Open adrn opened 11 years ago
Actually, let's keep this an enhancement. I realized that it complexifies the stacking. So we should just do this at the end to say "even if the data aren't pixel-aligned, modeling works, whereas either stacking gets worse or else involves some interpolation (which is scary)"
Add a flag in the code to allow the true centroid of the star to wiggle around a little bit between coadds.