First Fall EAA

A couple of weeks ago, I enhanced my syncstella.js script by adding a new find-rotation-center feature. The process works by capturing and plate solving three images in rapid succession, with the mount platform being rotated between each shot using its motor. By analyzing the plate solving results from these three positions, the script calculates the center of rotation and its relationship to the celestial pole.

I finally tested the feature today, but it proved more challenging than expected. The platform motor only moves quickly in one direction (for resetting during normal use), while the return movement relies on the platform's normal tracking speed—making the process quite slow. Manual repositioning was an alternative, but lifting the platform risks tipping the telescope. Additionally, the calculated values appeared inaccurate. After several attempts without satisfactory results, I gave up and just left the platform roughly aligned.

I also added an eval-focus feature to syncstella.js by utilizing PSF (Point Spread Function) fitting to measure focus quality directly from captured frames. I tried it a bit but it was more cumbersome to use that simply doing the focusing manually.

I also enhanced stackfit.js. I added a mosaic flag for multi-panel astronomical images and an annotate flag (enabled by default) to automatically generates annotated versions of processed images with coordinate overlays and celestial information leveraging plate solving.

The mosaic flag did not work as expected - I was trying to stitch and stack at once across all the subs. I ended up generating stacked images for each panel and doing the stitching in the Siril GUI.

The annotated images unfortunately were not very readable.

I captured calibration frames including darks, bias, and flats. I started with the L-Enhance filter on the initial nebula targets then switched to the L-Quad Enhance filter mid-session for the Iris Nebula and galaxy targets. As a result, I needed to capture a new set of flat frames since flats are filter-specific.

Observations: