How do you measure the IQ of a lens?

p.giannakis

Pan Giannakis
Local time
12:18 AM
Joined
Jul 8, 2008
Messages
6,444
Location
Stafford - UK
Coming from the field of psychology, I know it is relatively easy to measure IQ with people; you probably need the MMPI or a similar scale.
But how do you measure IQ in a lens and what does that mean?
 
As image quality is made up of a range of different factors (the most common being resolution / sharpness of course but others like distortion and chromatic aberration, vignetting etc also contributing) as far as I know there is no single aggregate weighted measure of IQ that takes account of all factors as you might find with IQ in the intelligence quotient sense (measuring spatial, verbal mathematical etc skills all in a single rolled up number).

This is made more difficult in the case of lenses because IQ is inherently a subjective issue in my view. For example its often said that certain lenses are "clinical" in their rendering. But many people prefer lenses that render a bit less crisply and more artistically - sonnar lens often are regarded as having this latter quality. A lens that might score highly on a technical score might be regarded as less good by some people because they want a different outcome.

Also it depends on the purpose for which the image is being made - if you are photographing something to reproduce it exactly for archival purposes you might want a lens with as few technical imperfections as possible, while for artistic purposes as noted some (myself included) prefer something slightly less clinical in rendering. For example I do not mind having a lens that vignettes. Many would regard this as a potentially unacceptable fault. For the sort of work I mostly do it can be an advantage - but if I want to remove it in post processing then usually this is a simple matter these days.

To make matters worse, lens performance changes as you change the aperture (typically being sharper as you stop down till diffraction sets in) and depending on whether the image you are shooting is closer or further away.

As you can see IQ in a lens depends as much in some cases about how the lens is being used as its inherent qualities. A lens that may have good IQ for one job may be not especially good at another task

The Modular Transfer Function is the most common measurement used to show IQ of a lens but it only really measures resolution. But it gives some idea of the optical performance potential of a lens.

http://www.nikonusa.com/en/Learn-An...hat-is-a-lens-mtf-chart-how-do-i-read-it.html

I kind of like the approach used on the following web site which uses a graph showing "blur units" to represent sharpness across the field- but to this other assessments have to be made also representing the other key factors affecting image quality as mentioned above.

I have no idea how blur units are calculated but the approach which is essentially a visual one works for me. The graphical representation is a 3 dimensional graph and as such you can see how the lens performs at each aperture setting across the field (ie in the centre and further out towards the edges) by viewing the 3D shaded area and by moving the slider up and down to change aperture. Think of the 3D shaded area as the sensor and the variations in the level of the area as showing how sharp or less sharp the lens is at that point on the sensor.

To use the thing, click on the graph to open it in a new window then try the above by "twiddling" the slider to see how changing aperture affects sharpness. You will see that the results change as aperture charges. (BTW the lower down the scale the better). For zoom lenses another slider is available along the bottom of the 3D graph.

Here is an example of an assessment made on a Nikon lens

http://slrgear.com/reviews/showproduct.php/product/127/cat/12
 
Not being trite but very serious, evaluate the photos it makes. Nothing else really matters.

I continue to believe that modern day digital measurements imply a level of precision that is only academic and bears little relation to the real world.

Personally, believe that modern day lenses vary from 95 to 98 on a 100 point scale while photographers vary from 3 to 95 on that same scale.
 
In a scientific sense, the perfect lens performance wise, looks flat as a pancake on all the measurement charts: MTF curves all flat at 100, distortion and aberrations flat at 0.
In real world, the best IQ lens, is the one that makes you open your wallet really fast, when you see what pictures it can make...
Just to prove the point, the last lens I have bought, is probably not much better than the bottom of a beer bottle...:
http://www.rangefinderforum.com/forums/showthread.php?t=138858&highlight=petzvar+kickstarter
 
The link to the Nikon site explanation is interesting enough, but does not explain why sagittal and meridional diagonals should score differently. I take it this is influence by aspherical elements and random manufacturing imperfections.
Sites are obsessed with lens sharpness, but they don't usually remember that the impression is formed as much by contrast as anything else. It was well known by shops that turning up the contrast would impress people. Contrast, especially shooting towards light sources, depends a good deal on glass coatings. Those coatings must be working hard because I've seen zoom lenses with 21 elements in them.
 
Back
Top Bottom