Hello from France
Actually the only thing that you can do, provided that the optical axis is perfectly set perpendicular to the film-plane, is to define the best infinity stop not at full aperture but at the actual f-stop used to take pictures e.g. f/16 or any f-number that will be your actual working f-number. And yes, this is a pain since doing so, you won't be able to rely on what you see on the ground glass for focusing at full aperture. May be that you'll have to re-label your distance scales, since aiming at the best trade-off at the working f-number for best sharpness on the whole field at every subject distance will yield something different from what the manufacturer has engraved for you!
If there is field curvature you'll have to make a compromise between the centre and the edges, nothing to do much about that, except using a curved film holder; a cylindrical film path is not a sphere, but this trick was used in the good old days with cheap rollfilm 6x9 amateur cameras fitted with a meniscus lens to somewhat improve image sharpness in the edges.
At least by defining at the real working aperture what is the best mechanical infinity stop, you'll get the best possible trade-off in the final image for distant objects, if you have some "best focus" shift between full aperture and f/16.
With ultra wide angle lenses, even modern ones, we have to forget about the Gaussian model of paraxial rays: actually the best focusing setting for far distant objects is no longer the fixed, gaussian focal point, independant of the actual f-stop setting, but another point located close to it and which changes when the aperture changes. And if the manufacturer has engraved the distance scales according to classical gaussian formulae (Newton's formulae, by reference to the gaussian focal point) then those scales could be wrong ...
Using a 75 mm focal length on a 6x17 format means that you are reaching the limits of the 100-105° of angle of classical designs like the super angulon.
I do not know what is the actual image size delivered by your 6x17 film back, but if we assume 56x170 mm, the diagonal is close to 180 mm ; this makes a total angle of 100° in the corners so you are not far from the absolute limits of the image circle of a classical 100-105° 75 mm wide-angle view camera lens.
With a 90 mm lens and a maximum angle of 90°, the problem would certainly be less critical, but this suggestion does not solve your real problem
And regarding DOF problems, as you know, DOF is not intrinsic to a given lens design, a given focal length nor a given film format.
DOF is actually defined by the final sharpness criterion you define yourself for the final print.
DOF is not defined by others for you and is certainly not defined by DOF calculators where you have to use some pre-defined settings that you can not change.
Whether you'll show contact prints of size 5.6 x 17 cm, of murals of 56 cm by 1m70, your definition of DOF will not be the same!
Now comes and interesting question: if I use a classical DOF calculator anyway, which setting or which format should I use?
If I us a circle of confusion (CoC) similar to the one in use for the 13x18 cm film format [Europeans like you and me never use formats in inches,
natürlich ], hence I'll choose something like D/1720 where D is the diagonal of the image. In 13x18cm this diagonal is 210 mm but here we have only 180 mm; so let's go for c = 180/1720 = around 100 microns, 0.1 mm.
Now let have a closer look at our mural print, 56x170 cm. Actually if I look a this image, my eyes will scan the format horizontally and I can consider that I have in fact 3 square images stitched together side to side, like if I had stitched 3 images taken with a 6x6 camera.
Hence the other approach for the choice of c is to take the same value as for 6x6 photography. On the Rolleiflex TLR the good old masters in Braunschweig have chosen c=50 microns for the DOF scales that one can see near to the Rolleiflex fucusing knob. And using a 75 mm you are using the same focal length as all 3.5 Rolleiflex cameras! 75/1720 yields c=43 microns.
So it means that even the estimation of what computed DOF can be fuzzy in the 6x17 format, with a degree of freedom of at least a factor 2 in the choice of the circle of confusion!!!
Bookmarks