Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Blue bloat reduction...

  • Please log in to reply
5 replies to this topic

#1 Devonshire

Devonshire

    Explorer 1

  • *****
  • topic starter
  • Posts: 58
  • Joined: 10 Mar 2017
  • Loc: S/W Ontario, Canada

Posted 13 February 2019 - 11:26 PM

I've been working on my image processing, and having the occasional success along the way - but one thing I've not been able to fix is the sense that my stars overall are a bit more bloaty than they should be. Trying to figure out whether this is fixable in processing or not.

 

I pulled down a copy of RawTherapee, and fumbled about with my raws until I managed to get it to display the RGB and L channels.  Aha! R and L stars are crisp and tight, G has a dark ring or two, but otherwise similar to R and L, but B is bloaty, as I expected it might be.  I gather that this is because the blue is slightly out of focus, relative to the other colors.  

 

I'm wondering if there's a way to preprocess the raws so that the Blue channel image gets shrunk back on top of (say) the Red.  Refocusing in software, if you will, using one or more of the other channels to define the result of an ideally focused lens, and then correcting the Blue image to conform to that.

 

Any ideas?

 

- Bob

 

Gear and Software involved:
Camera: D5300
Software: DSS v4.1, Startools v1.4, and RawTherapee v5.5
Scope: SV 102EDT



#2 Jerry Lodriguss

Jerry Lodriguss

    Vendor

  • *****
  • Vendors
  • Posts: 6905
  • Joined: 19 Jul 2008
  • Loc: Voorhees, NJ

Posted 14 February 2019 - 12:02 AM

You can try a star reduction in the blue channel in Photoshop, just use the minimum filter set to round.

 

Try to match the star sizes of each individual channel to the smallest original one.

 

Or, you can use the blue halo reduction tool in Photoshop's Raw filter.

 

Or, sorry I see you don't have Photoshop.

 

Nevermind.

 

Rosann Roseanna Danna


  • jstrandberg, Starman27 and Jim Waters like this

#3 james7ca

james7ca

    Fly Me to the Moon

  • *****
  • Posts: 6329
  • Joined: 21 May 2011
  • Loc: San Diego, CA

Posted 14 February 2019 - 01:12 AM

Sure, you can use software to shrink the star size in your blue channel but you aren't likely to get exactly the same star shape and brightness profile that you'd get on the other channels. So, it's a matter of some compromise.

 

If you are mainly concerned about matching the star size in each of the channels and you have some resolution to spare then you might shrink the blue and enlarge (blur or convolution) the red and green channels.

 

If you want to shrink the blue channel then you want to look for software that offers some kind of deconvolution action. I'm pretty sure Startools has a deconvolution routine, but I don't know what it is called in that application (perhaps some kind of sharpening).

 

You can also use unequal masking on the channels when you perform your histogram stretch.

 

Lastly, you can try a mild desaturation in the blue channel using a mask to isolate just the blue bloat.

 

In all cases, you want to use fairly mild adjustment and/or a little of all of these techniques (sharpening, masking, and desaturation).


Edited by james7ca, 14 February 2019 - 01:49 AM.


#4 rekokich

rekokich

    Viking 1

  • *****
  • Posts: 705
  • Joined: 14 Sep 2014
  • Loc: Virginia

Posted 14 February 2019 - 04:18 AM

Blue halos may be in part due to chromatic aberration. To improve the focus in the blue channel with refractors, move the camera ever so slightly inward, toward the objective lens. This may partially correct the problem by bringing blue into better focus.

 

While camera sensors are quite sensitive to UV light (displaying it in violet or purplish color), only the best 4 or 5 element APOs are well corrected for near-UV. Also, atmospheric Rayleigh scattering always plays a MAJOR role in creating blue star bloat with refractors and reflectors. It is exponentially progressive at shorter wavelengths.

http://hyperphysics....mos/blusky.html

 

Imaging through a good UV block filter reduces blue star bloat from chromatic aberration and Rayleigh scatter. Astronomik L3 filter and the Baader Fringe Killer block light below 420nm.. Tiffen Haze 2 A is a good cheap alternative, but it is not coated and suffers from internal reflection artifacts on very bright stars.

https://www.cloudyni...trophotography/

 

Finally, there are several ways to reduce star size in StarTools. All involve first creating a good star mask which is slightly larger than the stars. If star shapes are severely deformed, use the Repair module / Redistrbute algorithm. If stars have prominent halos, the star mask must be enlarged to include the halos, then use the Magic module / Tighten mode.

 

Rudy


Edited by rekokich, 14 February 2019 - 06:59 AM.


#5 Devonshire

Devonshire

    Explorer 1

  • *****
  • topic starter
  • Posts: 58
  • Joined: 10 Mar 2017
  • Loc: S/W Ontario, Canada

Posted 14 February 2019 - 04:57 PM

Interesting ideas...

 

It never ocurred to me that the blue bloat might be coming from UV, either at the chip or via scattering - and yet, lots of hits if I search for that.  Thank you!

Lots to research now, as well as taking a closer look at just shrinking stars in processing.

 

Couple of questions:

 

1.  The Baader UV/IR cut filter appears to pass a broad, flat, spectrum that is pretty similar to the Astronomik L3, and yet it doesn't get the ink that its sibling Fringe Killer gets.  Any idea why?

 

2.  Some filters (LP, UV, etc.) get some credit for reducing CA, but then get a knock for introducing their own color bias and reducing light throughput.  What's the color/throughput experience with the L3, the Baader FK, and the Baader UV/IR?

 

Thanks!

 

- Bob



#6 rekokich

rekokich

    Viking 1

  • *****
  • Posts: 705
  • Joined: 14 Sep 2014
  • Loc: Virginia

Posted Yesterday, 03:08 AM

Bob,

 

Excluding wavelengths of light which degrade the image is considered desirable. Every filter and every color camera sensor introduces some color bias. The change will be minimal with the Astronomik L3 and the Baader Fringe Killer, but the CLS-CCD filter will introduce a fairly prominent green color cast.

 

There are several ways to easily correct the color balance. One is to adjust custom white balance in the DSLR camera itself by using a previous image as a sample. Another one is to align color histograms in the Deep Sky Stacker. Freeware XnView has options for manual and automatic color and contrast adjustment which produces excellent results half the time. And, StarTools has options for correcting color in Color, Wipe, and HDR modules.

 

If you have an unmodified camera, and need to cut only UV and violet, I would recommend to start with the Tiffen Haze 2A which is very affordable, and will serve you well on all but bright stars.

 

For full spectrum modified cameras, I would recommend the Baader FK or the Astronomik L3 as a very close second.

 

Rudy




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics