Re: [Leica] focus shift due to filter?

["Henning J. Wulff" <henningw@archiphoto.com>]

At 2:00 PM -0700 2/21/01, John Collier wrote:
>The references to this matter that I aware of is in the "Legendary Leica
>Lenses" articles by Dick Gilcreast in the LHSA Viewfinder. He mentions it in
>the article on the 135mm Hektor and here is the relevant quote:
>
>"The one caveat I should mention in using the Hektor is in the use of
>filters with the lens in rangefinder mount. In common with most other long
>lenses, Leica's as well as everybody else's, the glass in the filter will
>change the infinity setting very slightly, making the lens focus just a
>little beyond infinity at the largest apertures. It is well to either stop
>down to moderate apertures or use reflex focusing when using filters at the
>largest apertures at very long distances. I first noticed this effect with a
>200mm f/4 Telyt many years ago, and it took a while to figure out what was
>wrong. The effect in the Hektor is less than in the Telyt, but still
>noticeable at infinity with the lens wide open."
>
>I have not investigated this myself so I cannot confirm it.
>
>John Collier
>
>>  From: "Austin Franklin" <austin@darkroom.com>
>>
>>  A claim has been made (elsewhere) that using a UV filter will cause a 'focus
>>  shift', which, would be more critical for a rangefinder...depending on how
>>  shifted the focus is.
>>
>>  What are the facts around this claim?  I take a Hasselblad with a 110/2, and
>>  a magnifying hood...critically focus on something ~3' away, and I don't see
>>  a focus change with a UV filter on or off...and all my Noctilux and Summilux
>>  images are dead on focus wide open...and I do use UV filters, and I do a LOT
>>  of close up work...I would think I'd see it there.
>>
>>  Perhaps there is some focus shift, but it is certainly not significant.
>>

This is strange.

If the filter is optically correct; ie, homogenous  and perfectly 
plane-parallel, there should be no problem at infinity. There is no 
opto-geometrical reason for this to occur. On a rangefinder camera, 
there is a theoretical shift of the object plane with respect to the 
measured distance, but this is a fraction of the thickness of the 
filter, so of no practical consequence. SLR's avoid even this 
miniscule error, of course.

A lot of filters aren't that good, especially w.r.t. being 
plane-parallel, and this will cause problems with well corrected 
lenses, especially longer telephotos. Very wide angle lenses also 
suffer more than normal lenses from bad filters.

If filter surfaces are not flat, they will act as supplemental lenses 
and change the focus characteristics. More common today are filters 
that do not have parallel surfaces. I had about 12 Harrison & 
Harrison filters in the 60's that all had this problem. At larger 
apertures with telephotos that caused a lot of problems. Most brands 
of filters that I have used, even Leica's, have had some samples with 
non-parallel surfaces. In the past 10 years, most filters have been a 
lot better.

- -- 
    *            Henning J. Wulff
   /|\      Wulff Photography & Design
  /###\   mailto:henningw@archiphoto.com
  |[ ]|     http://www.archiphoto.com