At 3:50 PM +0000 3/21/06, Malcolm Stewart wrote:
----- Original Message -----
From: "Peter Hancock" <[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Tuesday, March 21, 2006 12:16 PM
Subject: EOS to filter or not...
As a postscript to the recent thread on whether to put filters on by
default, I just tested my 100-400L with and without the rather ordinary
Jessops filter that I got with it. Not much at 100, but at 400 there is
a visible reduction in resolution and contrast. I'll try again after
I've given it a really good clean :-)
Peter
________________________________________
Peter,
Long focus lenses are quite critical about the quality of glass used as a
filter.
....and extreme wide angle lenses. Moderate focal length lenses are
least affected.
The most common defect in filters is that the two surfaces aren't
parallel. So you have a wedge. It can ruin the results from a high
performance tele. In the 70's I tested a number of filters (about 50)
and found that about 1/2 had some parallelism issues; all of the
Harrison & Harrsison, many Tiffen, some Hoya and no Nikon or Leica.
Since then I've checked all filters that I've bought. It's quite
easy. Put the filter on a dark cloth on a table and look at a
reflected point source of light at about 45 degrees. Now rotate the
filter. If the reflections from the two surfaces move relative to
each other, the surfaces aren't parallel.
This is harder to do with very good multicoated filters, but you can
improve the reflectivity by smearing the surfaces slightly with your
fingers. You really can't damage the filters by doing this, but some
camera store sales people freak out if you do this. They tend to be
the same ones who say that 'all filters we sell are perfect'.
--
* Henning J. Wulff
/|\ Wulff Photography & Design
/###\ mailto:[EMAIL PROTECTED]
|[ ]| http://www.archiphoto.com
*
****
*******
***********************************************************
* For list instructions, including unsubscribe, see:
* http://www.a1.nl/phomepag/markerink/eos_list.htm
***********************************************************
