HDR Orion processing

[BlackNebula (Mark)]

Hey everyone, I’m needing some advice on how best to process multiple sets of data on Orion for a HDR effect and to hopefully process without a blown out core.

I understand I would stack and process each set separately with calibration frames in Siril and take into Photoshop. And then use layer masks in Photoshop and then blend in the core. What I am not sure of is the following:

I have captured 4 sets of data, 30 x10secs, 30 x 30 secs, 30 x 60 secs and 30 x 120 secs. All done on a OSC (533Mc pro) and shot on the same night, same session, no change to camera orientation. Was this overkill? Did I only need 2 sets of data, say 120 secs and 10 secs?

How do you align the separate images after stacking and processing?

How do you merge and blend the layers with 4 final images? Or should I just use the longest and shortest sets of data?

Any help or pointers would be greatly appreciated!
Thanks
Mark

[Retrograde (Pete)]

Hi Mark,
since no-one has chimed in I'll have a go at answering the best I can.
I've only done this a couple of times with Orion as I don't do much imaging this time of year and have never done a really deep image of M42.

It's hard to know whether the four sets of data re overkill or not as each situation varies. Even 10 sec exposures can be too long for the core of Orion given how bright it is, although this depends on how much you stretch the image of the core.
Start with two layers and see how it goes & you could always add in intermediate ones if you think you need them.

Aligning the separate images shouldn't be too hard as they should be quite close and the same image-scale. It can be done manually in Photoshop using the move tool, and either 'blinking' between the two layers (by turning off and on visibility of the top layer) or you can use difference blend mode on the top layer (or a bit of both). (You could always try "auto-align layers" under the edit drop down - who knows you might be lucky?)

The hard part is blending it all together so that it looks smooth and seamless. I find the "select and mask" function (from the Select drop down menu) works well allowing you to feather any selection visually, so there's not an abrupt step-change between layers. A fair bit of trial and error is usually required.

Good luck and feel free to post up your efforts.

[BlackNebula (Mark)]

Hey Pete, thanks. Some good advice, appreciate it! Work has been manic so haven’t had a chance to give it go yet.

I’ll Start with two layers as you say and take it from there.

I did find a you tube vid that talks to the aligning, will give it an go and follow your steps. Hopefully the blending works, will post an image once done.

This one could take a bit longer than usual

Cheers
Mark

[Cyberman (Rob)]

Mark , I had a go at this a little while ago. I used DSS, siril and gimp. I used 3 layers in GIMP:5,30 and 60 sec. I stretched the 60sec images to get more of the outer nebula. The inner areas are still a bit bright, but I can see the trapezium. My camera was an old cannon 600 D in terrible Brisbane skies. Gimp had functions for rotating and offsetting the layers to get allignment.

best of luck with the merging. it was time consuming but fun. Rob

[Craig_]

Possibly you have captured too many sets of data (or at least more than needed), without seeing it is hard to say. You could try stacking the 10s and 30s data together, and the 60s and 120s data together, leaving you with two integrations. Stretch the first set for the core, and the 2nd set for everything else.



Bring them into Photoshop, open as a stack. Place the higher exposed integration on top. You can then merge them a number of ways. The Orion core is fairly easy to just blend in by hand, so take a brush, set hardness to 0, opacity to 100 and mask the core in roughly. Then reduce the brush opacity to say 50%, and feather in the transition until it looks natural.



More advanced methods would be using techniques like luminosity masking, which are effective but may be more tricky depending on your level of Photoshop experience.