Cosmetic Correction Still Confused

[sreilly]

Most of the post I see on using Cosmetic Correction relates to it being used in BPP which I don't use. The problem this creates is it doesn't specify how it is used properly. Is CC used on un-calibrated linear images or after calibration of linear images? Even though my images are dithered I still have many dark and cold pixels to remove. I've attached a cropped area of my image that shows the problem in the stars as purple dots. This is an image that has 13 red, 13 green, and 11 blue combined images all 15 minutes each taken at -15C. Calibrated in PI using a master dark that consists of 115 - -15C 15 minute frames. The bias used to calibrate the flats was a 500 bias frame master. Flats created in PI as well using the procedure for making master frames posted on the PI website and used many times without issue. I used CC not too long ago but with the weather there hasn't been any data to work with for a while and it seems I've either forgotten just how I did use it or I'm doing something wrong now.

Thanks,

[RickS]

BPP runs CosmeticCorrection after calibration and before registration.  That seems like the appropriate time to do it.  It must be done before registration, at least if you're using a master dark or a defect list.  Running it after calibration ought to give better results than running it before since any interpolated pixels will then be based on calibrated neighbours.

Cheers,
Rick.

[JoLo]

Steve, I don't use BPP either and run the individual steps in pre-process.  My flow is calibration, cosmetic correction, star alignment, integration....I think i picked that up on a post here on the forum some months back.

Joe

[cdesselles]

Thanks, Guys!  Very helpful thread...  8)

[jerryyyyy]

Helpful indeed.  Are there some basic guidelines for the settings on CC?  I see there is no help file. 

[Alejandro Tombolini]

Hi Steve, it doesn't seems to be a problems to solve with CC, it seems a rejection issue. Why don't you try a simple integration without rejection and compare the result?

Saludos, Alejandro.

[sreilly]

I actually found an issue a few days ago with my 8 position filter wheel and had to make an adjustment. I re-took the flats, eliminated the first series of images, 18 in all, and now seem to have eliminated the problem. I haven't needed CC before as I usually have enough images to use either Winsorsize or Linear Fit combine which has always done a good job of pixel rejection. Although I admit I'm unclear of making adjustments to the low/high rejection and range selection settings.

-Steve

[Alejandro Tombolini]

...and now seem to have eliminated the problem.

Excellent!
Saludos, Alejandro.

[JoLo]

jerryyyy.....y:

I have had good luck with using the auto detect setting, make sure you check the hot and cold pixel boxes.  The other involves using a dark frame, and from what i have read the auto detect works just as well, but i have not tried the dark frame method, personally.

Joe

[jerryyyyy]

Thanks.  I tested it with some subs from last night.  Did not have yet enough for Winsorized or Linear processing and there were hot pixels on the Percentage processing.  Adding CC clobbered them. 

Thanks, will add to my routine. 

Not sure that the Darks would do much as the stacking already includes the Darks.