I mainly stack with scripts and I usually stack with the weighting based on noise. Is there a way to get SIRIL to output to the console what the weighting was used for each light.
If not, is there a way to take a list of lights and conduct a series of steps that would have SIRIL output stats that would easily allow the weighting it would use to be calculated?
I take light until it approaches dawn light, so the last few will have more light and I’m curious to see how SIRIL weights those then I weight by noise.
indeed we don’t display this, to avoid flooding the log (we don’t even show this in debug)
If you want to compute them, you can run a seqstat command to extract the statistics for the sequence. You want to pass the option full to get the scale value in the outputs.
And then it’s just a few operations:
first you need to compute the normalization factor for the ith image (I’ll assume you have additive with scaling): n_i = scale_0/scale_i (scale_0 is the scale of the reference image)
then for each image compute w_i=1/(n_i^2 * noise^2)
sum all those terms and divide by the number of images to get a norm N
divide the initial weights w_i by this norm, and there you go!
Cheers,
Cecile