Background Neutralization and Color Calibration

[msmythers]

My thought was you need a representative background color sample from the closest neutral area to your main image. Large scale background color does not change with exposure time only signal to noise. Is noise an issue, maybe but certainly not a driving factor. Your not adding this image into yours. Your only sampling the neutral background in that image to set your background in yours.


Mike

[rdryfoos]

My thought was you need a representative background color sample from the closest neutral area to your main image. Large scale background color does not change with exposure time only signal to noise. Is noise an issue, maybe but certainly not a driving factor. Your not adding this image into yours. Your only sampling the neutral background in that image to set your background in yours.


Mike

  But it still requires 3 images from 3 filters and a full on channel combination (calibration?).  I have never heard of anyone doing this.  It seems a rather rare, unusual thing to resort to.  My original question was what do most people do.  People with small APOs or widefield setups (hyperstar, Epsilons, f3-5 astrographs) don't have this concern--but what about all the images of Nebula taken with long focal lengths?  It is hard for me to believe that I have stumbled upon a unique problem (though that is one of my fortes)

[msmythers]

I don't know what others do. I was just responding based on rob's suggestion and what I would try. I shoot extreme wide field from a Bortle 7/8 zone in a hot humid environment so I have all types of other color problems to deal with.

I hope you find a good working solution.


Mike

[Geoff]

The background area can be quite small--just a few pixels in width.  I recently took a pic of M8 and there is nebulosity almost everywhere, but there  are a few dark globules that can be pressed into service. I've never yet come across a situation where there isn't at least one tiny bit of blackness somewhere.
Geoff

[rdryfoos]

The background area can be quite small--just a few pixels in width.  I recently took a pic of M8 and there is nebulosity almost everywhere, but there  are a few dark globules that can be pressed into service. I've never yet come across a situation where there isn't at least one tiny bit of blackness somewhere.
Geoff

  I've always found that dark clouds are never as "black" upon zoom as they appear widefield, especially before histogram/curve manipulation.  I assume the software takes the average of the pixels in the preview?  (If anyone knows please chime in).  So what would happen if the preview was small like you suggest--a few pixels, and 1 or 2 of those pixels were  cold pixels?  Better yet, what if the preview was 1 pixel and it was a cold pixel?

[jkmorse]

Rodd,

In answer to "what do I do?" it's pretty simple.  I use the tools that work for a given image.  In the case of an image full of nebulosity I simply don't use BN any more than I use Dynamic Background Extraction since you need a clean background to be able to use that effectively as well.  In the latter case, I will test using ABE which works surprisingly well with lots of nebulosity and otherwise make sure I have a top notch flat master for each color to eliminate light variations that way.  With background neutralization, the same applies.  PI has a huge toolkit to work with but sometimes, no matter how much you like it and how much it works on other problems, a tool just doesn't work for what you are trying to fix.  I love pounding on things with a hammer but rarely use one when fixing a leaky faucet.

In cases like yours, I simply use other tools.  That set includes Curves Transformation, particularly the hue setting, SCNR, a touch of Histogram Transform, etc.  Its in times like these where the "art" part comes a bit more to the forefront, but, in reality its all a mix of science and art anyway, even with RGB images. 

And if you absolutely insist on using BN, then go with Rob's suggestion or ferret out those few clean background pixels as Geoff suggests.  Just make sure you do your pre-processing to eliminate those cold pixels you are worried about (which should happen naturally since you won't be doing any BN until after calibration and image integration, so that concern is a bit of a red herring).  There is a sequence and a set of tools that will fix any image.  A big part of the joy of this adventure is the process of discovery in finding those solutions.  That type of playing around explains why my workbook is now up to Rev41a.  I find a new way of doing things with every image I process.

Good luck,

Jim       

[rdryfoos]

Rodd,

In answer to "what do I do?" it's pretty simple.  I use the tools that work for a given image.  In the case of an image full of nebulosity I simply don't use BN any more than I use Dynamic Background Extraction since you need a clean background to be able to use that effectively as well.  In the latter case, I will test using ABE which works surprisingly well with lots of nebulosity and otherwise make sure I have a top notch flat master for each color to eliminate light variations that way.  With background neutralization, the same applies.  PI has a huge toolkit to work with but sometimes, no matter how much you like it and how much it works on other problems, a tool just doesn't work for what you are trying to fix.  I love pounding on things with a hammer but rarely use one when fixing a leaky faucet.

In cases like yours, I simply use other tools.  That set includes Curves Transformation, particularly the hue setting, SCNR, a touch of Histogram Transform, etc.  Its in times like these where the "art" part comes a bit more to the forefront, but, in reality its all a mix of science and art anyway, even with RGB images. 

And if you absolutely insist on using BN, then go with Rob's suggestion or ferret out those few clean background pixels as Geoff suggests.  Just make sure you do your pre-processing to eliminate those cold pixels you are worried about (which should happen naturally since you won't be doing any BN until after calibration and image integration, so that concern is a bit of a red herring).  There is a sequence and a set of tools that will fix any image.  A big part of the joy of this adventure is the process of discovery in finding those solutions.  That type of playing around explains why my workbook is now up to Rev41a.  I find a new way of doing things with every image I process.

Good luck,

Jim     

    Thanks Jim--I actually love ABE so i know where you are coming from.