ollypenrice

12 hours ago, gorann said:

Why not just try an SCNRgreen and see what happens?

Thanks, Goran. I never think to use SCNR green on non-linear images but your suggestion worked perfectly. It gave me complete control over the green-blue balance, so here it is more towards blue. I could have taken it further but I'm a believer in baby steps!

Olly

40 minutes ago, Xiga said:

The method I used is a little different.

I first processed the Ha by removing the stars and applying a bit more N-XT than usual. I tried sharpening it, but it was already quite faint and couldn't really take any. Then I added it as a new layer and double-clicked on the layer to open the Layer Style menu. From here, I changed the blend mode to Screen and de-selected the G and B channels. Then I used levels and Curves to really darken the Ha layer a lot, so that it wasn't affecting the sky background. Once I was happy with it, I changed the opacity of the layer to 80% and then made a copy of the layer, only this time I de-selected the R and G channels and changed the opacity to 20%. This way, I was able to weight the Ha 80% to Red and 20% to Blue, which seems to be the consensus for how the Pixel Math brigade do these things. 

Nevertheless, I always trust my eyes over anything else, and I admit I did like the effect. It made the Ha less of a fire-engine red, and gave it a slightly magenta hue, which I really liked, so I think I'll do it this way from now on 😃

That's interesting. The argument for adding Ha to blue arises from the fact that the fainter H-beta line, which is blue, traces the same gasses as Ha. I used to do this in the early days but seem to have got out of the habit. I'll take another look.

Olly

49 minutes ago, gorann said:

From both out images I conclude that the more detail you bring out the less it looks like an owl🤫

It is pretty green, I agree. I blended the OIII into blue and, in a separate image, into green then weighted them in Ps layers. I would like to have given the blue a little more weight but this dramatically lowered the visibility of the outer shell and the temptation to feature it was too strong!

Olly

25 minutes ago, gorann said:

 

First I process the Ha image until I am pleased with it. Then (in RGB mode in PS) I move the slider in

I would add one caveat to this. Knowing that my Ha image is destined to lighten the red channel, I don't process it as I would a standalone Ha. I go for more striking contrasts because this will be toned down in the blend, and I don't worry about noise in the background because it will be darker than the same regions in red so it won't be applied. Also I will often give a further stretch once the Ha is in place as a layer over red and I can see what it is or isn't doing.  The beauty of Photoshop lies in its showing the consequences of what you are adjusting as you adjust it.

Olly

6 hours ago, Xiga said:

Thanks Olly ☺️

I had a hard time processing this. I spent 4 nights on it, and had to start over a couple of times. This was my first time incorporating Ha into a broadband nebula image. I initially spent a fair amount of time looking up how to do it via continuum subtraction using pixel math in Siril. It worked, but then I tried just adding it in PS instead and found I could do exactly the same, only with so much more control, so I went with that. It was also my first time using Generalized Hyperbolic Stretch. I'm very impressed with it. I knew it would tame the stars well, but I was worried I would lose some of the dust along the way, but it seemed to work well. 

I continue to think that Photoshop is usually easier to control than anything else. I add Ha to red using blend mode lighten, though you can also use blend mode screen if you bring in the black point, I've read. I need to try this.

I no longer give any thought to the stars in stretching because I'm going to remove them anyway and replace them with stars of an entirely different stretch. I also remove NB stars now, when adding Ha or OIII to the colour channels. That way they have no effect at all on the RGB stars.

Olly

I haven't a clue what your unashamed cheat was and, alas, your sequence of events means nothing to me... I don't know if anyone else is any the wiser but, hey-ho, let's see!

Olly

@gorann posted a superb Owl which reminded me that I had not yet re-done mine using the x-suite software. This image has about 9.5 hours of Ha/OIII/RGB, so nothing like as much as I would have liked, but I've found more detail in the core using Blur Xt.  Recently Photoshop has added Camera Raw filters which are, essentially, from Lightroom and I think that the Texture and Clarity filters are well worth having.

This is from my TEC140 and Atik 460 mono, mounted on a Mesu 200.

Olly

Beautiful, Goran. Lovely fine detail in the main nebula along with fine outer region.

Olly

Outstanding, Ciaran. Absolutely everything is great in this image; emission, dust, stars, background (where there is any!)

Olly

On the other hand, high haze can go with good seeing and this does look really pinpoint and crisp.

Olly

Not just more outer signal: it has interesting structure.

Olly

11 hours ago, Clarkey said:

I know. Us poor folk who can't afford 12" refractors🤣

Boosted the green a bit - had trouble not messing up other part of the colour balance and making it look 'natural' without starting from scratch. No sure which version I prefer though...

I like the second one. The outer shell is mostly OIII and OIII is at the green end of blue, the opposite end to magenta.

A 12 inch refractor would be nice  but you'd probably reach the limit of the seeing with a 5 inch - thank goodness!

Olly

I agree with you on the processing challenges of this one. The dynamic range is monstrous and harder to deal with than M42. No doubt about it, though, this is a very good image.  In this case I do think the reflector star spikes are an issue since they dominate part of the nebulosity - but that's not your fault.

Regarding the colour, my feeling is that the Ha may lack contrast, meaning that it is pulling the blues generally towards magenta. I suspect that with more contrast the Ha would play less in the middle brightnesses and more in the high ones, which I think is true to the target - but I can't be sure. As for the green-blue balance, we know that OIII lies on the blue-green border, often known as Teal Blue. I'd up the greens a bit, but that's me.

Olly

I would never just accept what a plate solve gave me: I'd always want to refine the framing by hand. I look out for the stars which give an attractive framing and there can be off-shot glows or flares to avoid as well.

Olly

You've already bought the wrong kit once, as most of us did when we started. It would be a shame to do so a second time, and upgrading an EQ5 as you suggest would still not leave you with a fully satisfactory astrphotographic mount. I would be looking at buying an HEQ5 in the knowledge that this will work for you.

Think what we are asking of our mounts: we want them to track the sky with a precision amounting to an error no larger than half a pixel at the imaging camera. That works out to about one part in 2 million. When you have this, your mount's tracking is effectively perfect, in that it improving it wouldn't make a visible difference. If you have an autoguided HEQ5 this will be possible at reasonable resolution. Long focal lengths/very fine image scales would need slightly better but I would worry about that later.

HEQ5 and NEQ6 mounts come up on the used markets.

Olly

2 hours ago, AstroMuni said:

I was under the impression that OP was using for visual only. If you are saying that slower scopes would be better even for that then I need to do more reading up
🙂

Briefly, faster optics are harder to make, so more expensive. With reflectors the problem lies in minimizing optical distortions which require ever more sophisticated grinds and improved mechanical parts. With refractors, colour correction becomes harder as F ratios diminish and at some point only a triplet lens can deliver.

In both cases, accepting a slower F ratio will, in principle, result in better views when budgets are tight.

Olly

Splendid tidal streamers. Arp was an interesting character who shared Hubble's doubts about the redshift-as-distance interpretation and was in search of 'new physics' to offer an explanation. Neither of them found any but, as Feynman said, 'Science is a culture of doubt.' So far, I don't think he has been studied by a biographer but he deserves to be.

Olly

31 minutes ago, TiffsAndAstro said:

Don't think all that can be done in siril. Probably gimp though. Or combination of both.

Bit advanced for me right now I think though. 

Well, the 'subtract' method will work. Assuming you have blend mode lighten in your layers menu, there is another easy way which sometimes works nicely. Place the less-stretched original with stars on top of the highly stretched starless and set the blend mode to Lighten. Only the stars should be visible from this layer because the nebulosity and background are not lighter than the bottom layer image. You can adjust the brightness of the stars using levels. You can also bring in the black point on the top layer to make the stars tighter.

Honestly, this is simple stuff in a Layers programme.

Olly

BRAVO!

Olly

Just superb! What can I say?

Olly

The image scale is fantastic. This will just get better and better with more time, if so you choose. Very impressive.

Olly

I do it in Photoshop and follow the WolfieGlos method until I have the finally stretched starless image as a bottom layer and the less-stretched normal image as a top layer.

At this point you could use blend mode subtract in Ps but I think there's a slightly better way. Whether these blend modes are available in Siril I don't know, but in Ps it goes like this:

Invert both layers

Change blend mode to divide.

Flatten.

Invert.

This gives a stars only image which can be pasted onto the starless, as WolfieGlos says, with the blend mode screen.

We are now in exactly the same situation again. With your stars only on top you can simply use the mid point slider in levels to brighten or dim the stars to taste. You can also use the contrast slider, or Curves, to increase contrast. This will reduce the outer glow around brighter stars. In Ps the very bright stars can be lassoed and their contrast further increased to bring any bloat down still further.

This control over stars is a dream come true!

Olly

Can't say I ever expected to post this target and it's not my work, it's from the robotic observatory all sky camera run by Marco Verstraaten. But here it is, the Aurora from Latitude 44.

This is a dark site, reaching SQM22 at the zenith on a good night.  I stuck my head out at 2.00am but my view to the north was obstructed by the observatories and I simply saw a dark, starry sky.

Olly

7 hours ago, AstroMuni said:

At f4.9 its a faster scope compared to what you have (f6.5 ?) so images will be a lot better. 

I don't follow this. Visual images in budget refractors tend to be better in slower optics and this kind of telescope is possibly the worst of all options for imaging, with pretty well everything wrong. Poor colour correction, a lot of field distortion, a sag-prone focuser, etc.

Olly

Very good. These interactions are fascinating to see.

Olly