How do I adjust white balance in images captured with new QHY183C CCD Camera?

[wildbillinMT]

Greetings:

I have been beta testing the new QHY183C backlit cooled CMOS CCD camera.   There are a couple of threads about this camera posted on Cloudy Nights and on the QHY forum websites. 

Several software programs have been recommended by QHY and/or other beta testers for capturing images with this camera.  They include SharpCap, MaximDL, EZCAP-EQ, and Nebulosity.  Of these, it appears that SharpCap and Maxim DL have routines that allow one to adjust the white balance in the camera before capturing images.  However, I do not have MaximDL and SharpCap tends to crash frequently.  EZCAP is geared mainly to planetary and lunar imaging.   So, I found that Nebulosity V. 4.1 is the most stable program for deep sky image capture with this camera.  Therefore, I have been using Nebulosity V 4.1 to capture the images and then using Pixinsight to stack and process the images.  I should mention that Nebulosity does not have a routine to adjust the white balance of light subframes before image capture.   

There is an unusual aspect of this new CCD camera in that all light subframes have a distinct green cast to them after the raw .fit images from the camera are debayered in Pixinsight.  I have been using automatic background extraction (ABE) after stacking to remove most of the green in the image.  The green background is successfully removed, but the propensity towards an inordinate amount of green in the resultant target image remains even after the ABE process routine is employed.

QUESTION:  In Pixinsight, is there a way to adjust the white balance either in each of the debayered light subframes or in the final master light frame?   

I hope there is a staightforward method of doing this in Pixinsight and hopefully that method avoids "Pixel Math" routines which are still somewhat ancient Koine Greek to me.

Your expert help with this would be appreciated.

Thanks!

Bill in Montana

[pfile]

hi bill, the green is normal for almost every OSC out there, not just your camera. it's kind of a visual artifact generated by the STF process, which uses the statistical properties of the image (noise) to compute an auto-stretch. because almost every OSC has 2 green pixels per bayer quad which get averaged together during debayering, the green channel has 1.4x the SNR of the other channels. if you run STF with channels locked, the image ends up looking green due to the stronger SNR there.

in order to use STF on such images, you can unlock the channels, and you'll get a more balanced image.

ABE/DBE are simply normalizing the backgrounds as part of how the defaults are set up. since ABE/DBE are probably the first things you should do to an image before color calibration, you're doing the right thing.

the next step after that is to use BackgroundNeutralization, which after DBE/ABE may not change too much, but it's still the right thing to do.

after that you use ColorCalibration to fix the white point.

I'm sure there is a tutorial on this in Pixinsight Resources.

after your background is proper (either the DBE or BN step) be sure to re-lock the channels in STF because you don't want to know STF's idea of what the image "should" look like - you want to see what it "really" looks like.

after you've stretched the image then of course STF does not matter any more since you can see the image without any STF applied.

rob

[chris.bailey]

How about http://harrysastroshed.com/pixinsight/pixinsight%20video%20files/2013%20pix%20vids/ccfinal/ccfinal.mp4

Chris

[wildbillinMT]

Hi Rob and Chris:

Absolutely great and very useful information.   

I was beginning to wonder if this camera was different from all other CCD camera, but thanks for clarifying the "green effect" for me.

From the noted tutorial I was able to discover the PROCESS - NOISE REDUCTION - SCNR routine in PI.  It does the trick on the master light frame - eliminates the green aura and allows me to get
normal color renditions by further processing.

Many thanks for your kind assistance.  I like PI, and this makes me like it that much more.   https://pixinsight.com/forum/Smileys/default/smiley.gif

Bill

[wildbillinMT]

Hi Rob:

I forgot to ask you:   How do I unlock the channels in PI?   What routine do I select to accomplish this?   

Please advise.

Thanks!

Bill

[RickS]

How do I unlock the channels in PI?   What routine do I select to accomplish this?   

ScreenTransferFunction process.  Click on the icon at top left that looks like a couple of links in a chain to unlink and re-link the channels.  Then you'll need to redo the screen stretch.

Cheers,
Rick.

[pfile]

oh sorry for the confusion - the channel lock/unlock is part of STF. it controls how the screen stretch is computed - either same stretch for all channels (locked) or independently (unlocked). the little linked chain button in the STF process is what you are looking for.

rob

edit: ninja'd