depth of field problemsgreenspun.com : LUSENET : Large format photography : One Thread
concerning the depth of field if all movements are zeroed, the near and far limit with reasonable focus will be parallel to the film plane, with plane of focus perpendicular to the film plane, but: 1/ what will be the shape of the "depth of field area/volume"? will it be a cyclinder made of concentric circles ,or just "rectangular block"? 2/ then what will be the "depth of field area/volume" if movement adopted, e.g. tilting or swinging? thanks
-- benz (firstname.lastname@example.org), March 05, 2002
amend: the plane of focus will be parallel to the film plane if no movement adopted. benz
-- benz (email@example.com), March 05, 2002.
With no tilt or swing, the volume is a rectangular block... with tilt, the volume becomes a wedge with the plane of focus tilted, the near plane of focus tilted less, and the far plane of focus tilted more.
-- Glenn C. Kroeger (firstname.lastname@example.org), March 05, 2002.
With the lens plane parallel to the film plane, DOF forms a truncated cone and in regard of the film format a truncated pyramid. When you apply tilt, the cuts at the cone/pyramid are applied diagonal, forming an ellipsis/trapezoid at the intersection.
-- Thilo Schmid (email@example.com), March 05, 2002.
I would have to agree with the cone theory. I would envision a radial slice if you will, the thickness of which would be determined by what f stop is used and what minimum circle of confusion is elected to achieve the desired definition of acceptably sharp. The actual subject plane (the plane on which you focus) would represent a depthless layer one third of the way back from the front surface of the slab and two thirds of the way forward of the rear surface. The thickness of the slab, once again would be determined by f stop and the elected circle of confusion one could live with.
-- Robert A. Zeichner (firstname.lastname@example.org), March 05, 2002.
Pictures illustrating the effect of lens tilt show a rectangular depth of field area when the film, lens, and subject area in focus are parallel and a triangular area when the lens is tilted forward ala Scheimflug (sp? - why couldn't he have been named Smith or Jones?) with the peak of the triangle at the camera and the base at the subject. See, e.g., Adams "The Camera," p.151. However, I don't know whether these illustrations are scientifically accurate or whether they're oversimplified for illustration purposes.
-- Brian Ellis (email@example.com), March 05, 2002.
I think the .mov files Brian is referring to can be found at: http:// home.fox.nstn.ca/~hmmerk/
I found the articles, etc. of Harold Merklinger linked to that page to be very interesting as I started in LF.
(And thanks Thilo and Robert, I'll have to think about cones/ ellipsoids. Don't quite follow what you wrote, and my primitive understanding of DOF with movements was in terms of planes.)
-- James Lewis (firstname.lastname@example.org), March 05, 2002.
errrr, that's: http://home.fox.nstn.ca/~hmmerk/
(sorry, an extra space sneaked into the address I posted before)
-- James Lewis (email@example.com), March 05, 2002.
In reality, that "peak of the cone" or hinge-point, when using tilt, is at a point below the camera lens. Think of two panes of glass, parallel with the front element of the lens and locked with the lens. One pane at the near focus plane and one at far. Everything between the panes of glass will be in focus. It is more like a sheet who's thickness can be varied.
As you tilt and/or swing the lens, the panes move exactly with the lens.
So, as you tilt the lens forward/down, the panes "lay down" (with the tops moving away from you) in precise alignment with the lens - watching the lensboard gives you a good indication.
The idea that the shape of the depth changes from a rectangle to a wedge/cone, whatever, is misleading.
The depth of the space between the invisible sheets of glass does not vary as the tilt changes - it just seems that way because it gets intersected by something -(the ground plane). If it were happening in space without the ground to interfere, it would be plain to see. And you may someday wish to limit focus to a few points in space....
-- Matt O. (firstname.lastname@example.org), March 05, 2002.
Depth of field doesn't have a sharply defined 'cut-off' point or plane. The focus gradually shifts from sharp to unacceptably blurred away from the plane of focus.
There can be no fixed 'volume' to the focused space, since it depends on what an individual finds acceptably sharp, and on the amount of detail in the subject matter, the degree of enlargement of the final print, and its viewing distance. It also depends to a degree on the quality of the lens. Some lenses suffer from curvature of field, which would make the volume bulge or bow toward or away from the camera.
If you can imagine a space which gradually fades to nothing at two ill-defined near and far boundaries, then that's about the closest you'll get to visualising the focal volume. The image circle of the lens will define the circumference of this nebulous conical section.
-- Pete Andrews (email@example.com), March 07, 2002.