[Sticky] Background Neutralization versus Background Calibration - getting a correct sky background and colors - ASI2600 ZWO RGB Multiplier Levels

Mm, sorry I have no experience with that one. I found this answer about the multipliers in general by Mabula. Why don't you like the background neutralization?;

My advise would be to never use the RGB multipliers actually, the option is there though so you can experiment and other applications do offer it as well. I never do myself with Canon, Nikon Dslrs or color cmos camera;s like yours. Let me explain why:

If you start to use a multiplier higher than 1, you immediately start to throw away data that is on the right side of the histogram (so the highlights) giving you quicker burnt out stars at a stage in processing where it's not needed at all.

Furthermore, the color of the initial light frames, can be all sorts of color and it just depends on a lot of factors actually. I have seen all sorts of colors and all can be easily corrected after integration of all the frames.

Some factors that play a role:

• moon phase has a big influence
• sensor response in R,G & B broadband wavelengths (which is different for different camera models)
• your local light pollution
• the quality and transmission of the filters that are used ( in this case the CFA filters on your sensor)

In a nutshell, a lot of factors determine what color your sensor actually sees for the night sky. That's no problem and perfectly normal. A thing like chromatic abberation doesn't play a role in this sense (and APP can correct for that as well by the way).

After integration, you should correct the data for gradients and then perform background calibration. Then the sky background is corrected in a scientifically correct manner. Star Colors are a different subject but need to be corrected using the star color calibration tool after background calibration since it needs to know the background levels.

So I recommend to use the linear data as it is recorderd by the sensor, the colors can and should be corrected after integration I think.

Posted by: @michelaudette

Hi;

I got got the ASI2600.  Anyone knows what the RGB Multiplier Levels should be.  I dont like to neutralize the Background...

 

Michel

Dear Michel @michelaudette and @vincent-mod,

Setting R,G,B multipliers is like setting a white balance. Which factors you would really need for your camera would depend on the circumstances in which the light frames are shot. There is no such thing as fixed multipliers for a certain camera model if you are looking to set white balance correctly.

Please realise that with regard to Background Neutralization that the background neutralization will do a far better job on setting an initial white balance for your data then those RGB multipliers that you are looking for. And please also realise that Background Neutralization is NOT the preferred way to get a good sky background and good colors. It is just a simple start to better see what signal there is your data.

To get good colors, whether you use Background Neutralization or not in 5) Normalize, you need to perform Background Calibration which is fundamentally different than Background Neutralization.

Background Neutralization - is setting a white balance for the sky (making it gray) based on all data of your image, so on the entire histogram. That can never be very accurate and will even visually surpress faint nebulosity. Please take note though, the faint nebulosity is not removed, it is still there in the data, only it is hard to see visually due to the operation based on the entire histogram of the image. Background neutralization is only for seeing what is there quickly, it is NOT to be used as a final step to get a good background and good colors. But, and this is very important, the Background Neutralization step is non-destructive. Let that sink in for a minute now...

Background Calibration - is setting a white balance for the sky (making it gray) based on selected parts of your data that you choose in the Background Calibration tool. This is THE method to get a good sky background and is also required to get good (star) colors. You can only get good colors if the background is correct first 😉 .

Now whether you use Background Neutralization or not in 5) Normalize, you will always see that after Background Calibration the result will be identical if you choose the same parts of your data in the Background Calibration tool, proving immediately that the Background Neutralization is a non-desctructive oepration as well. So it is perfectly save to have background neutralization enabled in 5) Normalize and it will lead to the same result for sky background and the star colors if you use a proper workflow of Background Calibration and Star Color Calibration on your data.

Kind regards,

Mabula

@vincent-mod

Hi Vincent; 

Thank you for the explanation, makes sense to me now.

Cheers

Michel

@mabula-admin

Hi Mabula;

Thank you for taking the time to answer.  Learning a lot!

So if I get this right, the "Neutralize BG" on the right side (The post process side) should be checked off and Background Calibration should be performed (Process 9)?  Also, I assume that when selecting 4 areas, that I should select background areas free of stars?

Lastly, I have a tendency to use "Remove light Pollution", which I understand also performs a Background calibration.  There too I try to select backgrounds without stars but do select area of light pollution and area more in the centre (less affected) by light pollution.  Do you use it?

One important thing I did notice is that my precious signal levels goes way down. (Even worst in Pixinsight) when I perfect either of these tasks.  Is there a way to minimize this?

Michel

Posted by: @michelaudette

So if I get this right, the "Neutralize BG" on the right side (The post process side) should be checked off and Background Calibration should be performed (Process 9)? 

I think that's better when correcting with the tool yes.

Also, I assume that when selecting 4 areas, that I should select background areas free of stars?

No, not necessary as APP can detect the stars just fine. Just overstretch the image so you don't select nebulosity.

Lastly, I have a tendency to use "Remove light Pollution", which I understand also performs a Background calibration.  There too I try to select backgrounds without stars but do select area of light pollution and area more in the centre (less affected) by light pollution.  Do you use it?

Always a good idea to do yes, also there it's not needed to select patches without stars. Just don't select nebulosity areas (if possible).

One important thing I did notice is that my precious signal levels goes way down. (Even worst in Pixinsight) when I perfect either of these tasks.  Is there a way to minimize this?

We should have a look at that in more detail then I think.

Cheers, Vincent

Thank you Vincent.  You are a star.  I will gather some example as far as data lost.