Image calibration issue

[McZaustria]

Hi there,

I'm not sure if I assess the issue correct, so forgive me if I'm wrong.
Lights as well as darks show hot pixels (a lot if they are really hot pixels)
Now if I compare these "hot pixels" to other frames or the master dark, they are at the very same position.
Am I wrong assuming, the image calibration process should eliminate them by subtracting them from the light using the master dark?

I already tried to do it with or without bias and flats, there is no change. (Just for the fun of it, I fed them to DSS it actually did not show these "hot pixels")
If I go all the way and stack the calibrated and aligned frames, the "hot pixels" are multiplied by the amount of frames and slightly shifted so the all become visible and stars are elongated.

I've compared my steps to various youtube videos and it's all the same..
Any ideas what I could be doing wrong? I truly appreciate any input!
Thanks, Oliver

[pfile]

do you have "optimize" checked under the dark tab of ImageCalibration? if so try without it checked.

rob

[McZaustria]

Hey Rob,

thanks for your answer!
I have played around with the optimization threshold and actually get to a point where these pixels start to disappear.

Now, I have retaken my dark frames in complete darkness (to be sure there's no stray-light), haven't been using bias or flats at all and it's still the same.
Of course I can work around it with the threshold but would like to know what the issue is or what I'm doing wrong..

Oliver

[pfile]

well, the problem is simply that PI is going to scale the dark to minimize the noise in the calibrated frame. hot pixels do not obey the normal linear rules of dark current buildup, so they can be over or under- corrected in the calibrated frame while optimizing.

if you have residual hot pixels after calibration, the best way to take care of them is probably to use CosmeticCorrection... or if you have dithered your lights, you can just reject the residual hot pixels during Integration.

rob

[McZaustria]

Thanks for the explanation.
I generally thought it's "easy" to subtract fixed-pattern noise, plus the description in PI states something like " there's something wrong with your dark frames if you have to use the optimization threshold function".
That to be the reason, I wanted to get to the bottom of it.

Oliver