[Leica] Fighting windmills?

[imx <imxputs@knoware.nl>]

The question of the possible image dgradation by the use of filters will
never end, nor be settled by whatever MTF measurement results. The reason
for my statement is the fact that the question is falsely framed.
The real question should be this:
what optical properties and optical quality does a filter need in order NOT
to degrade the image quality?
First a few facts:
Photographic optical systems consist of several lens elements that work
together to provide the designed image quality and optical properties,
Some basic reflection on the nature of ray bending in such a system will
tell you that (theoretically) adding a lens-element (a filter that is) to
such a system will change the optical properties and thus the image quality.
This is known by all optical designers and also by practitioners in the
field of cinematography.
When we use a close focus lens as an attachment we just do this: alter the
optical properties.
A filter in front of a lens does the same.
The amount of change does depend on only one factor: the quality of the
filter itself. Why should Leica add a built-in filter to their best ever
lenses (Apo-Summicron 2/180 and 4/280), if they were convinced that a filter
inherently would degrade the image quality. It does not, what can be proven
by using a 2/180 with and without the filter.
IF a filter has truly plane surfaces and is made from first grade glass,
there is NO difference in optical performance. I conducted test with MTF
apparatus to study this effect. BUT a filter with a heavy amount of dust and
fingerprints will reduce the MTF values significantly. As will do a front
lens surface that is full of dust and fingerprints.
The focal length of a lens can be defined in diopters. If a filter is
without any defects its own diopter value will be zero, not giving a change
in diopter value or focal length. If the filter is tilted in its mount, or
not plane or has other defects, there will be a diopter value to add to te
lens.
So testing 'a' filter in front of a lens has no value at all, unless you
first measure the optical properties of the filter itself.
to summarize:
any filter can generate flare in some circumstances, the amount of
parameters involved is not known.
Any filter might change the focal length of the system, depending on quality
of the filter, focal length used and distance setting. Te larger the
diameter of the filter, the more likely that a defect will turn up.
Any filter might degrarde the image quality, depending on the quality of the
filter. Leica filters are known to be fully plane.

To prove that 'a' filter might degrade image quality is futile: it has been
done already and very scientifically and practically.
To deny that a filter might degrade image quality is futile too: there will
be a drop in image quality when using filters with defects.
To assume that the degradation caused by a filter is more important than
defoccusing errors, exposure errors or erors of movement, is simply not
having a clue of the relative importance of the components of the imaging
chain.

Erwin