M51 WIP, making the most of bad data

[Tzetze]

I got some time at the weekend to work on processing data captured on 30/03.

Only 6x subs each for LRGB, and unfortunately none of the L subs have tight stars. Will need to get more data on this when the weather allows.

So, I've ended up with stretched out stars and blurred details in the dust lanes but I think the colour balance and processing has worked out quite well. I used deconvolution and TGVDenoise for the first time in this process. The details in the luminance did appear sharper than in this final result, so maybe a reworking can bring about a sharper result after all.

Comparing against other images of M51, there should be a strong hint of blue in the yellow in the left of the image. I'm not sure if I've failed to catch it due to poor processing or not enough data. That's something else to strive for in the reworking and further data collection.

L: 6x900s at 1x1

RGB: 6x225s at 2x2

Edit: I should also point out that I've somehow damaged the centres of the brighter stars (inverted?) during the processing. I guess my star+background mask wasn't strong enough on the stars.

[Sunshine]

WOW is that ever a great image, smack in the face closeup too!

[paul mc c]

Unbeleavable.......thats amazing.

[Tzetze]

1 hour ago, Sunshine said:

WOW is that ever a great image, smack in the face closeup too!

 

59 minutes ago, paul mc c said:

Unbeleavable.......thats amazing.

Thanks very much, Sunshine and Paul.   =]

[AKB]

9 hours ago, Tzetze said:

I've somehow damaged the centres of the brighter stars (inverted?)

This seems to happen, for me, when smoothing/feathering star masks in PI.  Perhaps worth looking there.  Otherwise, amazing!

[carastro]

Very nice image and detail.

Carole 

[Knight of Clear Skies]

Looking very good, and could be a real cracker if you can add more data. What scope and camera did you use for this please?

[Tzetze]

Thank you all.

Sorry Knight, I should have included gear used in OP.

Astrodon LRGB, Atik 460ex, 10" Newt @ f4.7, NEQ6.

[Seanelly]

Friggin' awesome.

[wimvb]

On 09/04/2019 at 00:23, Tzetze said:

Edit: I should also point out that I've somehow damaged the centres of the brighter stars (inverted?) during the processing. I guess my star+background mask wasn't strong enough on the stars.

Very nice image. When you have the opportunity, replace the luminance with better data, the rgb data doesn't need the same amount of detail.

The star profiles (edge/inverted) are probably caused by too aggressive deconvolution. If you subtract a star mask from an L mask, you can protect the stars during this step.

[Knight of Clear Skies]

18 hours ago, Tzetze said:

Sorry Knight, I should have included gear used in OP.

Astrodon LRGB, Atik 460ex, 10" Newt @ f4.7, NEQ6.

Thanks. It seemed to be an awfully long focal length for a refractor but I was confused by the lack of diffraction spikes (did you process them out?).

[Tzetze]

On 10/04/2019 at 06:26, wimvb said:

Very nice image. When you have the opportunity, replace the luminance with better data, the rgb data doesn't need the same amount of detail.

The star profiles (edge/inverted) are probably caused by too aggressive deconvolution. If you subtract a star mask from an L mask, you can protect the stars during this step.

Thanks Wim. I was able to gather much more data and now have 5.5hrs luminance to process. It's looking much better so far but won't get a chance to work on it fully until the weekend at least.

What I have been doing to create a mask is to add a star mask to an inverted L mask. I think this should provide the same result as what you've described. I'll have to pay more attention to the quality of the mask when I next create one.

On 10/04/2019 at 07:42, Knight of Clear Skies said:

Thanks. It seemed to be an awfully long focal length for a refractor but I was confused by the lack of diffraction spikes (did you process them out?).

I didn't process out any diffraction spikes. There were plenty in the full frame image but none in this tight crop.

[Jkulin]

Mike, that is really lovely, colours are gorgeous.

The core I found was tricky but sorted in curves, but what a lovely image.

[Tzetze]

On 19/04/2019 at 13:18, Jkulin said:

Mike, that is really lovely, colours are gorgeous.

The core I found was tricky but sorted in curves, but what a lovely image.

Thank you John. Hmm, using curves to protect the core, good point. I think maybe just a gentle initial histogram stretch, but then stick with curves for protective stretching.

I've done another partial reworking of the image with some extra luminance gathered. There's now 5.5 hrs luminance and just 26.25 minutes each of RGB. I only reworked the lum and blended with already processed RGB. I think I'll process all again from scratch when I next get down to it.

Here's the latest WIP.

 

Some things to strive for on the next working;

Star reduction/removal of the smaller stars for a less noisy, cleaner looking background.

Use curves to not blow out the galaxy cores.

When it comes to the cores of the larger stars, is it best to try to protect these during processing, or can they be easily fixed in the later stages? If it's due to saturation of the well, surely the only option is repair?