p.1 #1 · To what extent is lens design a solved problem?
I took a hiatus from the online camera world for some time, as my hobby interests went elsewhere. My interest has returned, and one thing that I’ve noticed since paying attention to this world again is the sense that lenses continue to get better and better over time. For example, a decade ago, a 24-120 lens might not have won that many admirers and been regarded as a tool of convenience more than anything else, compared to other zooms that were ‘pro’, like 24-70/2.8 and 70-200/2.8 lenses, and certainly when compared to primes. Yet these days, the Nikkor 24-120/4 S is very highly regarded.
One might think that the move to the new Z mount perhaps allowed Nikkor to unlock some low-hanging fruit in lens design that the old F mount did not permit, but consider this: another brand that I’m interested in, Fujifilm, has managed to shrink the size of their 16-55/2.8, and re-released a bunch of their prime lenses besides, all for the same X mount. I also remember the first Sigma Art lenses being highly regarded despite their bulky size, and have gathered that equally good versions are starting to be released that are smaller. Correct me if I’m wrong, and please share good examples from other brands if you have them.
So, my question is, where does this room for development come from, and how much is there still to go? Are the physics of the way light travels through glass truly still being figured out by our lens makers, or — given that lenses have been around for centuries — is it a matter of incorporating new manufacturing techniques, tightening tolerances, or making use of better computational models?
p.1 #2 · To what extent is lens design a solved problem?
I think there are two main sources of progress. First, the computational design is a big one, not just in predicting what the design will do, but iterating on the design to optimize it, along with including tolerances in the design, so the performance of what can actually be manufactured is optimized. The second is computational lens correction. The design can allow flexibility in areas that are easier to correct in post. There will be a small penalty to correcting in post, vs not needing correction, but it allows for an overall better performance, or a smaller lens.
p.1 #3 · To what extent is lens design a solved problem?
Mirrorless cameras rely on in-camera corrections more than DSLRs did. Most people are content with the digitally corrected images because they never actually see the raw images. Mirrorless "raw" files are rendered on screen with considerable corrections by most editing software including Adobe. There may have been some improvements in lens manufacturing. But modern lenses are often hiding what would have previously been considered optical flaws like distortion and color fringing by applying correction profiles even when viewing in camera.
p.1 #4 · To what extent is lens design a solved problem?
There are a dozen or more optical aberrations, plus size, weigth, and cost considerations creating a massive number of compromise factors. Just look at that 3D pop lens thread. Lens design will never end.
