Jeremy Likness Astrophotographer Jeremy Likness has always been interested in the mysteries of deep space, but only recently began to explore them through astrophotography.

📅September 06, 2021 💬674 words ⌛2 minutes to read

Easily Normalize RGB with Auto-Linear Fit

Have you ever opened an image from a recent observation only to find it is washed out and looks purple, pink, red, blue, or green? This is common with my sessions. I used to manually use ChannelExtraction to separate the RGB channels, then LinearFit to normalize the channels, and finally uses LRGBCombination to apply them back to the original image. After doing this about a dozen times, I finally broke down and wrote a script to automate it. I call it Auto-Linear Fit.

I made a short video to explain the script and show it in action.

The breakdown is simple. Here is the main dialog:

Auto-Linear Fit Dialog

The options include:

Use minimum channel This picks the lowest intensity channel as the baseline for normalization. Unchecked this to choose the highest intensity channel instead.
Create new instance By default, the script will apply changes to the original image to preserve history. Check this to leave the original untouched and generate the result in a new instance.
Preserve individual channels The script will close the R, G, and B channels when it finishes unless you check this box.
Reject low/high This is passed to the LinearFit function to handle clipping low and high range values.
Lightness This will change the overall lightness (luminance channel) of the final image. Values below 0.5 will make it lighter, above 0.5 will make it darker.
Saturation Slide below 0.5 to increase the saturation and above 0.5 to reduce it.
Chrominance noise reduction The LRGBCombination can process noise and remove colored blotches and pixels that are introduced by the saturation function.