Rework of NGC 2070

[Atmos]

Skies are finally starting to clear up a bit in Melbourne but I don't have a camera arriving until Friday so no imaging for me :(
Plus, scope doesn't arrive for another week or so.

Looked back over my old data on NGC 2070, PI was telling me that I was getting ~ 1.75'/pix so I thought my might try drizzling and see how that went. The image itself is 12x400s RGB taken from my light polluted back yard so it's not overly deep and still pretty noisy.

Some of the brighter stars have a bulls eye, this is caused by the HDRTransformation, still trying to figure out how to eliminate it without causing a bunch of other issues. Masking the bright stars helps a bit.

I think I also took the red subs a bit too close to the horizon (compared to the G&B), brighter stars have red halos.

High Res Version (http://www.astrobin.com/full/242003/0/)

Criticisms most welcome :)

[Rick Parrott]

:thumbsup:A beautiful image!

[SpaceNoob]

Great image Colin, I can offer some advice for dealing with the star core issues. It will most likely be the brighter stars that suffer from this so try to mask them before applying HDR. Star mask would be suitable, or a range mask - you should only need to target the larger / brighter stars. Perhaps a bit of trial and error.

[Atmos]

Thanks :)



At the moment I am running a standard range mask which does help a lot (they look like eyes without it!) but I think I may have to play with Pixel Math and add a couple of masks together; the range mask and a bright star mask. I have a range mask to help protect the dimmer areas so that it doesn't look quite so flat.

Still so much to learn!

[RickS]

Looks good, Colin. I like the colours.

For a mask to use with HDRMT and LHE I start with a clipped luminance mask then I remove the stars using a PixelMath expression like "iif(star_mask>0.1,0,$T)". You can adjust the 0.1 to make the holes where the stars were bigger or smaller. Then smooth the mask a little with Convolution or removal of a couple of wavelet layers and voila!

Cheers,
Rick.

[Atmos]

Thanks Rick. What does the $T refer to in that expression?

I was just thinking of doing something as simple as adding a bright star mask to a range mask, would that increase the masking of the brighter stars or not make a whole lot of difference?

[RickS]

$T means the target image, Colin. Using $T means you don't have to edit the expression every time you apply it to a different image.

You could combine a range mask for the low signal areas with a bright star mask but I'd suggest you do it with the max() function rather than addition.

The reason I prefer a clipped luminance mask to a range mask for this purpose is that it varies the strength of the HDRMT or LHE effect depending on brightness. With a range mask it's all or nothing.

Cheers,
Rick.

[Atmos]

Oh that is a really good one to remember! I sometimes just change the names of windows to make them easy to put into the pixel math area :lol:

I guess I have been doing it the right way then, I have been using a clipped luminance mask as it is effectively a range mask :) Just easier type on the phone!

I think I'll have to do a bit of reading up on the Pixel Math routine, rarely do anything more than adding things together. Never used the LHE routine, just started a Google search on it now to figure out what it is.

[multiweb]

That highres looks terrific. Beautiful starfield. Well done. :thumbsup: