Page 2 of 2

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Wed Jul 28, 2021 6:51 pm
by Graeme1858
Here's another go using RGB>SHO

PixelMaths
ABE
SCNR - 0.25
ColourCalibration
BackgroundNeutralisation
ScreenTransferFunction/HistogramTransformation
SCNR - 0.25
CurvesTransformation - S curve for contrast. Increase RGB. Increased Red. Increased Saturation.
DynamicCrop

NGC357-02 3.png

I was going to try a star mask to tone down the magenta tinge to the stars but when I zoomed in on a preview there were a lot of other coloured stars too! So I left it.

Higher res here: https://1drv.ms/u/s!AtzUKEWck9gpv8gHIfE ... A?e=MofUk2

Regards

Graeme

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Wed Jul 28, 2021 7:38 pm
by Jockinireland
Hi Graeme,

Thats really good, I prefer it to your earlier one.

There is a neat trick in PI to get rid of your magenta stars

Invert the image - this turns the magenta stars green, run SCNR, Invert back again. Done! Give it a go.

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Wed Jul 28, 2021 8:07 pm
by Graeme1858
Jockinireland wrote: Wed Jul 28, 2021 7:38 pm There is a neat trick in PI to get rid of your magenta stars

Excellent!

NGC357-03-2.png

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Wed Jul 28, 2021 8:22 pm
by kc75star
Hi
Here's my version

https://www.dropbox.com/s/vmyu4ktatryoi ... e.jpg?dl=0

Processed with Pixinsight
- Star alignment
- Dynamic Crop
- Denoise
- LRGB combine
- SCNR remove Green
- Remove Magenta
- Selective color saturation

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Wed Jul 28, 2021 8:24 pm
by Jockinireland
Graeme1858 wrote: Wed Jul 28, 2021 8:07 pm
Jockinireland wrote: Wed Jul 28, 2021 7:38 pm There is a neat trick in PI to get rid of your magenta stars

Excellent!


Image
There y'go!! Nice one.

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Wed Jul 28, 2021 8:32 pm
by Graeme1858
kc75star wrote: Wed Jul 28, 2021 8:22 pm Hi
Here's my version

Lovely blues! Is it mapped RGB=OSH?

Regards

Graeme

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Wed Jul 28, 2021 9:08 pm
by Jockinireland
kc75star wrote: Wed Jul 28, 2021 8:22 pm Hi
Here's my version

https://www.dropbox.com/s/vmyu4ktatryoi ... e.jpg?dl=0

Processed with Pixinsight
- Star alignment
- Dynamic Crop
- Denoise
- LRGB combine
- SCNR remove Green
- Remove Magenta
- Selective color saturation
Oh! I really like that. The background clouds show through particularly well.

Nicely done!

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Thu Jul 29, 2021 4:46 pm
by Graeme1858
This one just keeps on giving!

NGC357-03-4.png

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Thu Jul 29, 2021 7:08 pm
by STEVE333
Graeme1858 wrote: Thu Jul 29, 2021 4:46 pm This one just keeps on giving!


Image

Whenever I use the SHO pallet I always use the SCNR process to remove the green. Just my approach.
You might enjoy the result with your image. Just FYI.

The picture below shows the SCNR settings I usually use.
SCNR.JPG
SCNR.JPG (18.57 KiB) Viewed 8472 times
Steve

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Thu Jul 29, 2021 7:50 pm
by Graeme1858
STEVE333 wrote: Thu Jul 29, 2021 7:08 pm Whenever I use the SHO pallet I always use the SCNR process to remove the green. Just my approach.
You might enjoy the result with your image. Just FYI.

Cheers Steve, I did use SCNR on Green but only at 0.25 because with the setting at 1.0 the result was a bit washed out. Perhaps you up the saturation later but at the time it looked ok!

Regards

Graeme

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Thu Jul 29, 2021 8:09 pm
by STEVE333
Graeme1858 wrote: Thu Jul 29, 2021 7:50 pm
STEVE333 wrote: Thu Jul 29, 2021 7:08 pm Whenever I use the SHO pallet I always use the SCNR process to remove the green. Just my approach.
You might enjoy the result with your image. Just FYI.

Cheers Steve, I did use SCNR on Green but only at 0.25 because with the setting at 1.0 the result was a bit washed out. Perhaps you up the saturation later but at the time it looked ok!

Regards

Graeme

Hi Graeme - I just tried it on your final image and got nice results. Of course that's just my opinion. You can always up the saturation.

Steve

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Fri Jul 30, 2021 6:01 am
by Graeme1858
I tried another go at SCNR at 1.0 on the my final image but is turned everything brown! Perhaps I'll try it from scratch again with a good dose of SCNR and see where it goes!

Regards

Graeme

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Fri Jul 30, 2021 3:03 pm
by STEVE333
Graeme1858 wrote: Fri Jul 30, 2021 6:01 am I tried another go at SCNR at 1.0 on the my final image but is turned everything brown! Perhaps I'll try it from scratch again with a good dose of SCNR and see where it goes!

Regards

Graeme

Hi Graeme - The SCNR will definitely add more golds and browns to the image. I tried it again with your image as follows:

SCNR
HistogramTransform to adjust black levels of Green and Red to align peaks with Blue (make background black)
Applied Blue ColorMask then increased Blue color saturation with CurvesTransformation
Applied Yellow ColorMask then increased Red and reduced Green & Blue to bring out more Golden/Brown color.
ExponentialTransformation to bring out fainter details.

The result is shown below
NGC357_SCNR.jpg

In the end it is really what looks best to you. If you like the original better, then, that is the right one for you. Everybody's personal taste is different and none are right or wrong (just my opinion).

Enjoy and have fun.

Steve

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Fri Jul 30, 2021 5:00 pm
by Jockinireland
STEVE333 wrote: Fri Jul 30, 2021 3:03 pm
Graeme1858 wrote: Fri Jul 30, 2021 6:01 am I tried another go at SCNR at 1.0 on the my final image but is turned everything brown! Perhaps I'll try it from scratch again with a good dose of SCNR and see where it goes!

Regards

Graeme

Hi Graeme - The SCNR will definitely add more golds and browns to the image. I tried it again with your image as follows:


In the end it is really what looks best to you. If you like the original better, then, that is the right one for you. Everybody's personal taste is different and none are right or wrong (just my opinion).

Enjoy and have fun.

Steve
@STEVE333 @Graeme1858

It is a really interesting to me that little or no green in SHO images has become the most popular way to do it. As I alluded to in my original post I tried to keep some green in mine because whenever I think of the "Hubble Palette" my mind immediately goes to the day I picked up the National Geographic and saw the original "Pillars of Creation" image - which had significant green in it - and nasty magenta stars :lol: . I was absolutely awed by that image and remain so to this day. The size (I remember the caption saying that each "finger" was wider across its short side than our entire solar system), the colours and what they meant, just the shear beauty of the image. So for me a SHO "Hubble Palette" should have some green in it. And honestly, probably because of my reaction to that original image I personally prefer some green in those images. But as Steve says, with NB colours there really is no right or wrong and we should do what we think looks the best (What pleases us most I guess)

But I decided to have a look at my image with the green removed using SCNR and this is what came out;
Rob_SoL_DLM_TD_SCNR.jpg
and to save you trawling back up, this is my original
Rob_SoL_DLM_TD.jpg
Now, thanks to the fantastic data from @yobbo89 these images are probably my best result from processing since I started AP - but which is the best of the two? I don't think its possible to say either is better than the other. But I can say that I still prefer the original with its green.

But the blue is pretty cool too. :D

Interested to hear your thoughts.

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Fri Jul 30, 2021 5:19 pm
by yobbo89
Jockinireland wrote: Fri Jul 30, 2021 5:00 pm
STEVE333 wrote: Fri Jul 30, 2021 3:03 pm
Graeme1858 wrote: Fri Jul 30, 2021 6:01 am I tried another go at SCNR at 1.0 on the my final image but is turned everything brown! Perhaps I'll try it from scratch again with a good dose of SCNR and see where it goes!

Regards

Graeme

Hi Graeme - The SCNR will definitely add more golds and browns to the image. I tried it again with your image as follows:


In the end it is really what looks best to you. If you like the original better, then, that is the right one for you. Everybody's personal taste is different and none are right or wrong (just my opinion).

Enjoy and have fun.

Steve
@STEVE333 @Graeme1858

It is a really interesting to me that little or no green in SHO images has become the most popular way to do it. As I alluded to in my original post I tried to keep some green in mine because whenever I think of the "Hubble Palette" my mind immediately goes to the day I picked up the National Geographic and saw the original "Pillars of Creation" image - which had significant green in it. I was absolutely awed by that image and remain so to this day. The size (I remember the caption saying that each "finger" was wider across its short side than our entire solar system), the colours and what they meant, just the shear beauty of the image. So for me a SHO "Hubble Palette" should have some green in it. And honestly, probably because of my reaction to that original image I personally prefer some green in those images. But as Steve says, with NB colours there really is no right or wrong and we should do what we think looks the best (What pleases us most I guess)

But I decided to have a look at my image with the green removed using SCNR and this is what came out;
Image

and to save you trawling back up, this is my original

Image

Now, thanks to the fantastic data from @yobbo89 these images are probably my best result from processing since I started AP - but which is the best of the two? I don't think its possible to say either is better than the other. But I can say that I still prefer the original with its green.

But the blue is pretty cool too. :D

Interested to hear your thoughts.
i like green. but i just think hso looks good with blue and gold in the 'tweaked' hubble pallet , green looks good in the less processed hubbel pallet , hso with magenta stars and that pallet looks stunning used on a target like crab nebula where alot of the elements are mixed together and makes a very vibrant colour mix . very subjective on what target you image, i realy like the dumbell in the cfht pallet

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Sun Sep 19, 2021 11:36 am
by Jockinireland
I'm bumping this older challenge because I wanted to show how the new StarXterminator plugin for Photoshop has performed (I have a post about it in image processing forum if you are interested).

Now I'm not usually much of a fan of starless images but this is an exceptional image that @yobbo89 has given us and I have to say that in this one starless, for me anyway, looks amazing. I've run it on both the green and blue versions which are 2 posts above.

If you right click on the image and "open link in a new tab" on each of the starred and starrless versions you can blink between them to see how good a job StarXterminator has done.
SoL_G_SX1.jpg
SoL_B_SX1.jpg
I'd love to hear your thoughts on this.

Take care everyone

David.

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Sun Sep 19, 2021 3:28 pm
by Sky
I've compared the StarXterminator with Starnet++ along with a free a star remover Photoshop action I also use. StarXterminator comes out on top since I'm not seeing the artifacts (especially on bright stars) that will pop-up occasionally with the other two. Fortunately, those artifacts can usually be dealt with in Photoshop.

However, with a purchase price of $60USD for StarXterminator puts it beyond what I'm willing to pay for a plug-in, since there are free alternatives available.

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Sun Sep 19, 2021 4:13 pm
by Jockinireland
Sky wrote: Sun Sep 19, 2021 3:28 pm I've compared the StarXterminator with Starnet++ along with a free a star remover Photoshop action I also use. StarXterminator comes out on top since I'm not seeing the artifacts (especially on bright stars) that will pop-up occasionally with the other two. Fortunately, those artifacts can usually be dealt with in Photoshop.

However, with a purchase price of $60USD for StarXterminator puts it beyond what I'm willing to pay for a plug-in, since there are free alternatives available.
Yes I agree it does a much better job than anything I have used up to now, but $60 does seem quite steep for a "1 trick pony" plug in.

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Mon Sep 20, 2021 1:03 pm
by yobbo89
Jockinireland wrote: Sun Sep 19, 2021 11:36 am I'm bumping this older challenge because I wanted to show how the new StarXterminator plugin for Photoshop has performed (I have a post about it in image processing forum if you are interested).

Now I'm not usually much of a fan of starless images but this is an exceptional image that @yobbo89 has given us and I have to say that in this one starless, for me anyway, looks amazing. I've run it on both the green and blue versions which are 2 posts above.

If you right click on the image and "open link in a new tab" on each of the starred and starrless versions you can blink between them to see how good a job StarXterminator has done.

Image
Image

I'd love to hear your thoughts on this.

Take care everyone

David.
wow ! , looks better then pixinsights basic star removal , very clean processing too ,very low noise .

Re: AP Processing Challenge 22 - Yobbo89 NGC3576

Posted: Mon Sep 20, 2021 3:54 pm
by Jockinireland
Hi @yobbo89

Yes I have run comparisons against Starnet++ in PI and this is much much better in terms if the residual artefacts. On the original release a few days ago it was killing some small background nebulosity (You could see that in my original post on the image processing forum) but I sent that info to him and he released a new AI version yesterday which resolves that problem. I did the above with the new AI.

I think its come out really well -but theres only one person can take the credit for the fantastic data!

There is some talk on other forums about it being made into a PI module but I'm not seeing the author (Russel Croman) giving any encouragement to that idea. I'm not sure how it would work anyway - he's charging 59 USD for it which I think is a bit much anyway - and I dont see how he'd get reimbursement at that rate as a PI module.

It is good though and I can see me paying for it.

Take care.

David