Star Color & L-Extreme narrow band filter

You can salvage star colors by carefully stretching your image (using arcsinh or masked stretch).
This will not give your true colors, but not totally white stars either.

For RBG: Stars are very bright, so you won't need much exposure time to get a good result.
Your subexposure times will greatly depend on the well capabilities of your camera.

Hello Tim,

It is doable and I've done it recently (my first attempt) in these two photos:

Pickering Triangle and Witch's Broom



Crescent and Soap Nebulas and its Surroundings

I found easier to replace the chrominance only, not chrominance+luminance, and the key thing is to build a proper star mask. In terms of how long, it depends on your camera and on your NB photo. Stars should be, more or less, of similar size and it will depend on your specific conditions. But in these cases, I took 10 x 1 min exposure for the stars.

There is another technique which I haven't tried which is to make a synthetic green and combine the stars differently to achieve a quasi-RGB. But my experience is that stars have different sizes in Ha and OIII with dual-NB filters and I think that does make difficult to apply such technique.

Cheers,
Andre

In Startools you can add NB data to rgb data assuming both stacks are scaled the same ,using the NB accent module .

Try StarXterminator, a photoshop plugin to remove the false colour stars associated with the extreme filter then replace with normal colour stars using the colour camera. Assuming you have Photoshop though.  You will need to use Match colour to bring back the colours as they get rather washed out in the process of combining the 2 images.

If you can take RGB only  - say with L Pro or Ir cut only then you replace the stars . I simlpy use Pixelmath and there is a lot of online tutorials to show you how. Here is l extreme with RGB

M20 Trifid

I use the photometric color calibration tool in Siril. It plate solves your image and then adjusts the white balance based on reference color measurements for the stars in your image.   PixInsight has a similar tool.

Thanks for your response - I have actually tried that in PixInsight. It really helped with the background color but didn't really change the color of the start. Perhaps, I wasn't using the correct settings??  Thanks Again

Tim Gehring:
While I have dabbled with Astrophotography in the past with my limited equipment taking pictures of solar system objects (sun/moon), I have recently (about 6 months ago) fallen down the rabbit hole of DSO astrophotography !  I have really enjoyed my journey to date. Feel free to view my images to date. 

I have been playing with the L-Extreme  NB filter and I really enjoy that data I capture. However, I would love to able to "fix" the start color to something approaching realistic. I have read that you can capture broadband images (same framing as NB image) and just swap out the stars using Starnet in PI and process the start as needed  and add to your NB starless image with Pixelmath as part of your post processing workflow. Sounds doable (probably harder than I think) but how many broadband images and how much exposure time?

Are there other ways to correct the star colors? 



Heart Nebula


Thanks in Advance,
Tim

Tim,
   There are a number of ways to correct this. I am using a canon DSLR with an astronomik UHC filter, which is the same as the filter you mention in this regard—it is also a quad-band narrow band filter. So it does the same things to your stars. I have also dealt with your issue. I can only speak in terms of the software tools that I use for processing. 

a couple of options:
1) if you are processing in photoshop you will want to set you white point & black point in a threshold layer using the color sampler tool.  You can later delete  this layer & it will retain your white & black points. There are a lot of YouTube videos on how to do this. I can point you to some if you’d like. set the black point is on what you know to be black sky. Set your white point on what you think is a yellow star like our sun, a class G2V star. There are websites were you can look up class G2V stars in your FOV. Our human brains are evolved to see the yellow light of our sun as “white light”, so this technique will color balance stars to this as a white point, which will compute as “looking right” to our brains.

this fixes it, though it takes practice & trial & error to learn to get this right. 

2) second option—I begin using deep sky stacker to stack my images. It is an excellent program for that, so my next comments are not a criticism of DSS. For freeware, it is phenomenal & overall does a good job. However, I switched to using astropixel processor for stacking, which is not freeware. It uses a bit more advanced stacking algorithms which I think do a little better & it also has some good pre-processing tools. Particularly, it has one of the best light pollution removal tools anywhere (and very easy to use) & it also has a scientifically based star color calibration tool. It works, though adjusting the star color calib tool correctly is kind of complicated.

I typically stack all my subs in astropixel processor & then do the light pollution removal. I will then save 1 version as a TIFF at this stage. If I take this TIFF straight into photoshop without even doing the APP “star color calibration” function the star colors are pretty close to correct right there from images that came from my quad band narrow band filter. This was never the case when stacking from DSS. Bottom line, APP does help with color calibration a great deal with narrow band, but it is not the end all be all. Certainly helps though.  

now if I do the star color calibration process in APP, it gets the star colors better, but that tool takes some finesse to get “perfect.” Even if you are not quite “right” stars & color balance are still far better out of the box as a stacked image than I saw with DSS. That is not a shortcoming of DSS, it’s just that APP is built to do more things than DSS is intended to. 

hope this helps.