AAARRRGGGHHH!!! Stars! Hate hate hate!

[Astrosurf]

I think you get the thrust of my feeling about my trouble with stars. I've tried two star processing techniques:

Enhance star colour: http://www.waidobservatory.com/tutorials/star-color/star-color.html

Star mask: https://www.youtube.com/watch?v=UZlpDFsugYk#aid=P-7xs7XJ7Jk

With the enhance star colour tutorial I get too strong an outline, even if I keep the saturation down. But with the star mask, even with gentle stretching I get harsh star halos, as shown in this image, which go when I re-invert the mask to positive, which defeats the object. What the ^*&(*&*(*()(**&%$!!! am I doing?

Not to mention my stars just look pretty poor before I even start processing. (I've yet to try a separate capture of shorter subs to prevent over-blowing the stars and combining in PS.)

I've had some advice but I can't seem to get it right. How does this guy do it??! http://www.astrobin.com/87820/

How can people use much the same equipment and get beautiful stars? Getting seriously put off AP.

Alexxx

[Thalestris24]

Hi Alex

Being a beginner and also a noob with processing I'd like to know too! I have noticed a couple of things... Anything that aligns the rgb channels tends to wipe out colour. Likewise with increasing contrast too much. Next time I do some processing I'm going to try and split the rgb into separate channels, process them separately, and hopefully recombine them at the end. Simples! If that works then, fingers crossed, star colour will be retained. Well, that's the theory, lol.

Regards

Louise

[Astrosurf]

I hope this thread will help you when people come on board. I also have what looks like atmospheric dispersion on stars which I didn't think happened! I've isolated the stars here, so the blue and red effect can be seen. I despair!

[ronin]

You're on the wrong site if you hate stars that much. :grin: :grin:

[lensman57]

I think you get the thrust of my feeling about my trouble with stars. I've tried two star processing techniques:

Enhance star colour: http://www.waidobservatory.com/tutorials/star-color/star-color.html

Star mask: https://www.youtube.com/watch?v=UZlpDFsugYk#aid=P-7xs7XJ7Jk

With the enhance star colour tutorial I get too strong an outline, even if I keep the saturation down. But with the star mask, even with gentle stretching I get harsh star halos, as shown in this image, which go when I re-invert the mask to positive, which defeats the object. What the ^*&(*&*(*()(**&%$!!! am I doing?

Not to mention my stars just look pretty poor before I even start processing. (I've yet to try a separate capture of shorter subs to prevent over-blowing the stars and combining in PS.)

I've had some advice but I can't seem to geta it right. How does this guy do it??! http://www.astrobin.com/87820/

How can people use much the same equipment and get beautiful stars? Getting seriously put off AP.

Alexxx

Part of the problem is that DSLRs and even some CCD OSCs suffer well depth stauration quite early on in the capture so the core of the stars appear pure white or close to it, as you staurate the star, the  outer halo shows more colour as most of the core area is now pure white. The best way of processing nebulea or DSOs is to remove the stars, process the nebula and paste the stars back in. Annie's actions for PS has an automated action that helps a lot. I use StarTools and PI and the procedures  for star removal are rather different to PS.

A.G

[Thalestris24]

+1 for StarTools! One of these nights, if the clouds ever go away for a few hours, I shall attempt to image a nebula! Will be stoked if I can manage to get an even half decent one! With nicely coloured stars too, of course .

AG - I'll bear in mind the well depth next time. Not sure if it's the cause of the lack of star colour in my case but could be a factor. I'm still getting used to the qhy8l - quite a change from the Canon.

Louise

[Astrosurf]

Thanks a.G. Please can you post a link to Annie's Actions?

I've been working at isolating the stars using the wand tool and other stuff, and I got a better image of my Eagle Neb but with some star bloat. I used the following tutorial as well (having to modify it for CS6) and it helped, but I was a bit clueless about what I was doing:

http://starmatt.com/articles/StarShaping.html

Here are before and after images:

Still a looooong way to go. I really need to get some captures with shorter exposures.

[lensman57]

Thanks a.G. Please can you post a link to Annie's Actions?

I've been working at isolating the stars using the wand tool and other stuff, and I got a better image of my Eagle Neb but with some star bloat. I used the following tutorial as well (having to modify it for CS6) and it helped, but I was a bit clueless about what I was doing:

http://starmatt.com/articles/StarShaping.html

Here are before and after images:

Eagle Nebula .jpg

Eagle Nebula New Processing.jpg

Still a looooong way to go. I really need to get some captures with shorter exposures.

Here it is,

http://www.eprisephoto.com/astro-actions

A.G

[Astrosurf]

Thanks A.G.

Can anyone tell me what techniques they use to improve star colour?

[lensman57]

Thanks A.G.

Can anyone tell me what techniques they use to improve star colour?

You can use the Annie's actions to increase the star colour, failing that you have to set an effective star mask in PS, there are a few tutorials on the web on how to do this, once the mask is done run a Gaussian Blur about 1~1.5 pixels on the mask to blur the edges and then use the stauration tool to taste, do not over do it as it would look false.

A.G

[Thalestris24]

Here it is,

http://www.eprisephoto.com/astro-actions

 

A.G

I don't suppose Annie's Actions work with cs2.... cs3 maybe? If cs3, I might try and buy a copy but otherwise PS seems a bit too expensive... I can split up my rgb fits images into separate mono files but don't really understand how you process them individually without skewing the colours. I did manage to preserve colour in a combined rgb in StarTools last night so some progress . I'm getting a debayered 1100d soon so hope to use it to add Ha/luminance - eventually.

Cheers

Louise

[lensman57]

I don't suppose Annie's Actions work with cs2.... cs3 maybe? If cs3, I might try and buy a copy but otherwise PS seems a bit too expensive... I can split up my rgb fits images into separate mono files but don't really understand how you process them individually without skewing the colours. I did manage to preserve colour in a combined rgb in StarTools last night so some progress . I'm getting a debayered 1100d soon so hope to use it to add Ha/luminance - eventually.

Cheers

Louise

I don't see why not.

[andrewluck]

I don't recommend spitting the RGB planes out to process them separately. You're just making life harder than it needs to be as colour balancing will be almost impossible. Get the colour balance right early on and then process RGB as a single unit.

If you look at brightness levels in the final image (8 bit) then anything with a pixel value over 200 will start to look washed out and lose colour. To avoid this you need to protect the stars when you're performing the first non-linear stretch. These days I use the Pixinsight Masked stretch tool pretty exclusively for this. This tool performs the stretch iteratively, using the output from each stage as a mask for the next. This way, the brighter something gets, the more masking it receives on the next stretch. This is very effective at controlling dynamic range in the image.

It should be possible to emulate this approach using Photoshop. You don't need to worry about generating star masks, just use the entire image as a mask for the next stretch and keep each stretch pretty small. Once this is done, then use another mask and increase the saturation for any object that's not background. Once the nebula looks right, then you can use a star mask to further increase saturation on the stars if required.

Of course, this all depends upon the stars not being completely over-exposed to start with. I don't generally find this as much an issue with my OSC QHY9 as it was with a DSLR. With DSLRs, keep the ISO low as this will increase the available dynamic range on the camera. With my 350D I stuck to 5 minute exposures at ISO 400 as a compromise. With brighter objects I reduced the ISO rather than the time.

Andrew

[Thalestris24]

I don't recommend spitting the RGB planes out to process them separately. You're just making life harder than it needs to be as colour balancing will be almost impossible. Get the colour balance right early on and then process RGB as a single unit.

 

If you look at brightness levels in the final image (8 bit) then anything with a pixel value over 200 will start to look washed out and lose colour. To avoid this you need to protect the stars when you're performing the first non-linear stretch. These days I use the Pixinsight Masked stretch tool pretty exclusively for this. This tool performs the stretch iteratively, using the output from each stage as a mask for the next. This way, the brighter something gets, the more masking it receives on the next stretch. This is very effective at controlling dynamic range in the image.

 

It should be possible to emulate this approach using Photoshop. You don't need to worry about generating star masks, just use the entire image as a mask for the next stretch and keep each stretch pretty small. Once this is done, then use another mask and increase the saturation for any object that's not background. Once the nebula looks right, then you can use a star mask to further increase saturation on the stars if required.

 

Of course, this all depends upon the stars not being completely over-exposed to start with. I don't generally find this as much an issue with my OSC QHY9 as it was with a DSLR. With DSLRs, keep the ISO low as this will increase the available dynamic range on the camera. With my 350D I stuck to 5 minute exposures at ISO 400 as a compromise. With brighter objects I reduced the ISO rather than the time.

 

Andrew

Hi Andrew

I lost the colour when processing in StarTools. I know what I did wrong now but haven't had a chance to put it right and since I don't have enough subs, I've not really put much effort in on this one. Since Olly Penrice extols the advantages of mono cameras I was trying to emulate this by processing the rgb separately. Maybe this is a mistake but I thought it would be interesting to try! I've not imaged a nebula yet... It's on my todo list!

I'm currently imaging with a qhy8l which gives me fits format files. So debayer via dss. It's possible to split the file into separate r,g,b files using imageJ but if I load them into StarTools it recombines them. Maybe I'll keep it simple and just use StarTools .

Cheers

Louise

[ollypenrice]

I never process the colour channels separately. You would have a heck of a job creating the same stretch in each and I can't see any reason to do this. If capturing with an OSC camera your channels have to be aligned - with one caveat; it would be possible for a mis-collimation or other quirk of the optics to disperse one end of the spectrum differently from the other. What is odd about Alex's image is that the red channel does look misaligned, with some red bloat showing to the lower right. To put it another way the stellar cores are offset to the upper left. This is quite original, Alex!

Two things come to mind for me. 1) Is the focus really nailed? I'm inclined to doubt it and perfect focus is critical to getting good stars. 2) Could the chip be slightly tilted? I'm fairly suspicious here, too.

I would only do complicated things like removing and replacing stars if there were a real imperative to do so and then only in parts of the image. The more you muck around the more you introduce artefacts. I've been forced into it when stretching the outer shell of M27 into visibilty. The fierce stretch also stretched the stars in front of the halo so I removed and replaced those. http://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/i-VTWp94c/0/O/M27_HaO111RGBWEB.jpg

So focus and/or chip tilt? Only guessing though.

Olly

[Physicist13]

what Olly says...

also though...and I've def seen this in my DSLR pics, is that the red channel being longer wavelength gives larger star images. so the final composite tends to have a red ring around them. Yours is a bit like this but skewed as well but cld be similar effect. This effect is exaggerated the more the focus is off, the more the stars are close to saturation and the more you want to stretch the image or boost the saturation.

Mostly I'm not very fussy!1 ...my pics for me so I tolerate it!

If being fussy, I sep out red channel and reduce star sizes by an amount to match those in the G and B channels (resonably easy to do in PI). then combign again and go from there....

Patrick

[Astrosurf]

Thanks guys. I noticed on another imager's pics that there was red one side of the stars and blue on the other, like you see when imaging a planet. I hope it's not my sensor! I use B masks with my scopes and lock the focuser, so I don't think my focusing's out, unless temperature change affected it here. Here's an image I took of the Andromeda galaxy with my frac, and processed by Mel Gigg. The stars look OK. Could the problem with the Eagle Neb, taken with my 200P, have been poor collimation and coma?

[Davey-T]

Sorry to say in my pic of M16 stars look like yours and that's with CCD and filters so may not be down to OSC effects, probably down to processing skills in my case.

I still like it though because it's the first nebula I imaged.

Dave

[Thalestris24]

Thanks guys. I noticed on another imager's pics that there was red one side of the stars and blue on the other, like you see when imaging a planet. I hope it's not my sensor! I use B masks with my scopes and lock the focuser, so I don't think my focusing's out, unless temperature change affected it here. Here's an image I took of the Andromeda galaxy with my frac, and processed by Mel Gigg. The stars look OK. Could the problem with the Eagle Neb, taken with my 200P, have been poor collimation and coma?

Um you defo have coma - it always shows as elongation in the corners. I don't know about collimation but I'd say you increased the star contrast too much and then tried to compensate the colour saturation. I only say this because I've done the same myself and got similar results..

[Astrosurf]

Very likely. I did try to reprocess while protecting the stars in various ways. I might try two sets of subs, as I said, shorter ones for the stars and longer for the nebula.