Originally Posted by vanheden
This may surprise you, but "sharpness" and "good color" are actually the same thing. When you use a "good color" lens to photograph flowers with bright and bold colors, those colors will come through at full strength. Whereas a lens without good color will record the flowers in a more dull, low-contrast way. (Some photographers actually use such lenses on purpose,particularly older ones,to attain a certain mood for the image. I prefer to capture good colors first and then make it dull in post production -- that's easier than doing it the other way around. [])
<meta content="text/html; charset=utf-8" http-equiv="content-type" />
The color of a lens is so important that all the manufacturers measure it and publish charts on it -- it's called MTF. A perfect "100%" MTF says that the lens will give you the best color possible of what you're photographing. A low MTF, like 20%, means the color is significantly dulled and diminished.
So there are three related factors:
- "Good" Color: this occurs when you have high contrast at low and medium spatial frequencies, especially.
- Sharpness: this is a subjective combination of contrast from various spatial frequencies (usually the highest -- but it depends on who is using the word).
- Resolution: this usually refers only to the absolute highest frequencies, and ignores contrast.
So, the primary factor in what most people consider good color is actually contrast. But there are two other factors as well:
- Spectral transmissivity
- Color accuracy
- Color matching (with other lenses)
The first factor relates to how some lenses have a red cast, or a yellow cast, or some other color filtering effect to them. To me, that's bad. I'd much rather apply whatever color filtering I want in post, but I don't begrudge anyone their preference for doing it in the lens and without needing an extra color filter.
The third factor, color matching, is when you try to get two lenses that achieve the same color -- even if that color is inaccurate and/or has a certain color cast you like or dislike. At least if they are the same, whatever effects you get will apply equally when you switch lenses.
<meta content="text/html; charset=utf-8" http-equiv="content-type" />
The effect of all three of these factorsis very, very minute, and completely overshadowed by things like the white balance setting. That is, you can take two lenses that are completely opposite in their color cast, apply a custom white balance, and to most people the color will be the same. For really critical color, you can use a colorchecker chart and a custom profile to get them closer. It's only the most advanced and difficult situations where you truly have to have lenses with perfectly identical spectral transmissivity in order to achieve identical color reproduction.
<div></div>
Personally, I love the 17-40 as a full frame landscape lens. I think it produces very good results on my 5D2 at f/8. But that's because the image is enlarged/magnified a lot less, and a narrower f-number can be used to achieve the same DOF and diffraction. The net effect of these is that the flaws don't show nearly as bad and the results are sharper and have better color. But I don't like to use it at f/4 or f/5.6, and I don't like to use it on a crop.
Think of the 17-40 as a large, expensive camping trailer. If all you have is a compact car (e.g. crop camera), trying to pull that trailer will make your camping trip miserable. A small tent (like the 17-50 f/2.8) is more appropriate for your small car and give you better results. When you get a large truck some day, then you can pull the camping trailer and enjoy all the benefits of the expensive camping trailer.
Reply With Quote