images: philosophy and modeling

On the plane home, I started writing a paper about our lucky imaging stuff, with a combination of computer scientists and astronomers in mind. This led me to write quite a bit about how we think about images: They are noisy samples of a PSF-convolved intensity field. The PSF should be the pixel-convolved PSF at the focal plane. That's important! Well-sampled images are easier to use than ill-sampled images. Telescopes can never measure the unconvolved intensity field; that's an object that doesn't really exist in empirical science. And so on. All this kind of discussion sounds very philosophical, but it has important implications for how the math works, what we present as our output, and the speed of our computations.