First go with StarX'

[OK Apricot]

I managed to find a tutorial on youtube, exactly the thing I was looking for with respect to creating a starless and star only layer, manipulating and merging the two at the end.

This is what I came up with after following along with the tutorial. The first was my first go at the Crescent, the second with StarX', separate layers and blending. There is obviously a lot of room for improvement here but I was just excited to see the difference, an almost 3D effect. Again I welcome any criticism!

Question - These purple/pink haloes around some stars - Where are they coming from? I shot this with the ZWO Duo Band filter, could it be here? Focus looks good to me, a result of stacking? 

 

[Adreneline]

15 hours ago, OK Apricot said:

Question - These purple/pink haloes around some stars - Where are they coming from? I shot this with the ZWO Duo Band filter, could it be here? Focus looks good to me, a result of stacking? 

I have no experience at all of using Duo Band filters so am far from qualified to help but in the absence of anyone else popping up I'll give you an opinion - for what it is worth - and hope it helps.

I used to own an ED80DS and it suffered rather badly from 'blue bloat' - not uncommon with doublets I understand. The optics are not bringing the r, g and b components together to a common focal point which results in the blue filter stars being bloated - and sometimes the red filter star. Mixing red with blue gives magenta - to my eyes a bad thing! I am very anti-magenta in images but that really is just me - lots of folk like it. Anyway, I digress. I found the inclusion of a UV/IR cut filter on the end of the reducer/flattener helped quite a bit, especially at the red end of the spectrum.

Quite how all this equate, if it even does, to the use of a Dual-Band filter I am not sure. In narrow band imaging you can get the same effect with the SHO palette where the S (very much in the red part of the spectrum) and the O (lurking between green and blue) get similarly bloated and produce magenta. The same happens with HOO in my experience.

Make of all that what you may! It may be of help and it may be a complete red-herring.

Hopefully someone who knows what they are talking about will now hopefully pop up and help you and correct me.

Adrian

P.S. Nice image, by the way, of a very challenging target.

Edited October 24, 2022 by Adreneline

[Budgie1]

16 hours ago, OK Apricot said:

Question - These purple/pink haloes around some stars - Where are they coming from? I shot this with the ZWO Duo Band filter, could it be here? Focus looks good to me, a result of stacking? 

The halos are caused by your dual-band filter. You can see the same on images taken with other dual-band filters, not just the ZWO filter. 

I'm not knowledgeable enough to give an explanation of what causes the halos with these filters, I just know I also see them on my own images when using a duel-band filter. I also don't think the addition of a UV/IR cut filter will make any difference in this case, because the wave lengths blocked by the UV/IR cut filter have already been removed by your dual-band filter. 

One thing you can do to reduce them is; when you remove the stars, the halos are often left on the background image. In PixInsight, I use the CloneStamp tool to remove the halos from the background image, you may be able to do  the same with a healing tool in the software you use.  

[tomato]

Yes, if you have the patience you can do quite a lot of cleaning up of the background on the starless image. I combine a lot of images from different scopes and the FOVs never line up 100% so despite my best efforts, lines can be seen on the final integration. These can easily be removed from a starless image, but I don’t know if I have the dedication to use the clone tool on hundreds of halos, might be ok if it is just on the bright ones.

[Budgie1]

6 minutes ago, tomato said:

 I don’t know if I have the dedication to use the clone tool on hundreds of halos, might be ok if it is just on the bright ones.

I don't do them all, just the obvious ones, which normally adds up to no more than 10, but most of the time it's just a couple.  

It also depends on when & by how much you reduce the stars. Doing the star removal at the linear stage and stretching the stars & background separately tends to leave only the largest halos. It also allows you to control the stars more easily, with less of a stretch than you would give the background. 

[The Admiral]

On 23/10/2022 at 20:48, OK Apricot said:

I managed to find a tutorial on youtube, exactly the thing I was looking for

Care to share the link?

I gather that most dual band filters cause star halos, though one or two of the more expensive ones claim not to. I'm thinking of the Antlia Gold.

Star removal certainly brings the target to the fore.

Ian

[OK Apricot]

Sure - 

Going to be fun playing around now that I've got this for reference, and once nailed down I can maybe think about trying to imitate the hubble palette with HOO. 

[ollypenrice]

I don't think the video above comes up with the best way of replacing the stars. They are selected by subtracting the starless from the starry to leave the stars. There's a better way. I also think that reducing them by using the minimum filter is very invasive.

The method suggested by Xiga here on SGL is, I find, preferable.

- Paste the starless on top of the starry.

- Invert both images. (Ctrtl I)

- Top layer active, change blend mode to divide.

- Stamp down. (Ctrl Alt Shift E ) which will give you a top layer. When this is flattened you have your star layer with stars at full intensity.

- Paste this star layer on top of the starless and change the blend mode to screen.

Now you have the stars on top as they were when you de-starred the image so you haven't necessarily gained anything - yet. But you can reduce them by using Levels and lowering the mid point (Mid point slider to the right.) This simply de-stretches them. If the stars look a bit 'hard' don't be frightened to give them a small Gaussian blur.

The million-dollar question comes before all this:  At what point do you de-star?. I'm still experimenting, but I don't find the linear stage to be good. I get good results at either a partial stretch or a full stretch. If you go for a partial stretch and then heavily stretch the starless, you can then replace the partially stretched stars and they'll need less de-stretchng (or maybe they'll need none.)  Experimantation needed.

I certainly suggest you try this method.

Olly

[OK Apricot]

Thanks Olly. I will try to give this method a go at some point. Still very early on in the stages of learning processing and photoshop as a whole, so give us some time and I will report back. 

So far with the short time I've had  practicing with StarXTerminator I've found that I first do a couple gentle curves, then sample the black and white points, balance the RGB levels, a couple more gentle curves and after that is where I run starless. It seems to bring the stars out enough to be able to do things with them, see any defects etc but not so much that they start to bloat.