Another crack at the Crescent...

[+Derbyshire Dave]

Hi,

 Posted an image of the Crescent Nebula yesterday, without fully realising that it ws, in fact, well to the blue side of the scale. Thanks for pointing it out guys. So this morning's task was to try and track down how I had managed to produce the blue babe.

 

 It took me quite a while to realise what (I think!) is going on. Sharing this as it might be a lesson not to just me!.

 

The data from the three filters, Sii allocated to red, Ha to green, and Oiii to blue actually had a bias in it. The blue, although there was less area under the graph, was actually offset to the right (i.e. brighter) than the other two signals.

 

The way that I used to process the data was alternative steps in Photoshop of curves and level, and had I folllowed that, the offset would have been 'killed off' early in the process.

 

 

What I had got into the habit of doing though, was to do all of this in one step in PixInsight. I think the blue amplitude was getting magnified disproportionally early on.

 

So, went back to the original Photoshop curves/levels/curves'levels process and the blue did not get overblown. It even allowed me to get a bit more brightness in the picture, which in my mind, is much better.

 

 

And this is the result.. As usual, click, then click on Flickr for full res..

 

 

[+RonC]

That looks well balanced Dave!  👍

[Sheila]

That is better perhaps boost the ha a little to get more of the surface data if you wish as well as the blue halo. But much better

[Clive]

That does look better Dave 🖒

I usually set the all the histogram peaks to the same value (since for most astrophotos that is the level of the sky background) and then use curves if required to match the shape of the histograms. However, I'm not sure whether PS shows the actual peak since it seems to clip them.

[+Graham]

That looks better Dave

[+Derbyshire Dave]

55 minutes ago, Clive said:

That does look better Dave 🖒

I usually set the all the histogram peaks to the same value (since for most astrophotos that is the level of the sky background) and then use curves if required to match the shape of the histograms. However, I'm not sure whether PS shows the actual peak since it seems to clip them.

Thanks Clive.

 

The hsitogram tool in Photoshop could be much better from an astronomy point of view. I've wondered, and looked, in the past for a plug-in better version, but not found one.

[Clive]

9 minutes ago, Derbyshire Dave said:

Thanks Clive.

 

The hsitogram tool in Photoshop could be much better from an astronomy point of view. I've wondered, and looked, in the past for a plug-in better version, but not found one.

I guess it was designed for everyday images that generally tend to have the histogram more evenly spread over the full intensity range. The same is true for Affinity. I've created an Excel spreadsheet that plots out the histograms so I can examine them in detail if I need to.

[+Derbyshire Dave]

19 minutes ago, Clive said:

I guess it was designed for everyday images that generally tend to have the histogram more evenly spread over the full intensity range. The same is true for Affinity. I've created an Excel spreadsheet that plots out the histograms so I can examine them in detail if I need to.

Impressed Clive. The PixInsight Histogram tool is good, now that I know where the issues are, I'll be having a go at replicating my Photoshop Curves/Levels/Curves in Pixy..

[+Glafnazur]

That's a lot better, I really like that 👍

[+Ibbo]

You are turning into a proper pixie.

 

[Neo_uk]

Great image ....just love narrow band ..