Almost flawless! 35mm f/2 IS USM review

Al Downie wrote:

Seriously, joking apart, in my view some lenses *do* display something special (at all apertures) which cannot be measured by MTF charts or specifications.

If it can be seen, it can be measured.

Bokeh is one of these things;

Bokeh can be measured:

http://www.rickdenney.com/bokeh_test.htm

...depth of colour is another;

Color can be measured.

...three-dimensionality is another.

If it's a real effect, it can be measured.  But first, it needs to be defined.

You can't describe them in specifications and it's got nothing to do with the sophistication of IS or resolving power or whatever. Some people see it, and some people don't - some people never see past sharpness tests. In my view, the 35IS doesn't have it.

It's a simple matter to test, obviously.  Take pics of the same scene with the same camera, same settings, but different lenses.  Post, say, five photos from each lens, and have people say how the photos differ.

I've actually seen such tests for various claims, and it never went any better than random chance.  Of course, there were always a few who would look for measurable signatures of the cameras / lenses being compared, and could tell them apart.  For example, let's say one lens had more distortion than another -- they would look for horizontals and verticals in the photo and correctly deduce on that basis.  But never have they done better than chance going off of "immeasurables", for the simple reason that, if you can see it, you can measure it.

So, no, it isn't all about sharpness.  But there are any number of other measurable quantities that can be seen and quantified without having to resort to "magic pixie dust" arguments about why A is "better than" B.