Poisson and Readout noise draws appear to be correlated - that is, the standard deviation of an image with just Poisson and Readout noise (no stars, no other errors) equals the sum of the standard deviations of an image with just Poisson noise and an image with just Readout noise. Is the same random number sequence used for both noise sources within STIPS?
A simple check is that a simulated blank image with readout noise enabled and a background of 144 electrons, with only readout and poisson noise sources enabled gives an image with mean 144 and standard deviation 24 electrons. It should, of course, be sqrt(144+12^2) =16.97 electrons.
Poisson and Readout noise draws appear to be correlated - that is, the standard deviation of an image with just Poisson and Readout noise (no stars, no other errors) equals the sum of the standard deviations of an image with just Poisson noise and an image with just Readout noise. Is the same random number sequence used for both noise sources within STIPS?
A simple check is that a simulated blank image with readout noise enabled and a background of 144 electrons, with only readout and poisson noise sources enabled gives an image with mean 144 and standard deviation 24 electrons. It should, of course, be sqrt(144+12^2) =16.97 electrons.