Lens qualities, other than IQ

[Steve Axford]

Over the years I have noticed that those lenses that I use the most are outstanding in areas other than straight resolution. There are 4 things that I noticed, only 1 of which gets much attention in reviews. These are
1. Colour. Very rarely covered by reviews apart from a few words. Colour can be how well the lens renders colour and the intensity of colours. The very best lenses often render colour much better than not so good lenses. When you use these lenses you always think, "Wow, the colour with this lens looks great", but how do you measure it?
2. Contrast. This is measurable, but few people ever do it (it must be complicated), but it does make a big difference.
3. Bokeh. This is critical if you shoot wide open or near to, but how do you quantify it? Consequently there is little space given to an analysis of it.
4. Distortion. This is covered extensively by most reviews as it can easily be measured.

I tend to look for these things in reviews and will often take more notice of them than any resolution tests, even though few words are devoted to them. Lenses like the Canon 50mm f1.8 ("Nifty fifty" ) are actually rated as being sharper than the Canon f1.2L, but in other respects the 1.2L easily wins (so it should too as it is a lot more expensive).

[jjphoto]

...1. Colour. Very rarely covered by reviews apart from a few words. Colour can be how well the lens renders colour and the intensity of colours. The very best lenses often render colour much better than not so good lenses. When you use these lenses you always think, "Wow, the colour with this lens looks great", but how do you measure it?

Spectral transmission varies between lenses and it certainly can be measured and displayed but without direct comparisons with known/reference lenses (ie lenses that YOU are familiar with) it may be meaningless. How many would appreciate the relevance of a subtle differences in a graph. A comparison of graphs would make more sense.

I've never been so concerned about the colour rendition of a lens mainly because I shoot RAW and I know I will have to tweak every image anyway, but the differences between lenses are there and fairly obvious at times. By way of a simple example, I normally boost saturation with all my Canon lenses but hardly ever do so with Leica R lenses used on the same Canon body, as the Leica R lenses seem to have more saturated colour. Olympus lenses seem the opposite, less saturated, more subtle and muted.

...2. Contrast. This is measurable, but few people ever do it (it must be complicated), but it does make a big difference.
...

Correct me if I'm wrong but isn't this what the low frequency MTF shows?

I think it's uncommon to find lenses with mediocre contrast these days as every lens is multicoated. Veiling glare is less common than it was just 15 or 20 years ago. Hot spots, a form of flare, still occur in specialised applications but these are uncommon.

...3. Bokeh. This is critical if you shoot wide open or near to, but how do you quantify it? Consequently there is little space given to an analysis of it.
...

The quality of the out of focus (OOF) areas is often and most accurately described as under/overcorrected for spherical aberration (SA) and this tells you if the background will be soft or harsh relative to the point where the lens is actually focused. This is the easiest and IMHO the most accurate way to describe bokeh. It is fairly commonly used although TBH I don't read many lens reviews any more and prefer to judge with my own eyes.

Undercorrected SA:


Overcorrected SA:


Incidentally the two images above where shot with the same lens but focused in different places.

Bokeh has other qualities too and these can be important, such as the shape bright points of light that are OOF. A round OOF highlight is less distracting than a hexagon, 'Ninja Star', doughnut etc. Generally the shape of the aperture blades determines this quality. Overcorrected spherical aberration also causes 'soap bubble' OOF highlights and this can be desirable for effect alone. mechanical vignetting of the aperture can also cause swirl, another bokeh effect.

You can manipulate bokeh to make it bad for effect, eg http://photocornucopia.com/1053.html

...4. Distortion. This is covered extensively by most reviews as it can easily be measured...

Distortion is often clearly stated by manufacturers although to me it is meaningless due to the subject matter that I photograph. Clearly it is important in many types of photography.

I think that to find all the information that you want or need about a lens takes a bit of effort by looking at many reviews but mostly by examining the images that show the various qualities that matter most to you. By way of example, LoCA is one of the most important qualities to me yet this is not often described however a simple photograph of the right subject will tell you everything you need to know about it.

JJ

[Steve Axford]

I agree, you can get some data on these things, but, as you point out, it is only useful if it is presented in a usable form. Colour information, for example, can be very complex and it also relates to our eyes sensitivity to colours, which varies, so it is almost impossible to compare quantitatively. You say "I've never been so concerned about the colour rendition of a lens mainly because I shoot RAW and I know I will have to tweak every image anyway.", but that assumes that the differences are simply a matter of saturation or white balance and they can easily be corrected in photoshop. From long experience photographing things like brightly coloured fungi, I would suggest that this isn't true. It's a bit like saying you can apply a polarizer in photoshop. As I said, it can be very complicated and it is very difficult to measure in any objective way, so I can understand why it is largely ignored, but for certain types of photography it can make the difference between an average and a spectacular shot. But, this does depend on the type of photography that you do.
Bokeh - as you say, you need to try it to see how it looks, to you.
Contrast - you could well be right that this is a legacy concern.
Distortion - it does concern me, but reviews usually cover it quite well.

[swifty]

The issue comes down to what is considered quantitatively better I think.
Attributes such as resolution (higher the better) and distortion (lower the better) can both be easily measured with a well accepted goal.
I'm sure your 1-3 can also be measured quantitatively but the goal is a pleasing rendition which is very much subjected to personal tastes.
So reviewers probably shy away from these subjective attributes in a bid to make their reviews seem more objective. But you'll find plenty of user opinions on things like colour and bokeh which I find useful to varying degrees depending on the user's experience with various lenses that he/she's comparing to. But nothing beats looking at the actual images so I tend to pour through as many examples of real world use whenever I'm researching a lens.

[jjphoto]

.... As I said, it can be very complicated and it is very difficult to measure in any objective way, so I can understand why it is largely ignored, but for certain types of photography it can make the difference between an average and a spectacular shot. But, this does depend on the type of photography that you do....

I do see differences from sensors, or brands of cameras, and as colour is the result of an entire system it's not particularly fair to blame or praise the lens alone for great or rubbish colour (exposure plays a part too). I suspect the colours you sometimes see in nature are extremes which may not fit any colour gamut at all. That they ever look good on a computer might be the result of using a specific combination of lens, camera, profiles/RAW converter etc. Change any one item and the magic is gone. If you used the same Canon lens on a Sony camera would the colours look as good?

[Steve Axford]

Good questions, and I really don't know the answers to them. For the moment I will have to treat it as an art rather than a science. The very best lenses/cameras do produce very good colour that allows great freedom with post processing. Lesser quality lenses/cameras allow much less freedom.