Fewer hot pixels without dark optimization?

[jfrech14]

I tried to see the difference between the two and saw less(clearly visible) noise without optimization. Is optimization for when there are no matching darks or for all the time? I am careful to always take matching darks for everything I take so I am curious if I need it.

Thanks!
Josh

[sreilly]

Josh,

Optimization is when you don't have matching darks. Lets say you have images that are 10 minute exposures and only have 15 minute darks. You are optimizing or scaling the dark to use with the shorter light exposure. At least that is my understanding. I do this with my flats. I have a master bias but my shortest darks are 300 second so by checking the optimize box I can use a longer dark. You can't do it the other way, shorter dark than light. At least this is my understanding. Personally I have a master dark library that is based on my normal exposures of 5, 10, 15, and 30  minute exposures. When I did M42 and had 30 second, 60 second, 3 minute, and 5 minute exposures, this being with a Tak FSQ-106 @f/5, I used my 10 minute master dark. I keep a seasonal library of 10, 15, and 30 minute masters which is what my exposures normally are at temperatures that range from -30 to -10 in 5 degree increments. Some just scale but I prefer to use darks matching the lights.

Hope this helps.

Steve

[Juan Conejero]

I tried to see the difference between the two and saw less(clearly visible) noise without

Don't confuse noise with outliers. Dark frame optimization minimizes noise after subtracting the master dark frame. However, hot pixels may be left undercorrected or overcorrected, depending on the value of the dark frame optimization factor, as a result of this process. This is normal.

The remaining hot/dark pixels will be rejected during image integration, provided you have dithered your frames (which you should *always* do). Or they can be also removed very easily with the CosmeticCorrection tool.

Is optimization for when there are no matching darks or for all the time? I am careful to always take matching darks for everything I take so I am curious if I need it

.

Dark frame optimization is always useful, even if your darks are well matched to your lights. However, you need a very good master dark frame, which implicitly requires a very good master bias frame. See this tutorial for more information:

http://pixinsight.com/tutorials/master-frames/

BTW, this residual hot/cold pixels problem is being addressed with a new dark frame optimization algorithm that we are implementing. It will be released with the next version of the ImageCalibration tool.

[jfrech14]

I always make sure to take at least 20 darks per exposure length for lights and flats along with at least 20 matching bias nearly each night I observe so I was wondering if it would benefit me but you are right. I didn't fully consider outlier vs noise which is why I put (clearly visible) in parenthes I will admit the images I did this rough test on were one of the nights I did NOT dither which fully explains the hot pixels remaining. Cosmetic Correction tool is something I have been trying to get around to learning so I will be sure to do that soon.

Thanks for the great response My curiosity is satisfied! Reading the optimization description it made sense to always use it but I was visually deceived.

The main reason I started to question dark frame optimization was because I have good masters of at least 20 and upwards to 40+ in the master but when I enabled dark frame optimization I would every now and then get that the frames did not match and I would get a k0=0. What could be my issues there? (this happens on both dithered and undithered subs)

Thanks again!
Josh