PDA

View Full Version : Depth of Focus - Circle of Confusion Confusion



Wilbur Wong
20-Oct-2005, 08:54
Does the "Depth of focus" vary with focal length? And how does that vary?

In practical terms, what does that mean to a photograph, rather than a theoretical absolutely flat target and perfectly flat film - all parallel to each other.

So say I have a scene with close foreground vegetation, lily pads and vertical water reeds mid distance, and high peaks in the back. Is a longer lens more tolerant of not getting my scheimflug perfectly set up? or is it an advantage to compose to shorter focal lengths?

Short of hitting the diffraction limit, and in practical terms of stopping wind blown flowers, what are the better strategies?

Thilo Schmid
20-Oct-2005, 09:20
Wilbur,
even if you asked for a non-theoretical view, perceived DOF is dependent on exactly three parameters:

1. the magnification ratio at exposure-time

2. the magnification ratio at print-time

3. the viewing distance of the print

You cannot ignore these more theoretical aspects, because they tell you, that there is no simple answer to your question. Your setup should be derived from the desired composition, i.e. the proper perspective (viewpoint) and then the framing (focal length). DOF should not be of concern in this stage. I hate to say this in a LF Forum, but the only parameter which will actually provide more DOF is the reduction of film size (at the cost of other limits, e.g. reduced possible print size). The inpact of the magnification ratio in 1. is higher than in 2., because the nature of the first optical image is different (3D to 2D) than the second (2D to 2D).

Dan Jolicoeur
20-Oct-2005, 09:34
Focal length makes no difference on DOF. Use the focal length that gives you the composition you are looking for. The means of getting a proper DOF are the same and repetitive irrespective of focal length.

Wilbur Wong
20-Oct-2005, 10:08
Am I confusing depth of FIELD and depth of FOCUS as one and the same or as two different aspects?

In a previous thread "super symmar80 xl not sharp" Mark Woods and Michael Briggs mention this in relationship of part of the film being out of the plane of focus.

So my basic question might be is the plane of acceptable focus "fatter" with varying focal lengths? In other words can you move the film further from perfect focus given the same acceptable "circle of confusion" definition, with different focal lengths.

Michael S. Briggs
20-Oct-2005, 11:48
Depth of field and depth of focus are different. Depth of focus is on the subject side and is what photographers usually thing about, e.g., when you ask about a scene with "close foreground vegetation, lily pads and vertical water reeds mid distance, and high peaks in the back" and wonder whether all of these objects will be in focus. The equations to calculate depth of field depend on the focal length of the lens.

For a photographer with a view camera, of course you can use swing and tilt to better align the focus with the objects of interest. Rather than using depth of field tables, I perfer to use the method described in an article on the main page of this site, where you measure the focus spread in image space between the near and far objects that you want in focus and than consult a table. The advantage is that this table only depends on the focus spread and not on focal length nor object distance.

Depth of focus is the allowed tolerance in the position of the film. For non-closeup photography, the equation only depends on the relative aperture (i.e., f-number) and the diameter of the circle of confusion. It doesn't depend on the focal length of the lens, or the distance to the (assumed not close) object.

The two are conjugate, so if you use up your depth of focus by an error in film position, the depth of field won't be where you expect it to be.

People commonly say that moving to a shorter focal lenth lens increases depth of field, but it also changes your view of the scene. If you reposition the camera closer to have the same view, depth of field decreases back to essentially the same as with the original lens.

As Thilo says, for the same f-number, you can get more depth of field by moving to a smaller format and the shorter focal length lens that gives the same view. This might be a solution if you can't stop down further with the larger format, perhaps because the exposure times would become too long.

John_4185
20-Oct-2005, 11:50
Depth of FOCUS is the flip-side of Depth of Field. It is the area in front of and behind the the film that is within an acceptable degree of focus (whatever one may consider acceptable.)

Longer lens == greater degree of depth of focus. Picture it thus: imagine the target circle of confusion as diameter D, a section of a cone. The long lens has a longer cone for D than the wide lens of the same aperture, therefore, there is a greater range near acceptable focus with the long lens than a short lens. Remember, we are looking at the film plane, not the subject here.

Keep in mind that in large format photography the scale of depth of focus is relatively small in terms of outcomes when we struggle with objects near eye level with great degrees of separation in the third dimension. In such cases we may concentrate first on a solution that uses a normal to wide lens and therefore depth of FIELD, and then work with depth of FOCUS on the ground-glass when do rear swings and tilts.

Kinda makes you want to pick up a box camera when you put all this in your head. Let your eyes lead and the rest will follow. :)

Wilbur Wong
20-Oct-2005, 12:08
As a corollary to this, would I be correct to think that longer lenses would be more forgiving of my set up not having the plane of focus set quite exactly, due to the much smaller angle of the "cone" which contains the circle of acceptable confusion with the longer lens?

Michael S. Briggs
20-Oct-2005, 12:10
"Longer lens == greater degree of depth of focus. ...... The long lens has a longer cone for D than the wide lens of the same aperture, therefore, there is a greater range near acceptable focus with the long lens than a short lens."

This is correct if "same aperture" means the linear diameter of the aperture, for example in mm. It wouldn't be correct for relative aperture, meaning f-number, because f-number is focal length divided by aperture diameter. Two lenses of the same f-number will have the same cone regardless of f-number -- the f-number sets the angle of the cone and thus how far you can get from the film and still be within the acceptable circle of confusion. Thinking about aperture in f-number rather than in mm is probably more useful for photographers, and in this case depth of focus does NOT depend on the focal length of the lens.

The equation for depth of focus (non-closeup case) is very simple: t = 2 C N, where C is the diameter of the circle of confusion and N is the f-number. For an relatively fast LF lens, f4.5, and using C = 0.1 mm, this gives t = 0.9 mm = 0.035 inch. Since this is half in front of the film, half in back, it is 2.5X the ANSI figure of +/- 0.007 inch for the tolerance of the depth of 4x4 film holders. If you stop down to f16, t = 1.6 mm = 0.06 in.

robc
20-Oct-2005, 12:13
depth of field refers to the subject(outside the camera) and depth of focus refers to the film.

your chosen acceptable DOField detemines your acceptable DOFocus which detemines the size of acceptable Circle of Confusion.

Altenatively you can work the other way, i.e. decide on the acceptable CoC which determines acceptable DOFocus which determines available DOField.

Either way you also need to consider other affecting factors such as aperture and focal length.

There is only one plane of sharpest focus in the subject which produces very very smal CoC on the film. Anything in front of that plane focusses behind the film and anything behind that plane focusses in front of the film. This is what creates a bigger circle of confusion and when that circle becomes to big it will look blurred when printed. But this also depends on how much you enlarge the print from the negative and what your viewing distance is. i.e. your ability to resolve detail in the print from any given distance.

What is an acceptable CoC is subjective and there are no rules. Some are pedantic about sharpness and other's less so.

Just don't forget that selective focus is a creative method for emphasing some subjects and therefore using bigger CoC either side of plane of sharpest focus can be a desirable thing. i.e narrow DOField is a useful tool for some situations.

N.B. there is no such thing as a plane of acceptable focus. There is only plane of subject sharpest focus and anthing else falls into acceptable DOField or acceptable DOFocus.

Michael S. Briggs
20-Oct-2005, 12:15
That should be "Two lenses of the same f-number will have the same cone regardless of FOCAL LENGTH".

John_4185
20-Oct-2005, 12:25
First, thanks to Michael Briggs, for picking up the detail.

As a corollary to this, would I be correct to think that longer lenses would be more forgiving of my set up not having the plane of focus set quite exactly, due to the much smaller angle of the "cone" which contains the circle of acceptable confusion with the longer lens?

Well, for all practical purposes you really do have to be carefull. The more precise you wish the outcome, the more precise your movements have to be, and our worst cases are where we have a great range of subject depth. It seems the stuff just off the middle suffers the most. :) Figuratively and literally. It's just so whacked. I think it's why so many LF photographers stop down as far as they can. F90 is the last resort of the scoundrel. I don't mean that in a bad way.

Leonard Evens
20-Oct-2005, 14:13
You have got all sorts of information, some of it incorrect. Believe what Michael Briggs tells you. He knows what he is talking about and he states things clearly so there is no ambiguity.

Let me add one additional fact. When you are in the close-up range, usually defined as less than ten times the focal length, then depth of FIELD depends to a high degree of approximation only on the relative aperture, the circle of confusion, and the magnification. The magnification by definition is the ratio of image size to subject size for a subject in the exact plane of focus. So, in a sense, you can say that depth of field in the close-up range is independent of focal length. That means if you choose two lenses of different focal length, you are within the close-up range for both, and you arrange the subject so that in both cases, the magnification is the same, then the depth of field about the exact plane of focus will be to all intents and purpose the same.

This last fact leads some people to claim that depth of field is independent of focal length. They can maintain that because even outside the close-up range, as long as the subject is not too far away, with magnification kept constant, depth of field doesn't vary strongly with respect to focal length. But for many typical scenes, including most landscapes including near and distant subjects, the statement is obviously false.

Thilo Schmid
20-Oct-2005, 23:51
"This last fact leads some people to claim that depth of field is independent of focal length. They can maintain that because even outside the close-up range, as long as the subject is not too far away, with magnification kept constant, depth of field doesn't vary strongly with respect to focal length."

This is obviously nonsense, because the focal length has a direct impact on the magnification ratio - for the same distance. The formula for calculationg DOF (which is rather complicated in case of a GF with applied movements) is first bound to magnification ratio, from which the focal length and subject distance parameters can be derived. Within certain limits you can then omit the subject distance parameter, because it will not have a significant effect. IMO it is better to think of magnification ratios instead of focal lengths. Close up work is then no exception to the rule. This is especially suitable for LF work where "close up" is reached sooner than with smaller formats.

Leonard Evens
21-Oct-2005, 09:09
Theo,

I don't know if we actually disagree.

Note that I said that WITH MAGNIFICATION KEPT CONSTANT depth of field doesn't vary strongly with focal length "if the subject plane is not too far away". (More precisely, I should really have said "with the magnification not too small", but since magnfication is related to subject distance and focal length, they are related statements.) Perhaps it is better to be more precise by actually given the formulas. I address the case where subject plane, lens plane, and image plane are parallel. (With tilts or swings, similar principles apply, but it is more complicated.) According to Jacobson (www.photo.net/learn/optics/lensTutorial) the DOF in front of the subject plane is given by

Nc(1+M)/[M^2(1 + Nc/fM)]

and that in back of the subject plane is given by

Nc(1+M)/M^2(1 - Nc/fM)] (but is infinite if Nc/fM is not less than 1).

Here N is the f-number, c the diameter of the maximum allowable circle of confusion, M the magnification, and f the focal length. This version of the formulas also assumes the pupil magnification is one, which is commonly the case.

As long as Nc/fM is 'relatively small', both of these will be essentially independent of focal length. Of course, just how small Nc/fM will be does depend of focal length. But if you choose the value of f to be the smallest focal length you are going to use and for M the smallest value of magnification, and as a result the quantity Nc/fM is small enough to be ignored in that extreme case, then it can be ignored in other cases you will encounter also.

On the other hand, the presence of the focal length in the formulas means that in many circumstances, even with magnification kept constant, the rear DOF at least can vary rather signficantly. and it may be infinite. I said that is commonly the case in much of landscape photography, and I believe it to be true.

Now, how does subject distance enter into this. Magnification M is related to subject distance u by the formula

M = 1/(u/f - 1)

or inversely

u/f = 1/M + 1

I don't know what you mean by "The formula for calculating DOF (which is rather complicated in case of a GF with applied movements) is first bound to magnification ratio, from which the focal length and subject distance parameters can be derived."
It is clear that the magnification is determined by the ratio of subject distance to focal length and vice-versa. You can't determine the subject distance and focal lengths from the magnification, only the ratio. Moreover, it is also clear from the formulas that the focal length appears explicitly as well as implicitly from M. So even if the magnification is constant, the results depend on focal length.

The trouble with all these discussions is that you can never catch in words just what the formulas tell you. If you have a precisely worded question, I can use the formulas to give a relatively clear answer in words. But, if you change the question, even if it doesn't seem like a big change to you, the answer might be quite different.

John_4185
21-Oct-2005, 09:51
Leonard Evans has a lifetime of renouned expertise in mathematics. I will trust his calculations.

Thilo Schmid
21-Oct-2005, 10:36
"I don't know if we actually disagree"

NO, I did NOT mean to disagree, but to give another explanation why it cannot be true that DOF is independent of the focal length. This is obvious, because the focal length has a direct impact on the magnification ratio at lease from the same viewpoint.

I don't know what you mean by "The formula for calculating DOF (which is rather complicated in case of a GF with applied movements) is first bound to magnification ratio...

I wanted to say that if you derive the DOF formula, you naturally start with magnification ratios and can substitute that with focal legth / subject distance to get other parameters.