me[31mPlease report this bug to: https://gitlab.com/free-astro/siril/issuese[0me[0m
There also seems to be a problem with the plate solving. the Image Information solver just freezes my PC completely.
log: Welcome to siril v1.2.0-beta1
log: Supported file types: BMP images, PIC images (IRIS), PGM and PPM binary images, RAW images, FITS-CFA images, Films, SER sequences, TIFF images, JPG images, PNG images, HEIF images.
log: Setting CWD (Current Working Directory) to 'I:\0 Astro\working~directory'
log: Parallel processing enabled: using 8 logical processors.
Successfully loaded 'I:\Siril\share\siril\siril.css'
Successfully loaded 'I:\Siril\share\siril\siril3.glade'
log: Searching scripts in: "I:\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_withbackgroundextraction
log: Loading script: OSC_Preprocessing_WithDrizzle
log: Loading script: OSC_Preprocessing_WithoutDark
log: Loading script: OSC_Preprocessing_WithoutDBF
log: Loading script: OSC_Preprocessing_WithoutDBF (2)
log: Loading script: OSC_Preprocessing_WithoutFlat
log: Loading script: RGB_Composition
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
Last available version: 1.2.0-beta1
Purging previously saved reference frame data.
log: Reading FITS: file result.fit, 3 layer(s), 3008x3008 pixels, 32 bits
log: There is no valid WCS information in the header. Let's make a plate solving.
log: Using already downloaded star catalogue
log: Catalog NOMAD size: 536 objects
log: Findstar: processing for channel 1...
using min_scale 1.862648, max_scale 2.262648
using trans_order 1
No limits set on relative scales for matching.
using trans_order 1
using trans_order 1.
Initial trans structure:
sig=1.0876e+00 Nr=45 Nm=0 sx=0.0000e+00 sy=0.0000e+00
TRANS based on matches only :
sig=8.8666e-01 Nr=290 Nm=329 sx=9.0360e-01 sy=9.1095e-01
After tuning with recalc, num matches is 329
sig=8.8666e-01 Nr=290 Nm=329 sx=9.0360e-01 sy=9.1095e-01
TRANS based on recalculated matches is
sig=8.8666e-01 Nr=290 Nm=329 sx=7.7426e-01 sy=7.4662e-01
Transformation Matrix:
-1.68066 +1.20478 +741.99168
+1.20467 +1.68081 -4345.08377
+0.00000 +0.00000 +1.00000
log: Saved focal length 375.04 and pixel size 3.76 as default values
log: Image center: alpha: 02h32m44s, delta: +61°26'55"
log: Normalizing on red channel.
log: Photometry radii set to 11.7 for inner and 21.7 for outer
log: Applying aperture photometry to 350 stars.
log: 323 stars excluded from the calculation
log: Distribution of errors: 178 no error, 15 not in area, 191 inner radius too small, 117 pixel out of range
log: Found a solution for color calibration using 27 stars. Factors:
log: K0: 1.000 (deviation: 0.000)
log: K1: 0.958 (deviation: 0.045)
log: K2: 1.509 (deviation: 0.198)
log: The photometric color correction seems to have found an imprecise solution, consider correcting the image gradient first
log: After renormalization, the following coefficients are applied
log: White balance factors:
log: K0: 0.663
log: K1: 0.635
log: K2: 1.000
log: Background reference:
log: B0: -2.20212e-02
log: B1: -4.19082e-02
log: B2: +0.00000e+00
log: The AutoStretch display mode will use a 16 bit LUT
[1, 0, 122;
0, 1, 79;
0, 0, 1]
[1, 0, -122;
0, 1, -79;
0, 0, 1]
log: There are not enough background samples. The background to be extracted cannot be computed.
log: There are not enough background samples. The background to be extracted cannot be computed.
log: Background extraction from red channel.
log: Background extraction from green channel.
log: Background extraction from blue channel.
log: Background with polynomial interpolation computed.
log: Execution time: 627.70 ms
log: StarNet: running. Please wait...
log: StarNet: Pixel values exceed 1.0. Rescaling to avoid clipping peaks.
log: StarNet: linear mode. Applying Midtone Transfer Function (MTF) pre-stretch to image.
log: Applying MTF with values 0.028223, 0.009455, 1.000000
log: Saving TIFF: 16-bit file I:\0 Astro\working~directory\starnet_result.tif, 3 layer(s), 2781x2802 pixels
log: Could not open the TIFF file I:\0 Astro\working~directory\starless_result.tif
log: Error: unable to read starless image from TIFF...
log: Execution time: 11.50 s
Error, signal 11:
e[1me[31mPlease report this bug to: https://gitlab.com/free-astro/siril/issuese[0me[0m
[#0]: in print_all_settings
[#1]: in kplotctx_border_init
[#2]: in _C_specific_handler
[#3]: in _chkstk
[#4]: in RtlFindCharInUnicodeString
[#5]: in KiUserExceptionDispatcher
[#6]: in end_and_call_remixer
[#7]: in gdk__private__
[#8]: in g_clear_list
[#9]: in g_main_context_check
[#10]: in g_main_context_iteration
[#11]: in g_application_run
[#12]: in main
[#13]: in main
[#14]: in main
[#15]: in BaseThreadInitThunk
[#16]: in RtlUserThreadStart
@Suttonrog ,
Ok, so the reason for the crash as such is the fact that it’s trying to open the recomposition window while starnet did not work. If you untick this option Recompose stars upon completion, the crash should not happen anymore.
No why starnet does not work is still a mystery… Tried testing on my PC, setting the working directory to a different drive, with spaces and ~ characters in the path and it completes without errors. Will keep looking…
