Re: [Leica] Lense testing...

[Rob McElroy <idag@pce.net>]

Austin,

Seeing as you have been given very few suggestions regarding your request for constructive input on the lens testing procedure, I offer the following;

1)  Make sure that the resolution and MT. tests are made not only in the center and edges of the frame, but also 1/3rd, and 2/3rds out from the center
of the image, and at the corners.  This will give you a much clearer picture of the optical performance of the lens across the entire field.  As you
know, the resolution and MT. in the center of a lens is often quite high but will fall off dramatically as you move toward the edges.  You may be
surprised at the figures you are going to get 2/3rds out with your 110 f2 Zeiss.

2)  If time permits you should make all of the above readings both tangentially and sagitally.

3)  Test the lens at the most widely used apertures. (Wide open, two stops down, minimum aperture - your choice, it's your test.)

4)  Make sure that the operator of the test is accurately making your measurements within the 24x36 or 6x6 image circle.  The image circles thrown by
the lens will be much greater than their respective camera format.

5)  Critical focusing of the aerial image is vital.  Look through it yourself to make sure the operator and you agree on the testing plane.

6)  What focusing distance are you doing your test at? (infinity, 20', 10', 5', 1' ?)

7)  If it was me, I would use the Hasselblad adjustable lens shade (extended to the maximum amount before vignetting) on each lens to make absolutely
sure there was no extraneous light hitting the front element or filter.  I would even bring black cards to hold up as gobos for the front element if
needed.

8)  I would also do a test with a light purposefully shining on the front element and on the filter to gauge the effect a backlit scene will have on
lens performance - with and without the filter in place.

Somewhere, I have the resolution and MTF data on your 110 Zeiss, that Zeiss published in the 80's.  It might be an interesting comparison to see if
your results are comparable (provided you can use similar test parameters).

Good Luck and I may be in the minority, but I too find the test interesting.  A question I have regarding the effect of the UV filter on lens
performance is:  Do large diameter filters on fast lenses have more of an effect (even if it's negligible or small) than smaller filters on slower
lenses?  The light rays passing through the center of a filter, pass through at a 90º angle to the filter's two surfaces.  The light rays coming
through the edge of a large filter (like over a Noctilux) are coming through the filter's two surfaces at a more oblique angle.  What effect, if any,
does this have?  Is that why Leica recommends not using a filter on the Noctilux - or is it perhaps only because of the increased flare potential on
the huge front element?

Do you think you can test both lenses in one hour ($175.00)?  It's a lot of work.

Here's an interesting side note on flare, ghost images, and the Nikon 80-200 f2.8 zoom.  In normal (non-backlit) situations the lens performs quite
well and is the mainstay in most photojournalists camera bag, but try and shoot a backlit subject where the sun (or other light source) is directly
hitting the front element, and you might as well use a coke bottle instead.  It is that bad.  In harshly backlit situations, fixed focal length lenses
still reign supreme.

Regards,
Rob McElroy
Buffalo, NY

Austin Franklin wrote:

> I decided to find out what the REAL effects of a UV filter are on image
> degradation.  I arranged for an afternoon of test time at a high end optical
> testing facility in Cambridge MA.
>
> I am bringing my Hasselblad 110/2 and Leica 75/1.4 with their respective
> Hasselblad and Heliopan UV filters, and will obviously test them with and
> without the UV filters.  They suggested as tests, both MTF and resolution.
> Center and out to the edges.  I also want to test focus shift.
>
> If anyone has experience with doing such testing and has some CONSTRUCTIVE
> input on the testing procedure, I welcome it.  Those who think this is
> silly, well, given that this only costs $175/hour, and one filter is near
> $100, I think that if I find out filters are useless, it'll easily recoup my
> money by selling all my image degrading UV filters.  If I find out filters
> do not degrade the image such that it is perceptible, well, then I won't
> waste my time debating this subject, as I will have facts to back up my
> conclusions, and that will be well worth the effort and expense.
>
> I will publish the raw test results, no matter what they are.  I fully
> expect this to be quite controversial, and I also fully expect to find there
> is 'some' degradation.  It is the significance of the degradation that will
> have to be 'analyzed' in order to draw any 'reasonable' conclusion to
> whether a UV filter does actually degrade the resultant image.