Calculating the focus spread at the image forming plane?

[l2oBiN]

It would be great to have a good explanation and a few examples how does one calculate the focus spread at the image forming plane. Also how does this change with tilts/sweings?

[ic-racer]

The focus spread is usually obtained empirically with a millimeter scale on your focusing rail. If you need to calculate it from your subject distance you can use the thin lens formula. It will always decrease when you tilt and swing.
If you want the entire explanation in great detail check this out, you can get these excellent books by Harold M. Merklinger from his website for free: http://www.trenholm.org/hmmerk/download.html
Understand what is in those books and you will be another 'resident expert' on the subject.

[Jeff Conrad]

The “focus spread” is related to the depth of field; the relationship is described in my paper Depth of Field in Depth (PDF), Equation 34. But in most practical photography, there’s little point in calculating the focus spread—as ic-racer said, it’s something you determine by focusing on the far and measuring the distance between the standards, focusing on the near and again measuring the distance between the standards, and then calculating the difference between the near and far distances. The process is much easier if you have a scale attached to the bed or focusing rail, and you can get greater precision if you also put a scale on the focusing knob, as described in Leonard Evens’s paper Some View Camera Formulas (PDF).

The required aperture to achieve the desired DoF is proportional to the focus spread, so anything you can do to reduce the focus spread will allow use of a larger aperture, reducing exposure time and possibly allowing the lens to be used at the aperture at which it is sharpest. For some scenes, especially landscapes, the way to reduce the focus spread is to use tilts, swings, or both. How do you know if you have the right tilt and swing? Try a few settings and see which gives the smallest focus spread—it’s described quite simply (but thoroughly) in QT’s article How to focus the view camera.

[Leigh]

the relationship is described in my paper [u l=http://www.largeformatphotography.info/ articles/ DoFinDepth.pdf]Depth of Field in Depth[/u l]...

That link is broken due to embedded spaces. I trashed the url html tags so you could see the text.

Here's the correct link: http://www.largeformatphotography.in...DoFinDepth.pdf

- Leigh

[Leonard Evens]

Jeff Conrad explained it well.

Normally, you determine the focus spread between where on the rail, at full aperture, you find your deisred near point and your desired far point. The reason for doing that is to determine the proper aperture you need to encompass that focus spread and get the desired depth of field. Jeff gives references for how to do that.

You could in principle calculate what the focus spread on the rail would be after specifying a near and far distance in the subject space, but it would be subject to several uncertain assumptions, and I don't really see much use for such a calculation. Unless you use an accurate rangefinder you wouldn't ordinarily know thsoe distances accurately enough. And if you did, you could determine the desired aperture without finding the focus spread.

After a tilt you can also measure a focus spread above and below the exact subject plane by seeing where on the rail an "upper point" is and where a "lower point" is. This focus spread would again be used to determine an aperture to provide adequate dof, as above. If you use the same formulas as for the near-far point method, you will end up using a slightly larger aperture than you actually need. But in most cases the error is negligible and in any event it will at worst only give you slightly more depth of field. It is seldom true that you want to limit your depth of field to an exact specified amount.

If you really want to understand these matters, you should read Jeff's article. If you are really a glutton for punishment, you could also try my articles
http://www.math.northwestern.edu/~le...s/pages/vc.pdf
and
http://www.math.northwestern.edu/~le...s/tilt_dof.pdf

[Leonard Evens]

Jeff Conrad explained it well.

Normally, you determine the focus spread between where on the rail, at full aperture, you find your deisred near point and your desired far point. The reason for doing that is to determine the proper aperture you need to encompass that focus spread and get the desired depth of field. Jeff gives references for how to do that.

You could in principle calculate what the focus spread on the rail would be after specifying a near and far distance in the subject space, but it would be subject to several uncertain assumptions, and I don't really see much use for such a calculation. Unless you use an accurate rangefinder you wouldn't ordinarily know thsoe distances accurately enough. And if you did, you could determine the desired aperture without finding the focus spread.

After a tilt you can also measure a focus spread above and below the exact subject plane by seeing where on the rail an "upper point" is and where a "lower point" is. This focus spread would again be used to determine an aperture to provide adequate dof, as above. If you use the same formulas as for the near-far point method, you will end up using a slightly larger aperture than you actually need. But in most cases the error is negligible and in any event it will at worst only give you slightly more depth of field. It is seldom true that you want to limit your depth of field to an exact specified amount.

If you really want to understand these matters, you should read Jeff's article. If you are really a glutton for punishment, you could also try my articles
http://www.math.northwestern.edu/~le...s/pages/vc.pdf
and
http://www.math.northwestern.edu/~le...s/tilt_dof.pdf

I've reread the second of these articles and corrected two or three typos. The first part of the article, while not light reading, is not too difficult to follow.

The article also explains why relying completely on what you see on the ground glass can, in some circumstances, be miisleading. The point is that if you use a high powered loupe to view the image on the ground glass, you in effect reduce the size of the largest circle of confusion you are willing to accept. That will have a significant effect on resulting depth of field calculations. You will end up choosing too small an aperture as a result. (In the above quoted material, I messed this up, saying "slightly too large an aperture" where I needed to say "too small". ) That will be an innocuous error as long as you don't have to increase the exposure time too much to compensate. But, in my photography, at least, subject motion, such as rustling of leaves in the wind, often means I want to keep the exposure time as short as possible, so underestimating the aperture size can be problematic.