What to do when SNRWeights are wrong?

[astrovienna]

I'm processing some data where some Ha and OIII frames have gradients (from moonlight) and are 45 degrees out of registration from the reference frame (because I shot them a year apart, with some hardware changes).  All of those frames have extremely high SNRWeights in Subframe Selector, but to me appear clearly to be no better than the rest.  How should I handle this for stacking?  I can just exclude them, but they're otherwise pretty decent frames.  I ran Local Normalization on them, but the LNorm output frames are still evaluated as having the highest SNRWeights.

Kevin

[RickS]

Hi Kevin,

You could use a FITS keyword to assign weights and tweak the values as you like using the SubframeSelector script or process.  It may be easiest to do separate runs with the old and new data with different weighting formulae.

Cheers,
Rick.