Astrometry_net - images failing to solve

[awesomecosmos]

Hi there,

I am writing code using Python's astroquery.astrometry_net package, and am attempting to use ast.solve_from_image from a test list of 4 FITS images, 2 with exposure length 60s, and 2 with exposure length 300s. I have found that all but one image fails to solve, and they each return the following error:

Traceback (most recent call last): File "C:\Users\ave41\OneDrive - University of Canterbury\Master's 2021\ASTR480 Research\ASTR480 Code\01 Data Reduction Pipeline\DataReductionPipeline\src\asp.py", line 93, in <module> test_wcs_header = ast.solve_from_image(another_test_img3,solve_timeout=1000, File "C:\Users\ave41\Anaconda3\lib\site-packages\astroquery\astrometry_net\core.py", line 406, in solve_from_image return self.solve_from_source_list(sources['xcentroid'], File "C:\Users\ave41\Anaconda3\lib\site-packages\astroquery\astrometry_net\core.py", line 305, in solve_from_source_list return self.monitor_submission(submission_id, File "C:\Users\ave41\Anaconda3\lib\site-packages\astroquery\astrometry_net\core.py", line 255, in monitor_submission raise TimeoutError('Solve timed out without success or failure', TimeoutError: ('Solve timed out without success or failure', 4857551)

For the image that does work, it is the first file in the directory, and it returns a WCS header as expected, and when I check on the astrometry.net page, it has solved successfully (http://nova.astrometry.net/status/4857455). However, the other three images return the error above, and when I check on astrometry.net at the time when the above error is returned, it says the job is queued, and after 5-10 minutes, fails to solve on the website too. Example of a failed solve: http://nova.astrometry.net/status/4857501

I have set a timeout of 1000 seconds in my code (https://github.com/awesomecosmos/Data-Reduction-Pipeline/blob/main/src/asp.py), and yet the images fail to solve within 3 minutes of running the program. I cannot find any more documentation on how to solve this error. Please can someone point me in the right direction?

(I had also submitted a Github issue regarding this, but they recommended I post here. https://github.com/astropy/astroquery/issues/2133)

Thank you!

[Bryan]

One possibility is that the images have too many stars.  60s and 300s are extremely long for plate solving.  I rarely use over 8s and this is with a DSLR at ISO 1600.  AT will work fine with less than 100 stars.

Bryan

[Bryan]

What indices are you using?

On Monday, August 9, 2021 at 3:45:06 PM UTC-6 awesomecosmos wrote:

[Paul Leyland]

Bryan <bcas...@gmail.com> commented

"One possibility is that the images have too many stars. 60s and 300s are extremely long for plate solving. I rarely use over 8s and this is with a DSLR at ISO 1600. AT will work fine with less than 100 stars."

I rather doubt this is the case. I often solve images with well over 2000 stars in them when working in the MW starfields around Sco & Sgr.

Failure to solve is generally cause by having too few stars in my experience. Setting the image scale incorrectly (-L, -H and -u), the plate centre (-3, -4) and search radius (-5) are by far the most common causes of failure.  Sometimes I set the scaling factor (-z) too high. Very occasionally I choose the parity (-8) incorrectly.

I recommend checking these in the order given above.

Paul.