15th Feb 2024: Astro Pixel Processor 2.0.0-beta29 released - macOS native File Chooser, macOS CMD-Q fixed, read-only Fits on network fixed and other bug fixes
7th December 2023: added payment option Alipay to purchase Astro Pixel Processor from China, Hong Kong, Macau, Taiwan, Korea, Japan and other countries where Alipay is used.
Hello,
After I finish with the integrations, I like to go to each of the color channels individually and remove light pollution for an even background.
Now before I combine them into HaRGB, I want to actually SUBTRACT out some of the background until it is just stars on a sea of black. Doing it this way lets me treat the background of each of the channels differently, for example, so I can preserve nebulosity in the Ha channel.
But first, I have a more immediate problem. If I just load or look at the green channel itself in the combine RGB tool, there is a spike in the histogram of purple. Where did that come from? If I use the bg slider under the combine rgb tool (I assume this means background?), then it makes even more purple if I try to drop the background. Is purple something like negative green??
Ideally, this is what I would like to do:
R,G,B: subtract light pollution and background separately, then combine and stretch. Tweak star colors if necessary.
Ha: separate light pollution and background subtraction, use a separate stretch
Now combine all signals.
Is this possible?
The histogram indicates that you are working on a RGB/mulit-channel image. If you want to do remove light pollution on the individual channels, simply split the channels first with split channels in 2) Calibrate ;-).
Then load the single channels into the Remove Light Pollution tool.
Afterwards, you can recombine all channels into a composite in the Combine RGB tool using the tools normalization setting (or not).
How you would get a green histogram with the purple spike is a bit of a mystery to me, what did you do with this image or data? Did you use the RGB multipliers perhaps in 0) Raw/FITS, to only show green ? Or did you kill Red and Blue in the combine tool ?
I would like to add that in my opinion performing Light pollution correction in the single channels of a RGB image is not the best nor efficient workflow. I always get faster and better results if I perform Light Pollution correction on the RGB data, because it will be much easier to see what you need and want to correct using overstretching and oversaturating the data while performing the correction 😉
And in my experience, if you treat the channels separately, and then combine them to RGB or HaRGB, you will again need to do correction because you will see that you are not there yet...
Background subtraction will simply be done with the black slider with or without a DDP auto stretch 😉 Be carefull that you don't clip to much pixels though on the black point.
Please let us know if this helps.
Mabula
Hello,
Thanks for the reply.
I get the same purple spike in the histogram, and purple color in the image if I load one of my raw light frames straight into the RGB combine tool while assigning it to green. (red frame has cyan spike, blue frame has dark yellow spike) If I view the raw frame outside the tool, then of course it is monochrome.
Regarding background subtraction with the slider to the right, the problem is that the Ha frames has a different level of background than the RGB frames. By the time I make the background of the star field (almost) black, the finer and extended features of the Ha nebula are gone. Yet when I look at the Ha signal by itself (in monochrome) there is a lot of detail. That's why I want to process these two sets of data (RGB and Ha) separately.
Can somebody confirm the problem?
Dear @bokemon,
If you only assign data to green, then it will be green in the RGB Combine tool (there is no way around that). The magenta spike in the histogram is becuase there is no data for red and blue which makes magenta.
I suggest that you first create a good RGB image without the H-alpha data using
- the Remove light pollution tool on the RGB data.
- and star color colibration on that corrected data.
Then you want to load the star color calibrated data into the RGB Combine tool (using no normalizaion in the tool to keep good star colors) and add the Ha-layer to red, probably mix it with green and blue a little bit as well and use the Ha partly as luminance as well 😉 Then it should work really well.
If you want me to demonstrate how to do this in a proper workflow in APP, simply share your RGB and H-a data and I will give it a go 😉 and I will describe how i do it.
Mabula
Thanks for the reply and sorry for getting back late. I still could not get the desired background subtraction to do what I want, since the bg slider inside the rgb combine tool can't make a background black, but instead maps it to the opposite color. And then the background slider on the right is for the overall image after everything is combined. In short, I want to treat the RGB background different than the Ha background because there is some faint nebulosity that I want to bring out, against a star background that is mostly black. I guess this might also mean a different stretch for the RGB vs Ha data.
I will very much appreciate you show how to do this. How do you want me to give you the files? each integrated file is 64 MB.
I attached an image which almost shows what I want, namely on the Rosette nebula there are faint branches to go up and to the right. I had to make a tiff file and edit it in some other image processing program where I can adjust the background for each of the colors separately. And somewhere along the way the star colors got a bit messed up (looking all white).
