How do you handle light-pollution and gradients?
Today I'd like to share one of my powerful methods in my post-processing workflow that I think will be super useful in your own astro workflow.
It is possible that you are already familiar with this method or already using it, but I thought it could be helpful for someone who does not know of this.
One of the biggest challenges of post-processing is eliminating the light-pollution and gradients, making the processing a lot harder.
One of the approaches I want to show today to handle this problem uses a method called frequency separation. It is a process that is frequently used (pun intended :) ) in portrait photography for skin retouching.
The basic idea behind this is to separate the rapidly changing brightness values from the slowly changing ones. In specific scenarios the three major frequencies on your image are the fine details like stars, the middle-sized details like nebulae and the very large entire image sized gradients like the light-pollution.
If there would be a solution to separate these into different layers, you could change their appearances separately.
If your image is suitable, which means these three things are not overlapping or at least not much, you can separate those with this technique.
How does it works?
So you can separate the low frequencies by bluring the image.
If you imagine what a blured image is, it is an image which doesn't contain the fine, rapidly changing detail any more.
Subtracting this image from the original will get you the image containing only the fine details (only the rapidly changing detail)
As a further step you can do it multiple times creating the three layers with different targets.
By creating this in Photoshop, you can duplicate the target layer and then blur the bottom one.
Then, choose the copied layer and click Image -> Apply image. In the configurations, choose "subtract for blending" and select the blurred layer from the dropdown.
If you choose subtract a scale and offset option will show up select 2 for scale and 128 for offset. This will shift the content of the layer so it will be normalized between -128 and 128.
After this is done you can select the Linear light for this grey looking layer.
The two layers, the blurred one and this grey one, produce the same image that you began with, but you can now place any manipulation between the two.
On the attached images, you can see how the three layers together produce the same image as before, but allowing you to change each separate component.
The gauss blur kernel is not actually truly equivalent to real frequency separation, but with large blur values, it produces a good approximation.
I also prepared a photoshop action that you can download and it will create the layers for you. You can download it from here:
https://drive.google.com/file/d/1Y5HBRAtkUWBoJVuqNTX_NwG-fX5PuDeb/view?usp=sharing
(Not perfect it places the new layers outside of the group, sorry for that)
This blog post was originally published in our Telescope Live Community.
The Community represents Telescope Live's virtual living room, where people exchange ideas and questions around astrophotography and astronomy.
Join the conversation now to find out more about astrophotography and to improve your observation and post-processing skills!