Scheimpflug Question

[nomennominatur]

Please bear with me on this question which may be inconsequential or due to a misunderstanding. However, I am puzzled by it and I did nor find a straigtforward answer in the books & documents available to me.

Distinguishing three situations:

(A) The Scheimpflug condition is met when the desired object plane, film plane and lens plane intersect at a common line. This can be achieved on a view camera by tilting the lens and/or back (conventional wisdom so far);

(B) The Scheimpflug condition is not met (i.e. the object plane will not be in sharp focus), if the three planes do not intersect at all;

(C) Now, what about the condition in which, departing from a Scheimpflug setup, a given amount of swing is added to the lens and/or back, resulting in the three planes intersecting at a single point (or do they)?

With situation (C), will there be a plane, a line, or a point of sharp focus? Is there any viable use case scenario for this or would it simply be ‚botched up Scheimpflug‘?

If this has been covered elsewhere, please give a pointer instead of flaming

Kind regards
Mathias

[ Edit: I should have called this thread „Yet Another Scheimpflug Question“ as I have just now realized that the thread title above has already been in use. However I do seem unable to change the thread title in retrospect... ]

[Pere Casals]

(C) Now, what about the condition in which, departing from a Scheimpflug setup, a given amount of swing is added to the lens and/or back, resulting in the three planes intersecting at a single point (or do they)?

With situation (C), will there be a plane, a line, or a point of sharp focus? Is there any viable use case scenario for this or would it simply be ‚botched up Scheimpflug‘?

In the C situation you have two planes in the space

1) The focus plane (determined by the camera position/orientation, bellows extension and tilt/swing)

2) The face of your object you want in focus

Two planes may match, may not intersect if being parallel and not matching, and may intersect in a line.

If the planes are intersecting in the object's face range you will see a line on the object in perfect focus, and a band in DOF. That line can be horizontal, vertical or inclinated.


To understand it easily please follow that reasoning.

With no movements you have a plane of focus perpendicular to the optical axis. If you tilt/swing the lensboard then the plane of focus (in the space/scene) tilt/swings in the same senses. If you tilt/swing the film plane then the plane of focus in the scene also tilts/swings but in the contrary senses.

Of course if you tilt/swing the front then also the circle of image shifts/rises, that can be compensated with shift/rise in the standards...

So, in practice, just tilt/swing the front in the sense you want to tilt/swing the focus plane in the scene, as you probably will use the tilt/swing in the rear then you have to do it in the counter senses.

Regards

[pepeguitarra]

I had the same questions and read and re-read (being an engineer, it should have been easy for me, but...). I really understood this principle after watching this video:

https://www.youtube.com/watch?v=gR4m70xr9mE

[nomennominatur]

Thanks for both replies!

@ Pere Casal: So this would be the case in some of the „pseudo liliput“ images of city vistas that one can sometimes find on the net, where a narrow band of focus (a line + dof) crosses the image either horizontally or at some weird angle. Thus situation „C“ will indeed represent a line of focus.

@ pepeguitarra: I do not want to sound ungrateful or nitpicking, yet although the video is a nice „quick & dirty“ instruction, the theory is reproduced in a faulty manner imho. The Scheimpflug condition, to my knowledge, does in no way necessitate the lens plane angle being the middle divider of the object plane/ film plane angle (as is being asserted at 1:35“). The Scheimpflug condition is fulfilled in every case where the three planes intersect in a common line in space. The condition of the lens plane being the angle bisector of the other two planes is merely one of an infinite number of possible angle divisions. It is the case for one given focal length with one given focussing extension and one given object distance & orientation. Please correct me, if I am wrong.

Kind regards
Mathias

[jp]

Rather than think about three planes intersecting with some unimaginable location of intersection, think of this.. Then you can apply it to focusing with both tilt and swing.

1. Understand the image is upside down in the camera.
2. Understand moving the lens outward focuses closer and bringing the lens inward focuses further.

If you tilt your lensboard down (typical scheimpflug situation), you are making a greater distance (closer focus according to my rule 2) at the top of the camera (and bottom of scene since things are inverted - rule 1)

[nomennominatur]

By the way, one wonders, whether we should not properly speak of the "Carpentier-Scheimpflug" condition. Although Scheimpflug significantly expanded and generalized the concept, the principle, in a limited application, had been filed by Carpentier from Paris before Scheimpflug entered his claim.

For those who like not only vintage camera technique but also vintage documents :

Patent_Scheimpflug_1904-1903_US751347.pdf

Patent_Carpentier_1901_GB190101139A.pdf

Kind regards
Mathias

[rdenney]

The image plane is the "gozoutta" and the film and lens planes are the "gozintas". We set the film and lens planes, and the image plane falls where it does based on those settings. The intersection of the film and lens planes, like the intersection of any two planes, is always a single line. So, the location of the image plane is the dependent variable, and it orients such that that it contains the line that forms the intersection of the film and lens planes (which comprise the independent variables, because we set them).

This image required tilt and swing to place the image plane so that it (approximately) skimmed the stair edges. The tilt moved the bottom of the image plane closer to the camera, and the swing moved the right side of the image plane closer to the camera. The image plane ended up sloping away from the camera to the upper left. But if the camera had been rotated so that the upper left corner was directly at the top of the frame, it would have required only tilt. The image plane always responds to keep the three planes intersecting along a line.



Of course, Scheimpflug is a model built on the assumption that the image plane is flat, which it is for most large-format lenses that are approximately symmetrical, but often is not for retrofocus designs. Getting a tilt/shift lens for an SLR to behave according to prediction when it has a curved focus surface can be a real challenge.

Therefore, your B condition does not exist. There is never a case where the image plane, if it is planar, does not contain the line of intersection between the film and lens planes, if they are planar, unless the lens and film planes are parallel. And when they are parallel, the image is still focused sharply, as long as the focus point is not beyond infinity. We could say that they intersect infinitely far from the camera, but that's not a very helpful description.

Rick "fun with geometry" Denney

[nomennominatur]

[...]
Rick "fun with geometry" Denney

1. This was fun as well as instructive! Thanks.
2. Suitable picture to demonstrate the effect of a combination of tilt/shift on the position of the plane of focus
3. Although I will dearly miss my condition „B“ I am not afraid of contemplating infinity

Kind regards
Mathias „appreciates a good answer“ Elsner

[JMO]

All interesting and helpful discussion here, but regardless of how you conceive of and understand the various "planes" being discussed, it helped me during my newbie learning curve to remember that when confronted with an image composition in the field (e.g., whether a complex cityscape, or redenny's stairwell just above, or whatever) I needed to make the lens (or film) plane movements with my 4x5 camera one axis at a time. In my landscape photography I typically use only lens plane movements, and start with the lens tilt, and then move on to lens swings, as needed. As you apply those movements in succession you should re-check your focus in relation to the various movements you've previously applied, but if you work systematically you don't need to get wrapped up in how to visualize what is going on with the "Scheimpflug Question."

[Doremus Scudder]

A few answers and comments to your questions:

...
Distinguishing three situations:

(A) The Scheimpflug condition is met when the desired object plane, film plane and lens plane intersect at a common line. This can be achieved on a view camera by tilting the lens and/or back (conventional wisdom so far);

Or you can adjust both the lens standard and the back...

(B) The Scheimpflug condition is not met (i.e. the object plane will not be in sharp focus), if the three planes do not intersect at all;

When all three planes are parallel, they do not intersect (at least in Euclidean geometry) and yet the object plane is in sharp focus and can be adjusted anywhere from the closest possible focus of your lens to infinity by changing the bellows draw. So you're premise here is false, unless, of course, you dabble in other geometries in which parallel planes intersect at infinity (there are a few of those...).

(C) Now, what about the condition in which, departing from a Scheimpflug setup, a given amount of swing is added to the lens and/or back, resulting in the three planes intersecting at a single point (or do they)? With situation (C), will there be a plane, a line, or a point of sharp focus? Is there any viable use case scenario for this or would it simply be ‚botched up Scheimpflug‘?

Planes always intersect in a line (lines intersect in a point). Tilting and swinging at the same time simply puts the planes involved at oblique angles to each other. Using both adjustments is useful when you need to place the image plane obliquely, e.g., near and high on the left, near and low on the right, far and low on the left and far and high on the right (hope that's clear...).

Hope this helps,

Doremus