How are depth of field and depth of focus related?

greenspun.com : LUSENET : Large format photography : One Thread

How are depth of field and depth of focus related? I know that if I reduce depth of field, I also reduce depth of focus but I don't know if the relationship is proportional or not...

I realize this is probably a basic optics question but the answer is escaping me for now and any input will be appreciated.

Thanks!

-- Jeffrey Goggin (audidudi@mindspring.com), November 13, 2000

Answers

It is proportional but not linear (as far as I can see). I've included the one link that I've found which mentions "depth of focus" - I'm very interested in this also and would appreciate any links that others might contribute.

http://www.smu.edu/%7Ermonagha/mf/mfbest.html

Wayne

-- Wayne (wdewitt@snip.net), November 13, 2000.


Jeff... Depth of focus equals fstop x cc desired on film x 2. Just like Depth of field, the greater the f stop, the greater the amount the amount of Depth of Focus. Fast lenses such as 2.8, at .03mm cc on film will only allow for a Depth of focus of .168mm. Most roll film holders would struggle with this requirment due to non straight film path. Maybe this is why all the real fast stuff is left to 35mm :-)

-- Bill Glickman (bglick@pclv.com), November 14, 2000.

It's basic trigonometry.
The exit pupil of the lens, for practical purposes this can be taken to be the physical diameter of the aperture, projects a cone of light on to every point of the film. Now, if the film is moved forward or backward from the apex of this cone, then the light will no longer come to a point; it'll form a circular patch of light on the film instead. This circular patch is aptly called a circle of confusion, and its size can be easily calculated from the diameter of the exit pupil, the distance of the focal point from the lens, and how far the film has been moved from the point of best focus. Also, by working back from the diameter of a given circle of confusion, we can find out the depth of focus.
The real crux of the matter is to determine how big the circle of confusion can be before the image becomes unacceptable. This depends on format size, degree of enlargement, and personal standards of what is acceptably sharp. For LF work, a figure of 0.02mm is about right (other suggestions are welcome).
Skipping some basic geometry, it turns out that only the numerical aperture of the lens, f/16 for example, needs to be known to calculate the depth of focus when the lens is focused at infinity. D (the depth of focus) = C(diameter of CoC) x N (aperture number). In the case of f/16 and 0.02mm CoC, D = 0.32 mm.
It's easy to see that the depth of focus is equal in front of, and behind the film plane, so this is plus or minus 0.32 mm, making the total depth of focus 0.64 mm.
So, the film or camera back can be displaced +/- 0.32 mm before the image will be noticeably blurred. This is completely independent of any depth of field, but will add to, or subtract from, any blurring due to depth of field allowance.
When the lens is focused closer than infinity, then the real effective aperture should be calculated.

-- Pete Andrews (p.l.andrews@bham.ac.uk), November 14, 2000.

Thanks guys - basic Trig.! Why did I think that it should be more complicated? In my case I was concerning myself with the relationship of depth-of-field vs. depth-of-focus between lenses of different focal lengths while not grasping the obvious.

-- Wayne DeWitt (wdewitt@snip.net), November 14, 2000.

The article by Steve Peterson, linked from http://www.cs.berkeley.edu/~qtluong/photography/lf/fstop.html derives the relationship. This is also available in Bob Wheeler's notes.

-- Q.-Tuan Luong (luong@ai.sri.com), November 14, 2000.


Pete's reply highlights that the depth of focus can be extremely shallow in consequence of which, the accurate position of the film plane becomes of extreme importance. Photographers generally seem obsessed with the sharpness of lenses and spend many hours and thousands of dollars in the search and purchase of the glass utopia. Yet, few take the time and make the effort to calibrate their cameras and film holders. An indication of how little importance is attached to the lowly film holder is the fact that the brand of filmholders which (in my tests) scored worse in calibration is the one that scores best in sales. Saving $20.00 in a film holder seems hardly worthwhile after spending thousands in lenses, or does it?

-- Julio Fernandez (gluemax@ora.auracom.com), November 15, 2000.

Funny you should mention this ... the reason I asked the question in the first place is because I'm optically calibrating the position of the ground-glass assembly in one of my cameras and I had some doubts about the conventional method of doing so.

Intuitively, it seems to me that rather than shoot a target composed of several closely spaced objects and seeing which one appears to be sharpest on film, a better approach may be targeting just one object and varying the focus +/- a few thousands of an inch (using accurate dial indicators or other measuring devices) for each shot and seeing which of the series is the sharpest. This way, you'd also determine exactly how much the ground-glass needs to be adjusted without going through another round (or two, or three) of testing...

Is this procedure also correct or have I overlooked something here?

-- Jeffrey Goggin (audidudi@mindspring.com), November 15, 2000.


Temperature, humidity, film type, the position of the camera (and probably the moon) will cause film to have an uneven surface that will vary from holder to holder regardless as to how close they are to spec. Measuring the plenum's surface depth with a dial-indicator depth gauge through a plate glass in which I've drilled several holes is the method that I use (with the back detached from the camera and the holders inserted into the back). Shooting at targets introduces the "film variable" which is uncontrolled in a standard (non-vacuum) holder. Since 99% of what we record does not fall at the plane of focus (unless you're doing repro of 2-dimensional objects) we're depending on depth-of-field anyway. I just make sure that I don't have any dog holders. I also make sure that I don't have any holders with a plenum shallower than my groundglass +.007" (film can't lie any flatter than +.007" above the plenum's surface).

-- Wayne DeWitt (wdewitt@snip.net), November 15, 2000.

Wayne: you make some good points. Additionally, colour film and negative film are not the same thickness. Velvia, Provia, etc are 0.009" thick and Vericolor and NPS is 0.007". Some variations can be expected in the 4th decimal place due to humidity. Because in holders the film sits in the retaining groves rather than against a pressure plate as in the Linhof holders the film thickness will be a contributory factor as to the film's position with reference to the GG. Unfortunately, Linhof no longer makes holders.

The difference between B&W and colour may not be substantial enogh to be for concerned about, but it would be nice if the manufacturer's colour divisions would send the 'uncolour' people an e-mail with the message: stop being so stingy and standardize! On a second thought, better not: in the corporate world the stingy always get the upper hand and we will end up with skinny 0.007" colour film..

-- Julio Fernandez (gluemax@ora.auracom.com), November 15, 2000.


Julio - Thanks, I thought Velvia felt thicker - I just never measured it. I once read that the original Tech Pan was on a .004" base - if true in LF that must have been a nightmare!

-- Wayne DeWitt (wdewitt@snip.net), November 15, 2000.


My earlier post referring to a "plenum" should have said "septum" (confusing my phototography and audiophile terminolgy - sorry). Glad no one dinged me.

-- Wayne DeWitt (wdewitt@snip.net), November 16, 2000.

Wayne, If you want to have a real "nightmare" sometime, try cutting a small hole in the center of an unexposed sheet of 8x10 and loading it in a holder without the slide. Then, by looking at the hole and the surface of the holder directly underneath you can get some idea of how much the film separates from this "calibrated" plane when you do various things that simulate actual use, such as: letting it hang film-down for a few minutes or blowing hot breath on it (simulating the condensation that would occur when a cold holder is used in a hot bellows). Different films will react differently but I've seen film "pop" as much as .100". You can use the information gathered by this test to fudge your focus a little when you work in these problem situations.

-- Bruce Wehman (bruce.wehman@hs.utc.com), November 16, 2000.

I came upon some interesting data today. I was sent a very old wooden holder which contained two sheets of Ansco film. Behind each sheet was a black paper. I wondered (and still do) if the paper was for spacing or anti-halation.

-- Bill Brady (wmbrady@olg.com), November 16, 2000.

Moderation questions? read the FAQ