fringe killer filter add blue back to central star

[alacant]

Hi
Firstly thanks for a great astrophoto product which allows us to create images, rather than having to jump through hoops to get anywhere near!

My request is that in:
filter > filter mode > fringe killer
would it be possible to add an option for the blue spectral lines to be added to the central star pixels, rather than removing them?

I feel sure that those of us with affordable telescopes would benefit greatly from this when attempting to whiten and otherwise colour correct stars.

Thanks again.

[admin]

Hi,

Would you be able to post a mock-up or some sort of image of what you're after?

Thanks!

[ChrisLX200]

I'm not sure if it's possible but... if you can apply the 'Magic' tool to just the Blue channel that would effectively shrink the blue halos without losing saturation in the star cores. If you cannot do it in one step like this (and I've never tried this so I don't know for sure) then you could, as an alternative, split your RGB image to separate R, G and B channels and then apply the Magic tool to just the Blue channel before re-combining to RGB again. It would take some guesswork for the amount of star shrinkage but generally you would be aiming to shrink the stars in the Blue channel until they approximate (on average) the sizes you see in the R and G channels. This can only be a guess - some stars will naturally be Blue and of course there will be all other colours as well, it won't be a perfect answer but you should be able to markedly improve an image that is suffering from 'Blue Bloat' in this way.

ChrisH

[alacant]

OK, I'll try the magic. I think that my stars are bloated because of the blue being out of focus. If I separate rgb, I can see the blue as circles wheras the red and green are points. Rather than just remove the blue using filter I'd like to be able to concentrate the blue energy on the same point as the red and green. I realise that I'm fighting affordable achros and understand this may not be an easy task.

[admin]

OK, I'll try the magic. I think that my stars are bloated because of the blue being out of focus. If I separate rgb, I can see the blue as circles wheras the red and green are points. Rather than just remove the blue using filter I'd like to be able to concentrate the blue energy on the same point as the red and green. I realise that I'm fighting affordable achros and understand this may not be an easy task.

You're definitely thinking in the right direction! This is chromatic aberration and what you really want/need is deconvolution of the blue channel. The only trouble with deconvolution is that it does not deal well (or at all) with overexposed areas such as... stars. That's because there is no data in something that is overexposed (it's clipped). That's why decon of the blue channel is very tricky. You could give it a try though (just load blue only through the LRGB module). StarTools' decon algorithm has a pretty unique ability to rather effectively coalesce over-exposed star cores into point lights without much ringing (to see it do that, make sure you don't use a star mask at all). That said, it's a tricky one and YMMV. Would be interested to hear how you get on!

[alacant]

OK
I've been going some time now. I can't for the life of me get those stars any smaller. The red and green LGRBs are small circles and look great in black and white on their own. The blue is fat and ugly. Here is the and auodev > wipe > dev 80% blue tiff:
https://drive.google.com/file/d/0B_uUt0 ... sp=sharing

I've tried deconv without a mask, with a mask, repair... Just how do I concentrate those blobs? Any clues gratefully accepted.

It's cloudy in Alicante ATM, but the good news is that I just got an old rusty EQ6 mount with bitumen lubricated bearings. If I regrease it, do you think it will fix that blue channel?!

[admin]

It sounds to me your problem is axial chromatic aberration, which finds its origin somewhere in your optical train. I don't think it's your mount (your stars don't trail).

What equipment are you using?

With regards to your data as-is and finding a post-processing solution, things get a lot trickier. Would you be able to upload your data so I can do some experiments?

[alacant]

Hi Ivo
NP:
https://drive.google.com/file/d/0B_uUt0 ... sp=sharing

The red and blue separations give tight stars. The blue, not.
TIA

[admin]

Yep, that chromatic aberration is a killer.

Try fixing acquisition/optics first, but there is there is one last thing you can do in post-processing.
In the Develop module ("stretch as-is" if you're otherwise happy with your stretch), you can use the luminance contribution mixer to take away all luminance contribution by the blue channel. This means that for any detail, only the red and green channel are used; color information is still used from all channels (including blue). This feature was mainly meant for planetary imaging where certain bands may be very polluted/unreliable/susceptible to atmospheric interference.

It's not a great solution for blue nebulosity (though you can apply the luminance mixing to just the stars - e.g. "fat stars" preset - using the Layer module and the undo buffer)

Your stars can definitely be salvaged, however it's a lot of work and custom processing; something that's fine if you know what you're doing but I wouldn't recommend it part of a staple work flow...

NewLRGBComposite.jpg (101.66 KiB) Viewed 20080 times

The above was accomplised by applying decon (in Planetary mode, e.g. no mask used at all) to the blue channel (all image pre-binned to 50%), shrinking the stars using the Magic module and re-merging it with red and green in the LRGB module.
The Develop module was used in the manner described above ("stretch-as-is") after I used AutoDev with a ROI over a star for the general stretch.
The final result (after switching Tracking off) was put through the Fringe Killer to remove any strong/deep blue.

StarTools is pretty powerful and will let you do all this if you really want/need to, but, again, it's always best to address these issues at the acquisition stage.

Hope this helps at all!

[alacant]

Thanks Ivo. That's just what I wanted. I think that with learning any new software, familiarity counts for a lot. Whilst I realise more or less what the modules do, I would never have come up with that workflow myself. I'm gonna learn it.

I take your point about correcting errors at source but I think it's good to have a go with what you have rather than not having a go at all. There's no time soon when I'm gonna be able to get an APO so for the time being, this is great. Thanks again.