OSC WBPP yields mottled master lights

John Kmetz

A few recent projects are giving me headaches. After doing the WBPP on the lights, flats and darks from my ASI071MC, the master light frames are showing a large scale mottling or splotchy backgrounds.  Unlike a gradient that is brighter on one side of the frame than another, these are lighter and darker random large patches, with hints of purple and green in these regions as well.

Am wondering if this has to do with checking the Separate CFA Flat Scaling Factor in the WBPP flats screen. This check is a recent addition to WBPP and I am not sure of the merits of using it or not.

Both a broadband and narrow-band projects are being wrecked by this splotching, which is not pulled out by ABE.

Thanks for any advice.

John

BlockHead

Hi John,

Hmmm... This is a complicated question- and it looks like you have a big project here. I think you need to simplify and just solve the mottling problem... You mention that you think it could be the scaling Factor parameter. I doubt it. None of these things affects a local part of the image. (Local) Normalization does though.

But you can test your theory and just do a smaller set of data without the scaling of flats. 

On another note- I am very concerned about all of the optimization I see- especially the optimization of short exposures. In the "Adam Block" universe of image processing this is a big no-no and will likely cause some issues. I am not saying this is the problem- but it is a well-known source of problems. (I have videos that explain and demonstrate the issue... please look in the WBPP set of videos where I cover this topic in detail.)

-the Blockhead

John Kmetz

Thanks Adam. I have now tried WBPP both with and without the Scaling Factor checked and this has not made a huge difference. The big patches are still there.

I have noted the optimization of frames, but this is the same thing I have tried successfully in two prior projects. I like to try a series of short darks, instead of precisely timed flat darks, and again this has worked in the past. I also tried another set of darks used last year and again no influence. And I have watched all your recent WBPP videos. All my calibration diagrams are what I wanted.

There must be something I changed in WBPP I am not thinking of, something that changed in recent updates, or perhaps some change with the OSC camera.

But trying a short project with few exposures may give better hints.

Regards,

John

BlockHead

My guess, without any further knowledge is the Normalization of the data. This is a change. 

You might try running *without* local normalization. 

You might also try the experiment of going through registration- and running NSG for the normalization of the data. So there are some things to explore. The final thing is the quality of the flats... but this is another big can of worms.

-the Blockhead