"kilopascal" wrote on 2002-12-08 18:00 UTC:
> I read through the link, but did not see anything that matches your explanation.

My source was a recent article by Robert Seetzen in c't 25/2002 page 28
(a very reputable computer technology magazine), which is in German,
therefore I did not originally quote it here. Here is my translation of
the relevant passage in the article:

  "The name 'Four Thirds' suggests an image diagonale of 3/4 inch, that is
   around 3.9 centimeters. In an interview published shortly after Photokina,
   a manager of Olympus USA wrongly attributed the name to the aspect ratio
   four to three widely used in digital cameras. Olympus and Kodak clarified
   after our enquiry, that 'Four Thirds' sensors will have a fixed image
   diagonale of 22.5 mm, without any prescribed aspect ratio. Whoever now
   asks for the meaning of the 'Four Thirds' name is referred to the odd
   calculation used for old vidicon tubes: a one inch tube can only scan an
   image area with a diameter of around 16 mm. As a basis for this image
   area calculation, Olympus and Kodak defined a value of 16.88 mm for a
   1 inch tube, leading to 22.5 mm for a 1.333 inch CCD. The use of the
   historic inch measure to describe sensor dimensions is by the way
   very common, as a look in the CCD specs of other manufacturers
   confirms."

So the 4/3 optics standard does in fact not prescribe any specific image
area aspect ratio at all. It only prescribes the diameter of a circular
imaging area, within which the optics must project an image properly.
This diameter is for a rectangular imaging area equivalent to the
diagonale of the image sensor field. This makes sense: Remember that
optical lenses are rotationally symmetric objects, therefore any aspect
ratio is irrelevant for an optics standard. It merely contributes to the
misunderstanding that the first sensors designed for the new standard
have the 4:3 aspect ratio of the old "academy" cinema format, which was
later also adopted by TV standards and PC monitors.

I hope this clarifies the issue. I would suggest, you retract your
incorrect letter to dpreview.

Markus

-- 
Markus G. Kuhn, Computer Laboratory, University of Cambridge, UK
Email: mkuhn at acm.org,  WWW: <http://www.cl.cam.ac.uk/~mgk25/>

Reply via email to