PixInsight Forum (historical)

PixInsight => Wish List => Topic started by: Andres.Pozo on 2009 May 16 06:19:39

Title: Inplace star alignment
Post by: Andres.Pozo on 2009 May 16 06:19:39
Hi,

it would be very useful for me if the result of an Star Alignment process could replace the source view. Also (probably) this could allow to add an star alignment process to a process container.

My objective is building a pre-processing process with this steps:


Title: Re: Inplace star alignment
Post by: Nocturnal on 2009 May 16 08:00:20
Hi,

I think DBE works best on high SNR images so it makes more sense to first combine your subs and then do a single DBE. At low SNR DBE will have a hard time distinguishing between nebulosity and gradients so it'll probably remove signal you'd really want to keep.

That said, it would be nice if the register process could modify the image in place. In fact I should add that to my debayer module.
Title: Re: Inplace star alignment
Post by: Andres.Pozo on 2009 May 16 09:09:47
Quote from: Nocturnal on 2009 May 16 08:00:20
Hi,

I think DBE works best on high SNR images so it makes more sense to first combine your subs and then do a single DBE. At low SNR DBE will have a hard time distinguishing between nebulosity and gradients so it'll probably remove signal you'd really want to keep.

The stacking with sigma-clipping works better if the images are normalized. My images usually have changing gradients and if I found an automatized method of eliminating them before the stacking the results would probably improve. I don't need a perfect flat background, only good enough.
Title: Re: Inplace star alignment
Post by: Nocturnal on 2009 May 16 09:24:42
Well that's an interesting point. Have you tried this by hand to see if the theory held up in practice?