We’ve been trying to tease out what is the ‘fundamental’ information about an optical system. We have our own thoughts, but I am most curious to hear what the community thinks? I imagine everyone has different aspects they care about, so it will be particularly interesting to see what overlaps exist! Thanks!
What do you mean by an optical system? An imaging system?
Definitely the optical throughput (etendue) of the system is important
Also, some measure of the information content of the system (# of pixels, or more accurately an integral of the MTF over the image space)
Lastly a range of wavelength over which the system operates (colors are good).
Parameters like field of view, distortion, relative illumination, chief ray angle, etc. I would consider less ‘fundamental’.
Thats a really interesting perspective! I was looking at it from the imaging system perspective. I traditionally have considered FOV, spot size, F/# etc as fundamental, but I agree, the information content is a highly useful parameter. And of course, the wavelengths (and in my opinion materials used) are also really important. Lastly, I always like to know the general shapes used, so get an idea how hard it would be to manufacture said system.
I think when talking about fundamental imaging system properties, their is the first-order perspective and then there is the physical perspective. In my opinion, the field of view, distortion, etc of the system are fundamental in the sense that unless a component of the design is changed, those properties will not either. They are also extremely impactful in the real-world use of such a system and therefore I find them fundamental.
I agree with you that from the first-order perspective these properties are not fundamental, but if the first-order design of an imaging system was all that mattered, we might not have a job 
.
I think this also points to a natural desire as humans to distill very complicated concepts down into single values. This helps us draw general conclusions and its an invaluable tool, but it has its limitations. If the first-order system doesn’t make sense, the physical system has no hope. However, the first-order system is not actually what gets built.
At the end of the day, I think its the blend of the two perspectives that really wins. So, its a question of how to represent the macro (first-order) data and the micro (physical) perspectives in a harmonious overlapping format.
In my opinion, the most fundamental properties of the optical system is the data needed to build the physical object. The behavior of that optical system is determined by the laws of physics and the optical elements. However, I understand that most of us cannot infer the behavior just from a list of radii, materials, spacing, etc., and therefore its helpful to build a layer of abstraction on top.
So maybe the question of, “what data do you think is fundamental to understanding an optical system?” is more appropriate to me. To that end, from the optical designer’s perspective, I want to know the lens prescription (surfaces, materials, spacings, focal length, field of view, wavelengths, aperture / pupils, etc) so I can recreate it in my software of choice and also have an intuitive understanding of the system. The specific aberrations and other behavior of the system is data that I would store in my mind when I’m working on the system heavily, but I wouldn’t consider fundamental.