Suppose camera A took an exposure of 1.5s

Camera B took exposures of 1s and 2s--so I would like to approximate the noise level of a 1.5s exposure by some combination of the 1s and 2s exposures.

The noise is Poisson / shot noise of the number of photons gathered. And the mean number of photons gathered is proportional to the time of the exposure. My goal is to make a blend of the 1s and 2s shot that would have the same SNR as a 1.5s shot.

I haven't touched statistics for years and my head is still spinning from trying to solve the problem--but maybe there's a simple solution here! Thanks