Question about front axis tilt on ebony 45s

[Julian Boulter]

Hi,

I have been using my new Ebony 45s with a Wista lens board with off center hole. As I understood providing the red dots of the rise / fall mechanism on the front standard are lined up then the lens will tilt on its optical axis.

However when I try focussing on something at the center of the ground glass and then perform a front axis tilt (forward) to bring the foreground into focus I notice that the center is now completley out of focus. Whereas if I try the same routine and use the back axis tilt (backward) to bring the foreground into focus the center of the ground glass remains in sharp focus as I would have expected.

Is this due to the position of the off center hole on the wista lens panel? I was suprised to find this as I understood that the Wista panel is interchangeable with the linhof type panels which the Ebony takes.

Thanks

Julian

[Ole Tjugen]

No, it's more likely to be due to the difference between the location of the tilt axis and the locatio of the optical node of the lens. Very few, if any, lenses have the nodal point exactly at the mounting flange.

[Leonard Evens]

Can you tell us what lens you are using?

If you tilt the lens (front standard), the tilt axis should pass through the front principal point in order to minimize any shift in focus. If the lens is positioned so that the lens axis does not pass through the tilt axis, as would be true for an off center lens, that won't be the case. Also, for some lenses, such as telephoto lenses, the front principal point may be well in front of the tilt axis.

Note also that it is not necessarily the center of the field which will remain in focus. It is the point where the original (untilted) lens axis intersected the gg. If there is rise/fall/shift of either standard, that may not be in the center of the field.

Actually, even if the tilt axis passes through the principal point, there will be a small shift in focus. I estimate that, for a 150 mm lens tilted 10 degrees, you would have to shift about 2 mm along the rail to keep a point along the original lens axis in focus. If there is a signficant distance between the front and rear principal points, that would induce a further shift, but in the opposite direction.

I'm not sure what you mean by back tilt. If you are referring to the back standard, and the tilt axis passes through the center of the field, there is no mystery about why that should remain in focus. The point where the image comes to focus on the gg is not changed. If you mean that tilting the front standard forward changes the focus but tilting it backward doesn't, that would indeed be mysterious.

[Julian Boulter]

Hi Leornard,

Nice to hear from you again. I have tried this with my 2 lenses, a 75mm Nikkor SW and a 150mm Rodenstock APO S, same result with both of them.

On the Ebony 45s when you raise or lower the front standard there is no change in the position of the tilt axis, that is the tilt rises and falls with the standard so nothing changes relative to the lens.

By back tilt I meant on the rear standard so if I tilt the rear standard axis tilt back to bring foreground into focus the center of the screen stays sharp, not so when I tilt the front standard axis tilt forward. In fact its very difficult to set the plane of focus using the front standard.

From what you and Ole are saying it seems that using the front axis tilt by focussing on the center of the screen and expecting fewer iterations of refocussing following tilt is hit and miss depending on the lens used.

Thanks

Julian

[Leonard Evens]

Julian,

I'm glad to hear that the tilt axis rises or falls with the standard. That is what happens with my Toho FC-45X, although I don't have axial tilt.

I'm also glad to hear that you were talking about tilting the rear standard when you said the focus didn't change. That is exactly what one would expect.

I have a Rodenstock f/5.6 150 mm Apo Sironar S. Is that the one you are referring to? I haven't done any careful testing, but I doubt very much that the front principal point is far from the lensboard. The rear flange focal length is 147 mm, 3 mm short of the focal length and I would expect the front principal point to be somewhere in front of that. If it were 3 mm in front of the tilt axis, if my calculations are correct, you would have only about 0.04 mm shift in the focus position along the rail with a 10 deg tilt. You shouldn't notice that.

I think Ole is talking about lenses in which the front principal point is a lot further away.

How off center vertically is the lens relative to the tilt axis? You ought to be able to estimate that with a metric scale. The rule of thumb is something like the following. If the front principal point is a certain distance horizontally in front of the tilt axis, divide that distance by 72 to get an estimate of the horizontal shift along the rail from a 10 degree tilt. If the principal point is displaced vertically a certain distance from the tilt axis, divide that distance by 6 to get an estimate of that horizontal shift.

If the center of the lens is 12 mm from the tilt axis, then a 10 degree tilt would result in roughly a 2 mm shift, which should definitely be noticeable. With an f/5.6 lens, the amount you could shift without seeing a difference is probably about 0.5 mm.

[Ole Tjugen]

Leonard, it's not the shift that does it. The 3mm offset of the Apo Sironar S is enough that the central part of the image will no longer be in focus after tilting, even if it stays in essentially the same place.

The only way to demonstrate this is to play around with a Carbon Infinity, I know of no other camera that has adjustments for nodal point offset! You're welcome to drop by and have a look at mine.

[Leonard Evens]

Ole,

I've been rethinking this and analyzing it from a variety of viewpoints. To avoid confusion, let me specify that I am referring to the image of something that is along the original lens axis, i.e. the horizontal line of sight, which is in focus to start with. That image point may or may not be in the center of the field. It would depend on the positions of the standards. It still seems to me that even if the tilt axis is off a few mm from the principal point, that, in iteslf, shouldn't make a noticeable difference in whether such a point stays in focus. But, as I noted earlier, it is true that there should be a noticeable shift in focus, even if the tilt axis passes through the prinicipal point, just because of the tilt. For distant points (where the focal length is much less than the subject distance), and modest titls, there would be a horizontal shift in where the image of such a point lies of approximately

one half the focal length times the square of the tilt angle in radians.

This formula follows directly from the lens equation and making some simple approximations. I would be very surprised if the behavior of a large format lens departed significantly from the lens equation, particularly for points not that far from the lens axis.

Suppose the tilt angle is 1/6 radian, which is close to 10 degrees. For a 150 mm lens, that would give (1/2) x 150 x (1/6)^2 = 150/72 or a bit more than 2 mm. With an f/5.6 aperture, you would certainly notice such a shift.

Since the tilt itself shifts the focus, I don't see how you can separate the different effects without making precise measurements on an optical bench, and even so, all you could ever measure would be the total effect, not that from any individual component. Of course, by varying the parameters and making enough different measurements, you could see how close to theory the lens behaved.

I don't have axial tilt, but I would find it surprising if the refocusing needed after titlting were really a serious imepediment to trying to visualize what what happening on the gg.

[Julian Boulter]

Guys,

I'm going to experiment over the weekend with both lenses and let you know the results.

Leonard with regards to how off center the lens is vertically relative to the tilt axis I would say its at least 12mm but I'll measure it and let you know.

Thanks

Julian

[Andreas]

My Technika has also axial tilt. I once tried to center one of my lenses (not a telephoto) with that axis as perfectly as possibly. The picture would then stop moving up or down when tilting the lens, but the subject would not stay in focus on the centerline of the ground glass. It appears that the actual focal length measured on that center line shortens when the lens is tilted. In some way we all have a zoom lens on our LF camera.

[Leonard Evens]

Andreas,

If you centered the lens so you didn't notice any significant vertical shift, then the explanation I gave above is probably the reason you noitced a shift in focus. But the focal length doesn't change. If it did, then other things such as the magnification would change, and that doesn't happen. It is just that the focal point is now in a different position because of the tilt. Using a modified form of the so called lens equation, which takes into account the tilt, I can calculate exactly how much that central image point shifts back from the lens. The exact equations are messy, but if the focal length is small in comparison to the subject distance and the tilt angle is reasonable, an estimate for that shift is, as I mentioned above, one half the focal length times the square of the tilt angle in radians. This can easily produce a noticeble shift of a few mm along the rail.

The situation is also complicated because you also have to worry about (1) the localtion of the principal points and (2) the distance from the tilt axis to the lens axis. For most lenses, the principal points are close enough to the lens board so type (1) shifts are very small. If the lens is not close to being centered on the tilt axis, which is apparently the case for Julian, that could also produce a noticeable shift. The extreme case of that is base tilt, where the tilt axis may be 100 mm or more from the lens axis, whence the shift would be something like 6 mm for a 10 deg tilt. In some cases, it may be possible to set things up so the inevitable shift described above is laregly cancelled out by a type (1) or (2) shift.