Color Correction in a non-APO refractor

[oopfan]

It's been a while since I posted any semi-useful content, so I thought this would be a good time to chew up a couple of megabytes of Robin's storage.

It seems like years now that I've been complaining about my refractor. As you can see from my earlier posts, it can not focus blue:
viewtopic.php?f=16&t=2313
viewtopic.php?f=16&t=935

I tried several solutions, but the best one is to eliminate blue light. That leaves only red and green. So, I replaced my standard luminance filter with a Wratten #12. That worked wonders, but then I had the problem of creating RGB channels from only RG. I found a solution to that, as seen here:
viewtopic.php?f=16&t=3396

I was thrilled with that solution, but I grew weary of orange and cyan stars. So, this brings me to the topic of this post. I discovered that Astro Pixel Processor (APP) has a tool called "HSL Selective Color." In the following screenshot, NGC 1664 demonstrates what my stars look like BEFORE applying the tool, and NGC 2158 shows what the stars look like AFTER the application:

NGC 1664 + NGC 2158.jpg (154.13 KiB) Viewed 1898 times

In the next screenshot, you can see what settings I used to push the orange stars towards red, and the cyan stars towards blue:

HSL Selective Color.jpg (61.78 KiB) Viewed 1898 times

Hopefully, I'll get some good weather soon, and try this technique on a galaxy, and report back.

Brian

[oopfan]

A wise man commented that I should make a before-and-after comparison of the same object. Here you go:

NGC 1664 BEFORE and AFTER.jpg (117.23 KiB) Viewed 1872 times

We could argue that it is a tad bit too blue, etc., but overall I think it is an improvement with minimal effort. I can tweak it.

You may ask yourself: "How does this affect me? I have an APO."

The answer is that everyone can benefit from this technique. Here is an example:

NGC 2158.jpg (92.17 KiB) Viewed 1872 times

Let's say it is getting late in the season. The weather is horrible, but there is a break in the clouds tomorrow night. According to your planetarium software, you've got only one hour before you lose the object behind the trees.

What you can do is this: Capture only Red and Green frames at Bin 2.

That is what I did here:

I took the Red and Green stacks of an LRGB composition that I captured one year ago. The color stacks are at Bin 2. I drizzled them twice to correct square stars due to under-sampling. And then, I applied my two-step process:

1. In the "RGB Combine" tool, inject a small portion of the red and green stacks into luminance, then inject 67% of the red stack in the red channel and 67% of the green stack in the blue channel. The remainder goes to the green channel.

2, In the "HSL Selective Color" tool, adjust CYAN and YELLOW by amounts I stated in my original post.


Brian

[turfpit]

Brian

The side-by-side clearly shows that the blue rings surrounding the stars have been removed.

Dave

[timh]

Thanks Brian,

That's interesting - so if I am following the thread correctly - even a short focus achromat that I had not even thought about using for astrophotography might be worth a go - via keeping away from blue using the Wratten #12 - and then optionally shifting the colour range up with the HSL tool (present also in Affinity and doubtless other software).

Another way might be to go narrow band (not for star clusters but for other stuff) with H alpha and O III? assuming here that OIII (teal) isn't too blue to focus

I am fortunate enough to have an apochromat ...but it is quite long focus and the achromat would potentially be good for to provide wider-scale views if flattenable.

Tim

[oopfan]

Hi Tim,

Thanks for your reply.

Regarding your achromat, I would do a focus test first. Your scope may not focus red but do well with green and blue. If that is the case, then I would use a Wratten #44A (minus-red):

http://karmalimbo.com/aro/pics/filters/ ... ilters.pdf
https://en.wikipedia.org/wiki/Wratten_number

The Wratten filter replaces luminance, but you don't necessarily need it. You can do what I did with NGC 2158 where I just used my standard RGB bandpass filters. Capture frames with the filters that do well in your focus test. If it does well in green and blue but not red, then only capture frames with the green and blue filters. Then, when you are in APP or PixInsight, inject small amounts of the green and blue stacks into the luminance channel, and then spread equal amounts into the red, green, and blue channels.

Using Bin 2 is up to you. If time is of the essence then I would use it. You can run the following tool to see if you will be under-sampling at Bin 2:
https://astronomy.tools/calculators/ccd_suitability
If you are near the green zone then your star shape will be good, but if you are severely under-sampling then you may need to drizzle to restore round stars.

Brian

[oopfan]

Here is a good Apples-to-Apples comparison of NGC 1664 with and without "HSL Selective Color". I paid particular attention to keeping everything else the same.

I am curious how others perceive the difference. Personally, I find the left-hand image to be cold, 2-dimensional, and boring. Whereas the right-hand image is warm, 3-dimensional, and interesting. I didn't notice before, but there are substantial differences in the faint stars too.

Brian

I recommend that you download the image by right-clicking it and selecting "Save Image As..." The image file is 1024 pixels across, but your browser is probably down-scaling it to fit within the window. When you download it, you can view it in its original resolution. Better quality, I think.

NGC-1664_W12-45x60s_R-70x60s_G-35x60s_APP_S15-3-25_SA-35-25_G95_Comparison.jpg (307.49 KiB) Viewed 1803 times

[timh]

Hi Brian,

I also prefer the right hand image. It maybe partly just aesthetic - but I also think that it is probably closer to the 'real' star colours - in the case of NGC1664 itself or at least judging by internet images of the object.

Looking through my own small library of star cluster images obtained using an 'apo' Newtonian reflector and PixInsights photometric colour calibration tool - i.e judging 'true' colour that way -- I think that you might be right to use the HSL tool as a default option but maybe not fully ..i.e. I can't see examples (which is not to say there aren't any) where the cyan / yellow would be right whereas white/ blue/ bronzes are more usual. I think that your HSL shift has taken things in the right direction. Here just as an example of one of my M44 pics - complete with nice diffraction spikes - see if you agree?

But to be honest I am not quite sure....looking again at my M44 image the colour palette seems to be somewhere in between the two of yours? Maybe you could try something in between ?

Tim

[oopfan]

Hi Tim,

I like your Beehive very much! And I agree that it looks like the colors fall somewhere between my two images. When and if I ever get another clear night I will try a non-stellar object like M51. I am very curious to see the range of colors.

Thanks,
Brian