Absolutely Stunning image you got there Bill awesome work.
@anata5127
9 ай бұрын
I know that you have originally been trained to do RGB processing, stretching and then star removal. Sometimes, it is better to stretch R, G and B separately. Additionally, I came to conclusion that stars should be extracted before stretching; and then stretching of nebulosity and stars could be independent. Much more effective.
@AstroVagabond
9 ай бұрын
Thank you for the feeback. I do remove the stars while the image is in a linear state, work the starless image and then rescreen the stars after applying a stretch to the stars.
@JonnyBravo0311
9 ай бұрын
Your processing has gotten quite good, Bill. My own tastes lean towards how it looks right around the 9:27 mark. The final product is a bit too red saturated for me. Again, just my own personal take on the image - nothing wrong with your representation at all. I'm curious to know what your pixel math formulas are for combining the luminance back with the RGB as well as combining your stars back with the starless.
@anata5127
9 ай бұрын
Me too.
@AstroVagabond
9 ай бұрын
I agree Jonny a bit saturated. When I blended the Luminance with the RGB simple Pixel Math "RGB + Lum" it really lightened the image. So to dial it back a bit I used RGB + Lum*.5. The image at 9:27 had the curves process applied with a bump in saturation. Things went south when I blended in the Luminance. I'm not confident yet on that process so it will be my focus when I do my next training sesson with Ron Brecher in January after my new moon trip. I focus on L+RGB as well as Ha+RGB techniques and then reprocess this image and post an update on what I did specifically in PixInsight. As to adding stars back in I stretch the stars being they were removed when the image was in its linear state. How I stretch lets me reduce the number and prominance of the stars. Once I have the stretch I want I use HistogramTransformation to apply the stretch. The I rescreen the stars using Pixel Math : combine($T,RGB_stars,op_screen())
@AstroVagabond
9 ай бұрын
See my reply to JonnyBravo. 👍
@JonnyBravo0311
9 ай бұрын
@@AstroVagabond the pixel math you use for recombining stars and starless is the same one I use. Just wanted to make sure you weren't using simple addition because that can cause issues and lead to clipped stars. I typically use the LRGBCombination process (Process -> ColorSpaces -> LRGBCombination) to combine the luminance and RGB data. That process allows you to choose the impact of the L on the RGB as well as do color saturation and color noise reduction. You stretch both your RGB and L images. Then you fire up the LRGBCombination process. Put a "check" next to the L and add the L image identifier to the text box. Then, uncheck the R, G and B. Now look at the "Transfer Functions". By default the Lightness and Saturation are 0.5. Play around with those sliders. Conversely to what you may think, the LOWER the value you put, the more aggressive the application of the function. You can also use the LinearFit process to match your Lum image with the extracted L of the RGB data. Same kind of thing as above... stretch both your RGB and Lum images. Once you're happy, extract the L component from your RGB image using ChannelExtraction (Process -> Color Spaces -> ChannelExtraction) in CIE L*a*b* mode. Next, open up LinearFit (Process -> ColorCalibration -> LinearFit). Put your stretched Lum as the reference image, then apply to the extracted L. Now, recombine using ChannelCombination (Process -> Color Spaces -> ChannelCombination) in CIE L*a*b* mode. Doing this ensures the flux is very nicely matched between your stretched L and stretched RGB images. Now you use the LRGBCombination tool as I described. By definition, adding luminance data to an RGB image will wash out the color. It is unavoidable - which is why the LRGBCombination tool has the saturation adjustment slider built in. Of course, you can just leave it at the default value of 0.5 and use the ColorSaturation (Process -> IntensityTransformation -> ColorSaturation) or CurvesTransformation (Process -> IntensityTransformation -> CurvesTransformation) to taste once you've created the combined LRGB image.
@AstroVagabond
9 ай бұрын
Thanks for sharing the flow. I first tried LRGB and got a really saturated result. I did not explore process enough to test the slider positions. Thanks to your information I'll use that technique / process on my Iris Nebula where I need to combine the L data. Much appreciated. 👍
Пікірлер: 10