Archived posting to the Leica Users Group, 1998/06/22

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: [Leica] Image assessment
From: Erwin Puts <imxputs@knoware.nl>
Date: Mon, 22 Jun 1998 20:10:43 +0200

I know the vast difference between lens evaluations (done in the name of
objective test reports) and user impressions (the most important as far as
the leica user is concerned).

To help sort out some confusion and give some common background for a
rational discussion, consider the following notes.
The basic concept used in image analysis is the minimum area of a point,
that will be seen by the unassisted eye as a solid point and not as a
circle. Obviously, any subject point, however small must occupy a certain
area in the image plane. The extension of an image point is dictated by
diffraction and is called the Airy disc. But in most cases this very small
image point will be less than ideal because of aberrations. These
aberrations will cause the fringes of the point to be diffuse and/or to be
so large that adjacent image points will overlap and not be distinguishable
as individual points. The circle of confusion is a practical tool to define
the limit of extension of a point. There is a widespread tradition to
equate this minimum area of a point with the concept of sharpness.
But sharpness is a subjective term and not dependent on resolving power,
but upon the ability of the lens/film-system to record coarse detail.
Resolution is the objective assessment of a film/lens system's ability to
record fine detail, recognsibly rendered.
Now resolving power is dependent on image contrast and therefore high RP
can be coupled to low or poor contrast at low spatial frequencies
(linepairs/mm).
But good contrast at low spatial frequencies gives the best sharpness
impression. Note that there is a diffenence between apparent sharpness as
perceived by the eye and measurable sharpness which is defined as
'acutance'.
That is the reason why coarse grain films give good sharpness impression.
But lack of course in the recording of really fine detail. With coarse
grain films it is impossible to see if some illuminance difference in a
small detail is real or caused by random noise in the grain pattern.

To appreciate the quality given by leica lenses one needs first of all to
differentiate between apparent and real sharpness and second to learn how
to record very small object details. These can only be recorded by fine
grain films as the grain noise will kill any subtle detail. But modern
colour neg films are all grainless films? The colourneg film is composed of
dye clouds and not of classical silver grains. Yes but this smooth cloud
pattern will actually reduce the ability to record very small details.
The real image quality then depends on the (interdependent) relation
between resolution, contrast and acutance.
Resolution and contrast are conveniently combined in the MTF graphs which
are horrendously difficult to analyse in sufficient detail. (But watch my
next article in PhotoTechniques).
One element of MTF graphs are easy to use: the socalled cascading function.
MTF figures can be multiplied to get a conclusion. So if a lens MTF has a
contrast figure of say 40% at 30lp/mm and the film has a 60% contrast at
the same lp/mm, the resulting contrast is 24%!!
Add this number to a low contrast detail in the shadows and you see
immediately that the recoring of fine subject detail is a problem and a
task.

Looking at film MTF's then is required reading for leica users. If one film
has a contrast of 90% at the very important 10lp/mm and another one has
100%, this one needs to be choosen to get maximum sharpness impression.

A film with a 100lp/mm resolving power at a contrast of say 20% is not much
use in most situations as the lens can record 100lp/mm with at most 5%
contrast as is below the discriminating power of the eye.

To be continued

Erwin