Image turns completely green after star calibration

Hi Eva @evas,

Thank you very much for reporting this.

Did you use any filter while shooting the data? Like an Light Pollution Surpression filter?

The green tint normally can be corrected easily with the green/magenta slider in the star calibration tool, did you try that? Simply move the slider to magenta to reduce the green cast. The green cast can be caused simply by a filter or the actual camera sensor and it's cfa filters attached to the sensor.

Perhaps you can show a screenshot of  the star color plots while performing star color calibration? Does the star population fit well with the model line in the plots?

Mabula

Hello Mabula,

Thank you for your answer!

I tried using the slider, but the results weren't great. I am not using any kinds of filters.

Sadly, I am not sure what exactly you mean with "Does the star population fit well with the model line in the plots?"

Maybe you can judge this from the screenshot:

 So far, the only tolerable solution for me is not using star calibration, and then using a Photoshop Plugin called "HLVG" to remove the green tint, but obviously, I would rather be able to solve this otherwise.

Best regards

Hi Eva,

Thank you very much for the screenshot, that helps a lot !

Looking at it, it is clear that your sky background is rather noisy and has both green and magenta patches... I think it will be better to use the HSL Selective Color tool to fix that first.

Start the HSL Selective Color tool on this image and try to remove the noise in the sky background like this:

Select from the color dropdown box all colors, and for data range select 0-background. Set SAT % to -100 and click on calculate. Try to stretch strongly and saturate too strong with the purpose to clearly see that noise in the sky background, like in the following screenshot:

Notice on the right side the HSL selective color settings. On the right side is the preview filter where i use the strongest stretch from the preset autostrethes with 30% BG and where I set saturation to the max and set the SAT TH to 0, this ensures the skybackground is showing saturated as well. Please note, that this is only a stretched preview now, the data is not yet altered. The settings of the HSL Selective Color tool are working on the not stretched /linear original data. And the preview filter is then stretching that result with it's settings.

If the range setting of 0-background does too little on the green/magenta patched in your sky background, increase that range to 0-background + 1*noise or even more to get more effect.

Once you get a better result, save the result. Open the Calibrate Background tool and perform it on the image. Then load that result into the Star Calibration Tool.

From you screenshot, I do not think that you are using the green/magenta slider, or if you do, you need to use it more aggresively.

"Does the star population fit well with the model line in the plots?"

The model line is the green line visible in the plots/charts. It coincides with the default model which is a gray line. If you would use the green/magenta slider, the green slider would not coincide anymore with that gray line. Inn your case, the green model line would be moved lower and to the right, which means to Blue and Red if you move the green/magenta slider to magenta and then recalculate the calibration.

Let me know if this helps 😉

Mabula

Hello Mabula,

Thanks a lot for the detailled explanation!

It seems like I am doing something wrong, however.. I tried to follow your steps exactly, but it just gets worse, sadly.

This here is how it looks with 0 - background:

Then +1* noise:

And +2* noise:

How is the result supposed to look like?

Best regards

@evas 

To me it looks like it is improving, the green and magenta patches are greatly reduced. Please note, that you are overstretching and oversaturating here just to see your data problem. It is not how you would publish your image of course. Since we are removing color noise in the sky background, Astro Pixel Processor is also automatically stretching stronger giving you the illusion that the problem grows where in fact it is the opposite. Because the auto stretch is increased, you see more, but the correction is made on the linear data and that is our purpose.

Now that +2* noise, would first need to go to the Remove Light Pollution Tool, same story there, keep overstretching and oversaturating and fix the remaining gradients in the sky 😉

Then proceed to Star Color Calibration...

Mabula

I truly appreciate your help, but I just can't seem to do it. 🙁

The green stars are still there after the HSL, just that the saturation of the green is lowered.

When I go to calibrate star colors, it will still make everything green again, and honestly, the sliders are extremely confusing.

There are 2 magenta-green sliders, and I don't know which one I need to use. I tested out all sliders yesterday for hours, but with no success.

Especially, since there are also 3 options from the dropdown to choose from, and each option does different things as well.

Also, changing the colors also changes the colors from the object. In my opinion, the galaxy itself has the correct colors already, merely the stars are just entirely green.

In another image I took, I have green stars on the right, and magenta stars on the left, while the galaxy itself also has the correct colors.

Following your steps with HSL there just desaturated the entire image here, which is of course not my goal, so I skipped this here as well.

The (or one of the 2) magenta-green slider in CSC makes it a tiny bit better, reducing the general green tint of the image, but the wrong star colors themselves stay the same.

Also, bright stars seem to be displayed in the correct color, only the dim stars have the green / magenta tint.

I know I can improve the background noise in the future, and it is reducable with Photoshop as well, so this is not my primary concern.

Merely the star colors are really driving me crazy. :/

Best regards

Hi Eva @evas,

Okay, maybe it is a good idea that you share your integrated result and I will have a look at it?

I can demonstrate what I would do with your data with explanation.

There is the possibility as well that your magenta/green tint on the stars is caused by optical aberrations and/or that your data suffers from strong light pollution complicating things perhaps. Both can be corrected with APP and I am happy to demonstrate it.

If you want me to do this, please upload your integrated/stacked result here:

https://upload.astropixelprocessor.com/

username: uploadData

password: uploadTestData

Please upload it in the folder : Eva-PostProcessing-Stack

and let me know once uploaded 😉

Mabula