Hello, I had some questions about Astrometry (specifically the web version) determines the orientation of the frame it solves. I am working on a code to display a star field based on the user input of a DCM to reach the desired point in the sky. The result of the code is being ran through Astrometry along with a real photo and while both agree on the center of the image, the orientation of one is flipped across the East-West axis with respect to the other. Namely, the real image shows an orientation of 160° E of N while the result of the code is shown to be at 20° E of N. What makes this more puzzling is that upon visual inspection, the reference centroids of the stars seem to line up in both images.
I have attached both results here for further comparison and analysis and appreciate any insights anyone has into this issue.