Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
In this Discussion
DBE
In the correction parameter discussion you said that division is used for flat field errors and subtraction for gradients and scattered light. In the small scale structures example with scattered light from an out of field star you used division for the correction. It "worked" but should you have used subtraction as it clearly was not a flat field problem.
Comments
Yes you are right, the scattered light is certainly additive signal. The primary point I was making was the (desperate) method of using a gazillion samples. This isn’t something I would normally do.
As far as subtraction or division... because only the sky is underneath that light, the math operation to model it isn’t critical. If the scattered light was over my object... then I might compare which operation is better.
Later this weekend I will find my icon and see If using subtraction matters. I am guessing not but it is a good experiment.
To this day I am not certain why division would not always be equivalent or better than subtraction where the s/n is low. I need to learn more about it.
-the Blockhead