swing lens cameras and focussing distance

I would have posted this in the original thread but it has got long
now and fragmented as to topic. I posted some detailed calculations in
the thread with the subject line "The opposite of a close-up lens?" on
April 6th 2004 two times and April 7th 2004 once and in addition on
April 14th 2004 showing an important link for people interested in
swing lens cameras, namely
http://www.funsci.com/fun3_en/panoram2/pan2_en.htm

I have posted a lot of technical stuff and maths backing it up. I use
swing lens cameras for maybe 10% of all my photography so maybe that
is a high proportion. I have a college degree (BSc) in Physics for
what it is worth and of course I studied Optics as part of that. I am
not claiming to be an expert on Optics, especially since I have earned
my living developing software for many years, but at least I can claim
to have some sort of awareness of the subject and a rudimentary
grounding. These cameras are important to me so I thought that anybody
who followed my calculations might be interested in what I have to say
next.

Here goes...

Now this is theorising and speculation but I personally do not believe
that anybody can make a lens with an accurate focal length. I assume a
0.5% error at least. So taking the Noblex 150, assumed fixed-focussed
at 10.4m, then this has a 50.93 millimeter drum radius according to
the film width and angle specs and there won't be much variation in
the examples they make. Fair enough! They say their lens is a 50.75mm
focal lens lens. You what?!? It could be anything from 51mm to 50.5mm.
There is nothing they can do to stop that unless they adjust the gap
between the front converging lens and the diverging lens behind it to
correct it. Do they do that? Do they do any adjustment at all? I
suspect that the manufacturers of these cameras do make some
adjustment but the most important adjustment, if you have followed my
calculations, is to make sure the secondary principal point of the
lens is exactly on the axis so that the magnification is correct and
so that as the lens swings round then parts of an image will always
stay on the same place on the film surface. Now I challenge anybody
here to find out from manufacturer-published specifications where the
lens is fixed-focussed to on any swing lens camera except the Noblex
range. I doubt you wil find it. And why? It could be that the only
adjustment they are making to the camera is to make sure the secondary
principal point is on the axis so that the image doesn't move on the
film as the lens swings round. But then the focussing distance will be
thrown out when they adjust this. So they will not be able to publish
any figures. So you don't know what distance your camera is
fixed-focussed to and you have got to take pot luck in getting one
that focusses at the distance you have a preference for. In my case I
will not pick bones and will say outright that I want my swing lens
camera to be focussed at infinity because I use it most at that
distance. But secondary principal point on the axis is the most
important thing otherwise you lose horizontal resolution in a big way.
The two are opposites. For me, at least, anybody claiming a distance
focussed to is setting the lens to focus at that distance rather than
doing the far more important thing of getting the secondary principal
point sited exactly on the axis. If Noblex are claiming figures for
focussing distance then they will be adjusting the lens to achieve
that figure and not adjusting the lens to make sure the secondary
principal point lies exactly on the axis which I feel they should be
doing instead.

It is possible to have the best of both worlds but it takes a lot of
adjustment. If you want a lens in a swing lens camera fixed-focussed
at infinity then you have got to make sure that the secondary
principal point stays on the axis at all time but you can adjust the
gap between the converging lens and the diverging lens behind it (for
as Tessar design) to focus it. This will shift the secondary principal
point again off the axis but by a process of adjusting them both you
will finally arrive at the optimum setting. But it seems to me that
the emphasis for the Noblex focussing models has shifted entirely away
from the position of the secondary principal point being on the axis
which to me (with calculations to back it up) is the most important
thing.