Smart Sharpen and Unsharp Mask

this is the main message area for anything solar :)
Post Reply
User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Smart Sharpen and Unsharp Mask

Post by Montana »

This afternoon as it is the most awful weather here I have been reprocessing my Solarscope photographs using the above mentioned tools. What a miracle these are and many thanks to Chris Schur and Ken Crawford for all the explanations. I will post my efforts but if you do happen to have further tips and criticisms I would love to hear them. All taken with Solarscope SF-100 / DMK41.

First a picture from last November:


2011-11-06 13-49-42 SS by Alexandra's Astronomy, on Flickr

compared to http://www.flickr.com/photos/alexandra4 ... /lightbox/ (Registax)

Now pictures from last weekend, full disc
http://www.flickr.com/photos/alexandra4 ... 535811777/
compared to registax http://www.flickr.com/photos/alexandra4 ... 535811777/

More:

2012-01-15 13-29-32 SS by Alexandra's Astronomy, on Flickr


2012-01-15 11-41-33 SS by Alexandra's Astronomy, on Flickr


2012-01-15 13-36-26 SS by Alexandra's Astronomy, on Flickr

I need to check out Chris and Ken's prominence techniques now!

Alexandra


User avatar
marktownley
Librarian
Librarian
Posts: 42270
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 20424 times
Been thanked: 10243 times
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by marktownley »

These are fantastic images Alexandra - they remind me very much of Lucas. It really does bring it home when you compare to the registax and non-registax images side by side. Which technique did you use though - the USM or SS?


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Re: Smart Sharpen and Unsharp Mask

Post by Montana »

Thanks Mark, I have been staring at the November image for the last couple of days as I was so mesmerised by it, the rest I did today. Well, for the 2x images Smart sharpen works the best, for the prime focus images Chris's iterative Unsharp mask works the best. However if I don't write down now what I have just done I will forget again.

OK, I have tried on the prominences and watched the tutorials and I have absolutely no idea what I am doing.

Can anyone tell me what a mask is, how you create one and what on earth it does??? it's a mystery. There seems to be many, vector, clipping, pixel etc and none of them seem to do anything apart from sit there. Buttons I can press, masks are just weird.

Regards
Alexandra


User avatar
marktownley
Librarian
Librarian
Posts: 42270
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 20424 times
Been thanked: 10243 times
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by marktownley »

Thanks Mark, I have been staring at the November image for the last couple of days as I was so mesmerised by it, the rest I did today. Well, for the 2x images Smart sharpen works the best, for the prime focus images Chris's iterative Unsharp mask works the best. However if I don't write down now what I have just done I will forget again.

Interesting; i've been playing around with both methods myself too... I found that the USM method works better on my lower contrast images - those (few) that i've recorded in single stack mode and when I use the 2.5x powermate. With the latter the image is very oversampled and lacks alot of contrast compared to the image scale I work at when recording data for my full disks. When in double stacked 'high contrast' mode I find I need to adopt a 'modified' smart sharpen approach to get best results: If I have the 'more accurate' box checked I find it far too aggressive and gives results similar to 'overcooked registax wavelets'. However, for me, I found the best method is a combination of Chris and Kens - I use the smart sharpen, without more accurate checked, but then apply this in an iterative way - maybe 3 or 4 iterations.

Have you tried with any your DS PST shots Alexandra? Would also be interesting if any other 'double stackers' have tried this and what results they got...

Now, might have to try these masks myself...


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
Ken Crawford

Re: Smart Sharpen and Unsharp Mask

Post by Ken Crawford »

Thanks Mark, I have been staring at the November image for the last couple of days as I was so mesmerised by it, the rest I did today. Well, for the 2x images Smart sharpen works the best, for the prime focus images Chris's iterative Unsharp mask works the best. However if I don't write down now what I have just done I will forget again.

OK, I have tried on the prominences and watched the tutorials and I have absolutely no idea what I am doing.

Can anyone tell me what a mask is, how you create one and what on earth it does??? it's a mystery. There seems to be many, vector, clipping, pixel etc and none of them seem to do anything apart from sit there. Buttons I can press, masks are just weird.

Regards
Alexandra

Those are very nice images!! Don't forget that in Smart Sharpen you can set the Remove Gaussian and that is the same algorithum that USM uses except you will have more control. So if you like USM then give that a try. Also, the More Accurate Check Box will do a double pass so if you don't have small details to bring out, try unchecking that box.

You don't have to use Masks (alpha channels), it is an advanced processing method that I feel is very powerful but with Smart Sharpen you can deal with the prominences with out the masks. I did not use masks in the Smart Sharpen tutorials. First master the tools without the masks then go on from there.

If you want to move into masks, maybe we can Skype for a one - one lesson.

Regards,


Ken Crawford

Re: Smart Sharpen and Unsharp Mask

Post by Ken Crawford »

[quote]Thanks Mark, I have been staring at the November image for the last couple of days as I was so mesmerised by it, the rest I did today. Well, for the 2x images Smart sharpen works the best, for the prime focus images Chris's iterative Unsharp mask works the best. However if I don't write down now what I have just done I will forget again.

Interesting; i've been playing around with both methods myself too... I found that the USM method works better on my lower contrast images - those (few) that i've recorded in single stack mode and when I use the 2.5x powermate. With the latter the image is very oversampled and lacks alot of contrast compared to the image scale I work at when recording data for my full disks. When in double stacked 'high contrast' mode I find I need to adopt a 'modified' smart sharpen approach to get best results: If I have the 'more accurate' box checked I find it far too aggressive and gives results similar to 'overcooked registax wavelets'. However, for me, I found the best method is a combination of Chris and Kens - I use the smart sharpen, without more accurate checked, but then apply this in an iterative way - maybe 3 or 4 iterations.

Have you tried with any your DS PST shots Alexandra? Would also be interesting if any other 'double stackers' have tried this and what results they got...

Now, might have to try these masks myself...

Wow Mark,

That is interesting! With your 2.5 powermate you have more information to be able to dig out as your images scale gets finer. So the more accurate checkbox should work better for you at that point. You have to be able to get the radius setting setting right on and the blend between the shadows and highlights correct for best results. If it looks over cooked then the setting are not correct between the Amount and Radius. Depending on the sharpening method (Remove Lens Blur or Gaussian) will also make a hugh difference.

Masking opens the gates for many other interesting applications for selective refinements in an image. Once you get up the learning curve you will find it is really very easy. Kind of like setting up your imaging train, focusing, getting the exposures just right, ect. ect. ect. :silly:

Well, back to my DSO project . . .


User avatar
Cschur
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 3807
Joined: Mon Oct 17, 2011 6:27 pm
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by Cschur »

Alexandra, those images have tremendos detail, your really on the right track now. no more globby overcooked looking wavelets for you!

I found that a radius of 2 for 2x barlow shots is right, and a radius of 1 for prime focus is best. Any higher, and you start getting globby again. Push the iterations until you see the noise, then use the history pallete to back off a few steps. Its a great tool if you havent used it.


Chris Schur

----------------------------------

Birch Mesa Observatory, Payson Arizona

Clear Creek Canyon Observatory, Happy Jack, Arizona
Ken Crawford

Re: Smart Sharpen and Unsharp Mask

Post by Ken Crawford »

Don't forget to duplicate your layer and apply the corrections to the top layer. Push it just to the point of noice or even a bit farther.

Then dial back the opacity to taste! I always go just a little farther than I should then dial back until I get excatly what I want.

Of course with Smart Objects you even do more :) but thats another story!


User avatar
marktownley
Librarian
Librarian
Posts: 42270
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 20424 times
Been thanked: 10243 times
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by marktownley »

[quote][quote]Wow Mark,

That is interesting! With your 2.5 powermate you have more information to be able to dig out as your images scale gets finer. So the more accurate checkbox should work better for you at that point. You have to be able to get the radius setting setting right on and the blend between the shadows and highlights correct for best results. If it looks over cooked then the setting are not correct between the Amount and Radius. Depending on the sharpening method (Remove Lens Blur or Gaussian) will also make a hugh difference.

Time to dig out my archived hard drive and revisit some of my data from last summer - I will have a play round with the sliders and see what I can do...


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
Bill E.

Re: Smart Sharpen and Unsharp Mask

Post by Bill E. »

Alexandra, Wow they came out great. Congrats :bow:


User avatar
swisswalter
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 17948
Joined: Thu Feb 06, 2014 4:28 am
Location: Switzerland
Been thanked: 1 time
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by swisswalter »

Hi Alexandra

you are doing very fine with the new tools. Masks are puzzling for me too, but they work fantastic. It just needs time and experimenting


Only stardust in the wind, some fine and some less fine scopes, filters and adapters as well. Switzerland 47 N, 9 E, in the heart of EUROPE :)

from 7 am - 7 pm http://www.nanosys.ch

from 7.01 pm - 6.59 am http://www.wastronomiko.com some times vice versa ;)
Derek Klepp
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 12900
Joined: Wed Oct 19, 2011 10:02 am
Been thanked: 171 times

Re: Smart Sharpen and Unsharp Mask

Post by Derek Klepp »

All this is way beyond my comprehension but extremely interesting and Alexandra those pics look sharp. I,m glad I,ve archived my AVIs.
Thanks Derek


User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Re: Smart Sharpen and Unsharp Mask

Post by Montana »

Thank you so much everyone, I just love these results. I have only used the filter twice and I have fallen in love already. The filter is so flat (no sweet spots) and seems to be on band and balanced everywhere and not even a hint of Newton's rings. So it just goes to show that it is the human and their Photoshop skills which are the most crucial.

Ken, thank you so much for your offer, I've never used Skype before. This afternoon I am going to listen to a few more of your tutorials even if they are not relevant to solar, I feel I really must get the grasp of what a mask is and I really need to work on my prominences.

Kind regards
Alexandra


User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Re: Smart Sharpen and Unsharp Mask

Post by Montana »

Mark, I have tried these methods on PST pictures and I have not got good results at all. I always need to do at least a tiny bit of wavelet 6 sharpening first. It seems that the PST doesn't have enough detail for Photoshop to grasp, but saying that it is probably me that is doing it wrong in the first place!

In case this is helpful to anyone (although all photos seem very different) this is what I did (I made notes):


Solarscope processing
2x Barlow
Open avistack file
Image/adjustments/curves - move the middle down a little
Filter/Smart Sharpen – advanced, more accurate, lens blur
Sharpen- radius 2.8, amount 402
Shadow – fade amount 1, tonal width 51, radius 1
Highlight – fade amount 1, tonal width 12, radius 1
2nd round, Smart sharpen, same as above but
Sharpen – radius 0.5, amount 25-50% depending on noise.
Then adjust curves middle and bottom to taste, then brightness and contrast.

Prime focus
Open avistack file
Image/adjustments/curves - move the middle down a little or not depends.
Press the Ha 0.5 Pix sharpen button(I did an action key) (this is 0.5 pixel, threshold 0, amount 25%) 14 x and a couple of Smart Sharpens (first change image to 220 pixel size, then Smart Sharpen of the same using Gaussian blur).


User avatar
Cschur
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 3807
Joined: Mon Oct 17, 2011 6:27 pm
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by Cschur »

Hey Mark, I did the same thing after I dumped wavelets - I have been going back to the first few shots I did with my Lunt100 and have tried them again without wavelets. It has been very interesting to compare the two methods, smart sharpen and iterative USM, and see what the differences are. Between me and Ken, we are going to provide this group with many new methods to try on our solar images. All of us will be producing the best images we can, and THAT is what really gets me excited here, to see everyone improve! Ok everyone, lets see your re processed Halpha images using either method - NO WAVELETS ALLOWED! Show me what youve got...


Chris Schur

----------------------------------

Birch Mesa Observatory, Payson Arizona

Clear Creek Canyon Observatory, Happy Jack, Arizona
User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Re: Smart Sharpen and Unsharp Mask

Post by Montana »

Hi Chris, any chance you could be a little more detailed on what you do to your prominences :) :)

I'm still producing rubbish, unless what I've been playing with is rubbish of course :(

Kind regards
Alexandra


User avatar
Cschur
Way More Fun to Share It!!
Way More Fun to Share It!!
Posts: 3807
Joined: Mon Oct 17, 2011 6:27 pm
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by Cschur »

Hi Alexandra, I do the same thing as I do with the disk, use a radius of 2, and hit the iterations until it looks grainy. then back off with the history pallete to the last good image. Now if you want to use masks to only sharpen the proms and not the sky around it, which is where the noise shows up first, then you must eventually learn how to use masks. You have much more control over what get sharpened when you use a dupicate layer and mask then a global adjustment in smart sharpen. But masks are something you will have to learn from the manual you got with your copy of photoshop. Once youve read that, or a book on masks, then you can try masks on the image. Layers and masks are SOOOOOO important for astronomical image processing. You must master them first before attempting to use them on a solar image. As a first go at it, create a duplicate layer of your image before sharpening. then click the mask icon to creat a blank mask. Sharpen the image to suit. for now, try hand painting the white mask with a black color with the paintbrush to only see the sharpening on the proms and spicules. Then flatten. If this sounds confusing - it will be until you get familiar with the layers and mask concepts. This is something I cant really teach you on line, you will have to break down and read the manual! However, I will try for your sake to put a tutorial together on the basics of prom masking. Until then, play around with it and see what you can do.


Chris Schur

----------------------------------

Birch Mesa Observatory, Payson Arizona

Clear Creek Canyon Observatory, Happy Jack, Arizona
Ken Crawford

Re: Smart Sharpen and Unsharp Mask

Post by Ken Crawford »

Alexandra,

Chris is right, using masks is the best way to approach to coloring selective processes like sharpening, color adjustments, brightness adjustments, ect.

This is a tutorial that uses masks that has helped many DSO imagers. But if you follow along you can see what simple masks are.

http://www.imagingdeepsky.com/Tutorials ... tream.html

Or just Google Photoshop Masking - compositing and there are many tutorials out there that explain masks. I prefer having someone show me how to do it in a movie (like looking over the shoulder) vs reading about it :)


User avatar
marktownley
Librarian
Librarian
Posts: 42270
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 20424 times
Been thanked: 10243 times
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by marktownley »

I've been using the layers, magic wand tool and feather function with my solar images to differentiate processing between the disk and proms (bear in mind I do 'one shots' 99% of the time) and have found this to be quite effective, however I understand the power and importance of masks so this is also on my 'to do' list regards learning it...


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Re: Smart Sharpen and Unsharp Mask

Post by Montana »

Thanks everyone!

Ken, I watched all your tutorials this afternoon and I thought, yes I understand masks now, then tried it myself and forgot everything :S I think I need to watch them all over again. It's so complicated, I need to think how to apply it (thats the hard bit), it's slowly coming though. I definitely find your method of explanation very good (sometimes a bit fast), I need to watch it 10x to get it into my head. I think it also depends on the image, I tried Smart Sharpen on this first image from last November and I was very pleased with the result. However my prominences from last week will not process in any way all I get is no sharpening but strange pixels instead. This is also what I get with my PST images (see second image). Why does this happen? or is it the case of rubbish in rubbish out!




Kind regards
Alexandra


Ken Crawford

Re: Smart Sharpen and Unsharp Mask

Post by Ken Crawford »

Hello Alexandra,

Please send me the 16bit master frame and I will take a look at it.

Send it to: kenrc20 at sbcglobal.net (remplace at with @ and remove spaces).

Ken


User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Re: Smart Sharpen and Unsharp Mask

Post by Montana »

Thanks Ken, I will do so this evening :)

Regards
Alexandra


User avatar
marktownley
Librarian
Librarian
Posts: 42270
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 20424 times
Been thanked: 10243 times
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by marktownley »

Ok everyone, lets see your re processed Halpha images using either method - NO WAVELETS ALLOWED! Show me what youve got...

Next time I get a new batch of imaging data I plan on doing a side by side comparison off all these techniques. I've been playing around with them ever since Chris USM tutorial using my archive data, but is not in the ideal file format I normally work with for master processing. I have to be honest at this stage and say i'm not getting as good results with either the SS or USM method as I am with my own processing workflow - but I do understand the benefits these methods have and as such at this stage i'm fairly certain i've just not hit on the ideal settings for the native image scale and contrast i'm working at. I will persevere!!! I know when I do hit on the ideal settings it will bring a big smile on my face... :)


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
Pedro

Re: Smart Sharpen and Unsharp Mask

Post by Pedro »

I still use wavelets to process my solar images. I find it a must for WL (solar granulation). I also use wavelets for H-alpha images but final processing is always done in PS CS5. As I mentioned before I change the middle value at the wavelet tab (Registax) to 1000.

It helps to read the info available at registax webpage:

http://www.astronomie.be/registax/previewv6-3.html
http://www.astronomie.be/registax/linkedwavelets.html
http://www.astronomie.be/registax/linkedwavelets1.html
http://www.astronomie.be/registax/denoi ... arpen.html

For full disk I use the default option (wavelets) but for HR I always use the gaussian option. Use it with care and process the final image in PS.

best


User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Re: Smart Sharpen and Unsharp Mask

Post by Montana »

Hi Pedro,
I don't understand where you put 1000 in the wavelets tab? any clue where this is?

Regards
Alexandra


Pedro

Re: Smart Sharpen and Unsharp Mask

Post by Pedro »


Hi Alexandra:

See the enclosed pic (wavelets - registax 6).


NandoPG

Re: Smart Sharpen and Unsharp Mask

Post by NandoPG »

Hi Pedro,
Do you use Linked Wavelets with layer denoising and deringing or just the default wavelet filter ?

Regards,

Fernando


Pedro

Re: Smart Sharpen and Unsharp Mask

Post by Pedro »

Hi Fernando:

I only use the wavelets filter. Any denoising can be done in PS. Do not push to hard in Registax.

I guess the pic I sent is self-explanatory (see the result - middle square). I only used the first three layers.

best


Pedro

Re: Smart Sharpen and Unsharp Mask

Post by Pedro »




more on the wavelet tab (Registax 6)

before and after pics are examples of images processed only with Registax

best

Attached files


NandoPG

Re: Smart Sharpen and Unsharp Mask

Post by NandoPG »

Pedro,
I adjusted my Registax 6.0 like you showed and sharpened a solar image getting a result not so good, like the settings were too agressive. Then I reset the image and changed the wavelet filter to 1000 as you indicated. With this adjust, the action of the sliders was much more smooth and the result for the same position of the sliders was good. Is this what you would expect ? I got that right, this 1000 on the wavelet filters desensitize the action of the sliders making them more smooth and thereby more controlable ?

Fernando


Pedro

Re: Smart Sharpen and Unsharp Mask

Post by Pedro »

Hi Fernando:

That's it. Change the center value in the wavelets tab (500 to 1000) and you will get better results and more control over the wavelet layers.

best


User avatar
Montana
Librarian
Librarian
Posts: 34560
Joined: Mon Oct 17, 2011 5:25 pm
Location: Cheshire, UK
Has thanked: 17667 times
Been thanked: 8787 times

Re: Smart Sharpen and Unsharp Mask

Post by Montana »

Thanks Pedro, this looks very interesting, I shall have a go using this setting tonight :)

As an aside, Ken kindly looked at my prominence image and why I couldn't sharpen in PS, it is as I feared, it is the fault of the human behind the steering wheel. It is too over exposed :( oh well, it's a learning curve which is good fun.

Regards
Alexandra


User avatar
marktownley
Librarian
Librarian
Posts: 42270
Joined: Tue Oct 18, 2011 5:27 pm
Location: Brierley Hills, UK
Has thanked: 20424 times
Been thanked: 10243 times
Contact:

Re: Smart Sharpen and Unsharp Mask

Post by marktownley »

Well, it's taken me nearly a week of tweaking and twiddling but i can confidently say that with the raw data I get out of avistack, using the smart sharpen method I am getting reults that is equally comparable to that that I get from Registax. I am hoping with a bit more perseverence on this I might actually be able to improve on it.

Thank you both Chris and Ken for giving me the impetus to challenge my established processing techniques and to push the envelope of solar imaging further still :)


Image
http://brierleyhillsolar.blogspot.co.uk/
Solar images, a collection of all the most up to date live solar data on the web, imaging & processing tutorials - please take a look!
Post Reply