My Favorite Deconvolution Tutorial for PixInsight

[STEVE333]

The link below is to my favorite Deconvolution tutorial. I find it easy to follow and it produces good results. Rather than selecting the PSF stars manually I use the newish script PSFImage from Hartmut Bornemann to select them automatically.

https://www.youtube.com/watch?v=97OyeSR76Fs

If you have a favorite please feel free to provide a link for us to share.

Steve

[ram]

Thanks Steve! This is my only one: https://www.lightvortexastronomy.com/tu ... tails.html - but I also now have modified it to use the PSFImage script as you advised me instead of picking all those stars myself as per this tutorial (still was instructive in telling me what goes one under the hood).

I'm not a big fan of using videos to learn (algebraic thinker) so prefer written tutorials but I do have your YouTube video as one of the things to check out and try out, next target.

(I've just now preprocessed my Ha, O3, S2 master lights for the starless technique again on sh2-132 - just about to start running starless on it but I did run deconvolution on all of them using the LVA tutorial instructions.)

But I'm going to try to get out tonight to capture some more IC1805 if clouds don't roll in...

--Ram

[STEVE333]

Thanks for the Light Vortex Astronomy tutorial Ram. I read through and it is pretty similar with a few small differences. I would think good results would be obtained with either one.

Steve

[ram]

I agree - the settings you sent me for your deconvolution were very similar. The primary difference is in the NR choices (gaussian vs. poisson, and # of layers, and layer param) I tried it both ways BTW, yours and the LVA (which I had also modified to make my layer 2 identical to yours) and I couldn't tell the difference visually.

--Ram

[yobbo89]

Clicking on 50 stars is such a pain haha, I'll give the script a go.

[STEVE333]

Clicking on 50 stars is such a pain haha, I'll give the script a go.

Hi yobbo - I actually had to click on a lot more than 50 stars when I did it manually. There are requirements for a star to "qualify" and, with my setup, I only averaged about one star out of four or five that qualified (right "shape", right amplitude). Also, because I had to zoom-in to select the stars, I had to navigate around the image with the zoomed-in window to all portions of the image to spread the selection across the image. So, it took me quite a while to get the 50 good ones.

I really have appreciated the script. I just click on Evaluate and go away for a couple of minutes, and, it's done.

Cheers,

Steve

[STEVE333]

I agree - the settings you sent me for your deconvolution were very similar. The primary difference is in the NR choices (gaussian vs. poisson, and # of layers, and layer param) I tried it both ways BTW, yours and the LVA (which I had also modified to make my layer 2 identical to yours) and I couldn't tell the difference visually.

--Ram

Thanks for comparing the results and sharing with us. Good to know.

Steve

[Benjamin]

Another YouTube take on Deconvolution (and other things) from Sean at Visible Dark. Also deals with the PSFImage script. I prefer to clip the blacks a little when making the decon mask (so not quite the full histogram stretch as given by the STF).

[STEVE333]

Another YouTube take on Deconvolution (and other things) from Sean at Visible Dark. Also deals with the PSFImage script. I prefer to clip the blacks a little when making the decon mask (so not quite the full histogram stretch as given by the STF).

Thanks Benjamin I'll check it out.

Steve

[Benjamin]

The Visible Dark channel has lots of cool things. The DBE is also pretty cool and works!

[STEVE333]

Some good information, but, it almost put me to sleep. Sorry, just being honest.

However, I did appreciate his discussion of the Global dark and Global bright settings in the Deringing section of the Deconvolution tool.

Thanks for sharing.

Steve