Crescent Nebula NGC6888 & First Light (ish) with new AM5 mount

[Ouroboros]

Not my best effort, but it is my first (ish) light with my new AM5 mount.  Taken with ASI2600 on my SW Evostar ED80. 

This was the first proper imaging run with this mount. I’d done a few test runs with it previously. I’m happy with the guiding. It’s a very nice little mount …. and a joy to carry outside and set up. I’ve had a few issues with my new ASIair and getting the mount and Air to play nicely, but I won in the end. For now! We all know that when the gremlins go away that they’re only hiding temporarily!  

Anyway, 41 of 60 x 120s subs were good. Cloud clobbered the rest.  Processed in Pixinsight.  Cropped, DBE, SPCC, , BlurXTerminator, NoiseXTerminator. Used StarXterminator to split stars from nebular to process each separately.  Applied General Hyperbolic Stretch to both. Tweaked with curves etc. Recombined with Pixelmath.

I’ve tried to reduce the star content of the image but it tends to look a bit insipid if I do that. I don’t think there’s enough nebular data to stand on its own.

 

[Ouroboros]

I have been wondering why this image looks so poor given the reasonable appearance of the separate images: starless nebula and stars. 

I have recombined them using the formula ~{(~starless)*(~stars)}     I read this is supposed to give a better rendition of colour compared with a simple addition of images in Pixelmath.   Any thoughts on this? 

Here are screenshots of separate images:

1. starless

2. Stars

3. Zoom in on stars.  Quite nice I think.

Edited August 25, 2023 by Ouroboros

[windjammer]

Stars are too prominent in the final mix?  The xstars and starless look good as you say (xstars a bit noisy?) I don't use PI except for the basics, so would combine xstars and starmask in Affinity Photo - the blend mode of the layers would be 'screen' or 'add', with live adjustment layers (levels, brightness/contrast etc) on the star layer to turn them down a bit - judgement by eye!  Not sure what the pixel maths for that would be!  

Simon

[Ouroboros]

15 hours ago, windjammer said:

Stars are too prominent in the final mix?  The xstars and starless look good as you say (xstars a bit noisy?) I don't use PI except for the basics, so would combine xstars and starmask in Affinity Photo - the blend mode of the layers would be 'screen' or 'add', with live adjustment layers (levels, brightness/contrast etc) on the star layer to turn them down a bit - judgement by eye!  Not sure what the pixel maths for that would be!  

Simon

Thanks for your comments, windjammer.  Yes, I think maybe I’ve over stretched the nebula given the quality of the data.   I should probably back that stretch off a tad. I have also been experimenting with adding or stretching the stars in different ways. Simply adding fainter stars (xstars + 0.2*stars for example) doesn’t look right. What looks better is stretching the stars to keep the main stars bright whilst suppressing the carpet of fainter stars.  The question is how to do that successfully? So far I’ve tried using curves with some success. Unfortunately, suppressing the faint stars also truncates the halos of the bright stars which gives them a slightly odd look.

[windjammer]

I had a play with the image and star mask - hope thats OK.  A bit of Topaz Denoise on the xstars.  Star layer is levels and brightness and contrast - the levels is a bit black clipping and gamma control: don't know what the pixel maths is to emulate the gamma control. The brightness and contrast should be easy to pixel math it.  2 versions here - more or les stars!

 

 

 

 

 

[windjammer]

ps:  star field mismatched with nebula: I took your xstars nebula but redid the star mask from your first image: but not the same image dimensions!  Anyway, ideas to work with!

[Ouroboros]

20 hours ago, windjammer said:

ps:  star field mismatched with nebula: I took your xstars nebula but redid the star mask from your first image: but not the same image dimensions!  Anyway, ideas to work with!

Sorry for delay in responding.  Busy yesterday.  Thanks for having a go at my image. Yes, the  images posted were screen shots and weren’t matched in size. You’ve done a good job in showing what’s possible, and certainly much closer to what I’m aiming at. I just have to work out how to do that in Pixinsight. Essentially it looks like you’ve done something like I was referring to earlier - retaining the brighter stars whilst suppressing the carpet of fainter ones. I’ll have another go later. 

[peter shah]

Lovely renditions both....good detail and colour....nice work

[Ouroboros]

Picking up on some of your comments, @windjammer, I’ve had another go. I went right back to an earlier image just after stretching to non-linear.  I applied some TGVDenoise to the starless image. Then used curves and colour saturation followed by a bit of local histogram equalisation.  I combined the image with the stretched star image in pixelmath. Finally I used Bill Branshan’s star reduction script to dial back the stars. I could even apply a few more iterations of that I think. Anyway, I’m happier with the result. Cheers.