[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Rollei] Digital Daydreams (long)
- Subject: Re: [Rollei] Digital Daydreams (long)
- From: "Richard Knoppow" <dickburk >
- Date: Fri, 2 Jan 2004 22:53:25 -0800
- References: <68DDE3FC-3C6A-11D8-A325-000A95894CEE
- ----- Original Message -----
From: "Bernard Cousineau" <flatbroke
Sent: Friday, January 02, 2004 6:57 PM
Subject: Re: [Rollei] Digital Daydreams (long)
> > This ignores the location of the rear nodal point, the
point from which
> > rays appear to be projected in any lens, whether simple,
> > retrofocus. Read nodal point location determines the
> > incidence of the light on the sensor.
> I've been thinking about this (it takes me a while because
optics is not my forte), and I think that the relevant
measurement is the exit pupil, not the rear nodal point.
> The rear nodal point for a retrofocus lens that is shorter
than the mirror box (let's say 35 mm) will be behind the
last lens element. To be more specific, it will be at one
focal length's distance from the focal plane. I had another
look at the
> Zeiss spec sheets, and in every case that I calculated,
the actual focal length is exactly the same as the back
focal distance minus or plus the position of the rear nodal
point (depending whether it is located behind or in front of
the last lens
> For example, the 3.5/15 has a back focal distance of 36.3
mm, a rear nodal point located 20.9 mm behind the last lens
vertex, and a focal length of (36.3-20.9=) 15.4 mm.
> If I look at a ray trace for a retrofocus lens, the rays
behind the lens seem to come from the exit pupil, and not
the rear nodal point. I did a quick search for such a ray
trace and found one on page two of
> This is a ray trace for a 2.8/19, which means that the
rear nodal point should be behind the last lens element
(19.4 mm in front of the focal plane, going by the specs).
The diagram clearly shows that the rays appear to be
projected from within the
> lens, roughly where one would expect the exit pupil to be
(I couldn't find the exact position of the exit pupil or
rear nodal point for this lens on the Leica site).
> Back to Zeiss lenses. Here is the location of the exit
pupil for a few lenses (most of which were available in
Rollei QBM mount, btw):
> Lens EP at inf. (mm in front of film plane)
> 15 55.7
> 25 54.4
> 35 52.4
> 45 42.6
> 50 67
> 100 75
> Again, the trend is for the exit pupil to be located
within a narrow range for retrofocus lenses, and then to
increase for longer non-retrofocus lenses.
> This seems to confirm my earlier suspicion that retrofocus
wide angles don't cause problems for full frame digital
sensors because light rays at the periphery of the frame are
not hitting the sensor at a particularly acute angle. In the
> above, a 35 mm lens should create somewhat more "deep
well" vigneting than a 15 mm lens. The worse lens for this
should be the Tessar 2.8/45, which is non-retrofocus.
> Since Nikon sells a very similar four element 2.8/45, it
would be interesting for someone to run a test comparing the
vigneting of this lens and of an ultra-wide on film and on a
full frame digital sensor. If my theory is correct, the 45
> suffer from relatively more darkening at the edges on
digital then a much wider lens would.
> I don't have access to a full frame digital, so this is
all speculation on my part.
The rear nodal point, also called the rear principle
point, is, by definition, one focal length from the focal
plane back toward the lens when focused for infinity. It is
the rear principle point which is affected by a telephoto or
reverse-telephoto (retrofocus) lens.
The exit pupil is the image of the stop as seen from the
back of the lens. The location of the pupils depends on the
power of the elements between it and the outside of the
lens. If the overall power is negative the pupil will appear
to be further away than the stop, if positive, it will
appear to be closer. One can measure the location of the
pupils of an actual lens using a camera which is capable of
close focusing. The first step is to focus on a reference
point, say the rim of the lens barrel. Then look at the iris
diaphragm (stop) of the lens and move the camera until it is
sharply focused. Don't adjust the focus of the camera, move
the whole camera. The distance the camera moves to refocus
is the distance from the reference point to the pupil.
Because the pupils are determined by the magnification (or
reduction) of the elements in front of them they can be on
either side of the stop. Both pupils can be on the same side
of the stop. In fact, this is true of Tessars. The front
group of a Tessar is negative, the rear positive, so both
pupils appear to be in the rear part of the lens, i.e., in
back of the stop.
The entrance pupil is the point a lens on a panoramic
camera should be rotated to eliminate relative motion due to
parallax of near and far objects in the picture. However, it
is definitely the rear or second principle point which is of
interest in determining the telephoto or reverse-telephoto
ratio of a lens.
Telephoto lenses have strong positive elements in front,
reverse telephoto lenses have strong negative elements in
front. This condition is not exclusive to these lens types
but is always found in them. An interesting lens is the
Roosinov wide angle lens upon which most modern symmetrical
wide angle lenses are based. It is essentially two
reversed-telephoto lenses back to back. This lens has the
property that the stop varies in effective size with image
angle reducing the light fall off by a substantial amount.
This effect also takes place to some extent in all
Los Angeles, CA, USA