Calibrate or not to calibrate RGB filters

More information,

I then stretched in the same way I always do the two images.

Top the uncalibrated RGB filter with SPFC and SPCC treatment version

Bottom the caibrated RGB filter with SPFC and SPCC treatment version

Which one looks more natural to you?

regards Rainer

📷 image.png

SonnyE · Nov 28, 2025, 09:11 PM

Hi Rainer,

I happen to like the top picture.

If I had to guess I would probably say the bottom picture, to me, appears to have a bit of light pollution in it.

I’ve signed off on post-processing my images over the years, preferring to use a OSC camera (ASI2600MC Pro) with one of two light pollution filters. After stacking in ASI Studio, I rarely do any more to my images. Often not even cropping out the stacking artifacts.

I began with OSC cameras, evolved to using an ASI1600MM Pro (EFW and ZWO supplied filter set. The “bundle” they used to offer). But I reverted back to a OSC ASI2600MC Pro and only LP filters. One an Antlia Quad, and now a Svbony SV220 3nm duo LP filter. Which is giving me brighter reds than I personally like.

I did run some test on my ZWO filters once out of curiosity and the seven provided filters (1.25”) were within 10 on a 1000 scale. So I decided they were probably good enough for a hack like me.

But based on what I see in the two images, I like the upper one the best.

Call me a Post-Processing Minimalist. 🤔

Hi and thanks,

Interesting.

The upper image has a lot more blue in it then the bottom one.

What you call light pollution is just IMHO a bit brighter and it is solvable. I wanted to make a screenshot of the RGB values but the readout window does not stay open when trying to choose the screenshot crop. So there is no way to show the RGB value differences.

Below a crop of this image using the RGB filter calibration

📷 image.png

Rainer Ehlert · Nov 28, 2025, 09:53 PM

SonnyE · Nov 28, 2025, 09:11 PM

Hi Rainer,

I happen to like the top picture.

If I had to guess I would probably say the bottom picture, to me, appears to have a bit of light pollution in it.

I’ve signed off on post-processing my images over the years, preferring to use a OSC camera (ASI2600MC Pro) with one of two light pollution filters. After stacking in ASI Studio, I rarely do any more to my images. Often not even cropping out the stacking artifacts.

I began with OSC cameras, evolved to using an ASI1600MM Pro (EFW and ZWO supplied filter set. The “bundle” they used to offer). But I reverted back to a OSC ASI2600MC Pro and only LP filters. One an Antlia Quad, and now a Svbony SV220 3nm duo LP filter. Which is giving me brighter reds than I personally like.

I did run some test on my ZWO filters once out of curiosity and the seven provided filters (1.25”) were within 10 on a 1000 scale. So I decided they were probably good enough for a hack like me.

But based on what I see in the two images, I like the upper one the best.

Call me a Post-Processing Minimalist. 🤔

Hi and thanks,

Interesting.

The upper image has a lot more blue in it then the bottom one.

What you call light pollution is just IMHO a bit brighter and it is solvable. I wanted to make a screenshot of the RGB values but the readout window does not stay open when trying to choose the screenshot crop. So there is no way to show the RGB value differences.

Below a crop of this image using the RGB filter calibration

📷 image.png

When I first got my newest camera a couple of years ago (ASI2600MC Pro) (a OSC) I tried it for a while with not filters at all. But I live in one of those higher Bortel areas, I say between 7 and 9 depending on what neighbors have left their outdoor lights on. 🤔

My unfiltered images displayed the typical dusting of LP interference. So that was my experience with infringed images. After installing an Antlia Quad LP filter, I’ve been really pleased with my out-of-the-camera images. Until recently.

Cuiv put up a page <link< on using an Svbony 3nm narrowband in the LP of Tokyo, Japan where he is and raved about the LP improvements to his images. So, I’m in the beginnings of doing Narrowband LP filtering to see what it might do for me with a same filter. So Far, So Good. Though a bit reddish for my tastes. But I’ve learned others seem to like bright images, brighter than I do.

My post was simply based on my observation, from my little corner. I hope no offence was incurred. None was intended.

Incidentally, I tried Mono with my ASI1600MM Pro for 3 ½ years, but reverted to my love of OSC cameras with my ASI2600MC Pro a bit over 2 years ago. I’m a Post-Processing Minimalist.

No offence was intended. 😉

John Hayes · Nov 29, 2025, 07:02 PM

Rainer Ehlert · Nov 28, 2025, 04:47 PM

Nowadays having PI with SPFC and SPCC my question is

¿does it still make sense to run a RGB filter calibration? Look at the screenshots below…

Thanks Rainer

No…it does not make sense. If you use SPCC properly, you can achieve correct color calibration. Your example images don’t look right. The background is clearly not neutralized in the top image and that’s a red flag. On my screen, it appears to have a bluish cast. Remember that the readout window is only showing you the values of the central pixel in the window; not the average value of the pixels in the window. To achieve proper color balance, you need to achieve a neutral average background and the desired white balance, which is a mathematical transformation of the RGB values from wherever they may start—either calibrated or not.

John

Thanks John,

I have set the read out options like following and so it should not be only one pixel but from 225 pixels.

📷 image.pngquote:” The background is clearly not neutralized in the top image and that’s a red flag.”

Both images have no processing at all except Plate Solve after LRGB combination and then SPFC and SPCC. No stretching at all be it GHC or HistogramTransformation so the histogram I presented is from the linear image.

It is interesting you mentioned that the upper image has not neutralized background and nothing mentioned about background from the bottom image which has also not a neutralized background at all.

In the same way as you do I see a bluish cast on the top image which is ¿nonexistant in the bottom image? which was made with LRGB combination and RGB filter calibration applied.

quote:”If you use SPCC properly, you can achieve correct color calibration.”

I do LRGB combination, Plate Solve and then apply SPFC and then SPCC. ¿is this the correct way? Thanks…

regards Rainer

John Hayes · Nov 29, 2025, 08:25 PM

Rainer Ehlert · Nov 29, 2025, 07:29 PM

Both images have no processing at all except Plate Solve after LRGB combination and then SPFC and SPCC. No stretching at all be it GHC or HistogramTransformation so the histogram I presented is from the linear image.

It is interesting you mentioned that the upper image has not neutralized background and nothing mentioned about background from the bottom image which has also not a neutralized background at all.

In the same way as you do I see a bluish cast on the top image which is ¿nonexistant in the bottom image? which was made with LRGB combination and RGB filter calibration applied.

quote:”If you use SPCC properly, you can achieve correct color calibration.”

I do LRGB combination, Plate Solve and then apply SPFC and then SPCC. ¿is this the correct way? Thanks…

regards Rainer

Hi Rainer,

Simply running SPCC on a LRGB image does not necessarily produce a color calibrated image—as you have discovered.

The proper way to color calibrate is to first create a RGB plate solved image. Then remove any gradients. There are a lot of tools for that job. Then perform SPCC using the appropriate white balance. In most cases the best white balance with be the “Average Spirial Galaxy”, which is the default. If the background comes out with a blue-ish (or any non-neutral) background, the background limits in SPCC need to be adjusted. If in doubt, you can tell if the background is right by running BackgroundNeutralization. If the background tone changes, there’s a problem with the settings in one of those two tools. When the background is properly neutralized, you should see very close to R=B=G for empty space no matter where you sample the background and it should appear to be a neutral gray. That’s what your second image looked like on my monitor and why I didn’t mention it. Once your RGB image has been color calibrated, stretch it. If you’ve done everything correctly, it should look like a spectacular image at this point. The background should be neutral and the color vibrant with a proper white balance at the high end.

Now it’s time to combine it with your Lum signal. The Lum signal should have gradients removed as well and it should be stretched with the same parameters as your RGB image. You also need to roughly match the brightness and dynamic range of the Lum image with the brightness of the RGB image so you may need to use HDRMT before doing the combination. It is very important to adjust the Lum signal so that it does not produce saturation at any point in the image. Adding the Lum signal should help to bring out additional detail and show much fainter stars than what the RGB data shows. I go over all this stuff in my GalaxyStudio presentation on TAIC so it may be helpful to review it there if you haven’t already.

My point is that in order to get proper color calibration, you do not have to calibrate your filters, which was your original question. The procedure you’ve described strikes me as being very similar to using the LinearFit tool to match RGB color channels, which again, is not necessary if you are using SPCC to do color calibration.

John

Hi John,

Thanks a lot for this explanation 😎