Cosmetic file generation

Post Reply
Kevin Gurney
Posts: 15
Joined: Tue Nov 27, 2018 5:42 pm
Location: Sheffield, UK
Contact:

Cosmetic file generation

Post by Kevin Gurney »

In my search for a solution to my 'strange artefact' problem, about which posed recently, I started to think about the cosmetic file...
I understand that the dark is scaled automatically for the exposure, but what about the cosmetic file? Do you just make one for the 'reference dark' (with longest exposure you use)?
Does its contribution get similarly scaled?

Kevin
David Boyd
Posts: 161
Joined: Wed Dec 11, 2013 5:50 pm

Re: Cosmetic file generation

Post by David Boyd »

Hi Kevin,

Hot pixels don't scale with length of exposure in the same way as dark current, so you cannot "correct" for them in the same way. They are also temperature dependent so very sensitive to the temperature at which your camera is operating. The normal approach is to remove them and replace them by a local median (or some such). The master dark can be used to identify them by plotting a histogram of pixel values and identifying where the downward slope intersects the horizontal axis. Beyond that point there is usually a long tail of higher value pixels. These are the hot pixels in the image and you can make a decision where to set the threshold for applying the cut. That is the approach I use in ISIS.

Best wishes,
David
Kevin Gurney
Posts: 15
Joined: Tue Nov 27, 2018 5:42 pm
Location: Sheffield, UK
Contact:

Re: Cosmetic file generation

Post by Kevin Gurney »

Hi David

Thanks...
I realised after I posted that cosmetic information couldn't be scaled because it is just a list of pixels - all intensity information has been lost.
I can also see that using the cosmetic file derived form the unscaled-longest-exposure-dark is OK, as although it will be somewhat agrresive in its identification of hot pixels in a scaled dark, using a median to 'patch over' these will work OK

Cheers
Kevin
Post Reply