Exploiting Photshop's Equalize function.

[ollypenrice]

The Ps Equalize command is found in Image-Adjustments near the bottom. It produces a rendition which has hugely exaggerated contrasts and, in the past, I've used it only to check for faint signal which a normal stretch might miss or for defects like faint residual dust bunnies or visible joints in mosaics. However, this is something I concocted for our recent IC348 to 1333 dusty mosaic. I've never felt I was very good at extracting dusty contrasts so I tried doing it as follows. (This is a quick demo of the procedure, not a careful attempt on a 'real' processing job. Images are screen grabs.)

Basic image fairly fully stretched:

Copy of above with Equalize applied. (Don't panic!)

In Layers, open original, make a copy layer and add a layer mask...

Copy and paste the Equalized version onto the layer mask and give it a good Gaussian Blur. (You can also tweak the mask in other ways, such as with Curves to add or reduce contrast etc.)

Go to Window-Arrange in the top toolbar and ask for a new window for the original image. Now simply stretch the top layer through the mask...

You can use Levels or Curves for this final extra stretch, which will probably be only a very mild one.

Flatten and you're done. My 'careful' version is here: 

 

I'm sure others will have done this before but I hadn't come across it and you might find it helpful on some targets.

Clear skies,

Olly

 

 

 

Edited October 7, 2019 by ollypenrice
Typo

[Davey-T]

Thanks for that Olly, I always bookmark or save your PS tips, my main problem in SE London in actually capturing the dusty bit's to start with 

Dave

[Filroden]

Nice method to quickly enhance contrast. The PS method is certainly simple to understand and control.

For PI users, there is something similar in the LocalHistogramEqualisation. It doesn't need the mask and has a live preview option. Don't ask me what parameters to use - as ever with PI, trial and error! From the manual:

"This process implements local histogram equalization with configurable limitation of maximum contrast enhancement. It is based on the CLAHE method (Contrast-Limited Adaptive Histogram Equalization). The main purpose of the process is to enhance local contrast and visibility of structures in low-contrast regions of the image. The process is designed to run on non-linear (already stretched) images. Histogram equalization takes the histogram and computes a transfer curve, which grants more brightness range to higher histogram peaks and less brightness range to histogram valleys. In other words, large areas of similar brightness get more contrast. Local histogram equalization works on individual pixels and computes a transfer curve from the histogram of a pixel neighborhood. The classical histogram equalization algorithm has the drawback of giving most contrast range to high narrow peaks, like a uniform noisy background. This problem is solved with the contrast limit property of the CLAHE method. This parameter limits the maximum slope of the transfer curve and prevents narrow peaks from getting too much contrast, effectively reducing noise promotion."

[carastro]

Many thanks for that Olly, copied and saved into my useful tips folder.

Carole 

[ollypenrice]

53 minutes ago, Filroden said:

Nice method to quickly enhance contrast. The PS method is certainly simple to understand and control.

For PI users, there is something similar in the LocalHistogramEqualisation. It doesn't need the mask and has a live preview option. Don't ask me what parameters to use - as ever with PI, trial and error! From the manual:

"This process implements local histogram equalization with configurable limitation of maximum contrast enhancement. It is based on the CLAHE method (Contrast-Limited Adaptive Histogram Equalization). The main purpose of the process is to enhance local contrast and visibility of structures in low-contrast regions of the image. The process is designed to run on non-linear (already stretched) images. Histogram equalization takes the histogram and computes a transfer curve, which grants more brightness range to higher histogram peaks and less brightness range to histogram valleys. In other words, large areas of similar brightness get more contrast. Local histogram equalization works on individual pixels and computes a transfer curve from the histogram of a pixel neighborhood. The classical histogram equalization algorithm has the drawback of giving most contrast range to high narrow peaks, like a uniform noisy background. This problem is solved with the contrast limit property of the CLAHE method. This parameter limits the maximum slope of the transfer curve and prevents narrow peaks from getting too much contrast, effectively reducing noise promotion."

 

Yes, I do use PI's LHE (usually taken into Ps as a layer) and suspect that it must exploit similar underlying principles. I also ran Noel's Enhance Local Contrast on the original image before using the Equalized mask. However, the Equalized Mask method did more for me than either of them this time. How often that would be the case I don't know.

Olly

[Stub Mandrel]

This is good. I've often seen detail in a locally equalised image that I haven't been able to 'layer' in without spoiling the overall image.

I've got a problem though - I can't past into the layer mask for some reason? Even with the mask selected pasting creates a new layer.

How do I do this?

Edited October 7, 2019 by Stub Mandrel

[ollypenrice]

1 hour ago, Stub Mandrel said:

This is good. I've often seen detail in a locally equalised image that I haven't been able to 'layer' in without spoiling the overall image.

I've got a problem though - I can't past into the layer mask for some reason? Even with the mask selected pasting creates a new layer.

How do I do this?

It's slightly obscure! At this stage...

... have your spare copy of the image selected (Ctrl A Ctrl C) then place the cursor over the blank layer mask symbol in the Layers menu and Alt-Click. This will throw up a blank layer mask screen and Ctrl V will paste the selected image onto it. It will appear in greyscale, which is what you want.

BTW, a difference between this technique and other contrast enhancements is that this one is lighten-only. It pins the darkest parts where they are but doesn't darken them.

Olly

[carastro]

I just tried it on my Dragons image.  Could not get the method to work exactly how Olly described it, (possibly because I didn't quite understand all the steps), BUT, pinching some of the idea, I managed to greatly enhance my Dragons image.  See relevant post.  Considering it is only 2 hours of Ha taken at Kelling and RGB taken from Bortle 8 also only a small amount I am very please at this result. 

Carole 

 

Edited October 9, 2019 by carastro