DOF-scale changes with format?

Q.G. said:

"Lots of choices. Pick one." "Define" (in the "pick one" sense). "[...] perceptible by a person"
And still "an objective scale"?

Doesn't work.

The second approach, based on arbitrarily chosen ("let us say") dimensions, does no better.

As you say: "The criteria may be arbitrary"
And there the whole things come falling down. DOF is not an objective thing.


A statement like "The second method is independent of factors such as magnification, viewing distance, photo size, etc." even cannot be a thing related in any way to DOF.
DOF, in essence, is dependent on factors such as magnification, [etc.]

Click to expand...

I think you are, perhaps, confusing the concept of "objectivity" with the concept of arbitrariness.

In most cases a standard is arbitrary, even though it may be objective. For example, a licensing body may decide that a passing grade on a professional certification exam is 70%. That is an arbitrary choice. (Why not 72%?) However, it is an objective criteria. In other words, once the standard is determined one does not have to subjectively discuss whether a score of, say 69% is a passing score.

Choices for standards are almost always arbitrary. For example, a definition of optical resolution based on the Rayleigh criterion is arbitrary. However, it is objective because two different people can perform the measurement and come up with the same result without having to apply subjective judgment.

Similarly, choices in definitions of depth of field that I discussed above may be arbitrary, such as the choice of a standard viewing distance. However, they are objective in the sense that they can be unambiguously measured.

No, no.
I do know what 'objective' and 'objectivity' are supposed to mean.
So that's not it.

Just take one of your arbitrary choices: "perceptible by a person".
Who would that be?

The DOF criteria were set using large panels of people, with the criterion they ended up 'defining' being some sort of statistical mean.
No other way. No 'objective', or even common, thingy in sight.

"The most important thing about DOF, by far, is that it is a perceptual thing. Not an 'autonomous' entity with an absolute dimension. (Despite all the formulae and calculators people like to let loose on it)."

A viewing distance may be measured 'objectively'.
But so that 'arbitrary choice' may be 'objective' (we can set up a thread about whether 'objective' is a meaningful concept at all. But another time perhaps ).
But we're not concerned about viewing distances, but with how an image looks (!) from that viewing distance.
Totally and utterly not objective.

The definitions of DOF and CoC are well known, and anybody can look them up. Simple simple.

For DoF scales DoF gets smaller as the format gets smaller.

In this case you're assuming constant distances while varying FoV/enlargement. So a smaller format has LESS DoF due to what you're keeping constant. Put another way, you're projecting exactly the same image, but cropping it and enlarging it more. So your CoC has to go down, and so does your DoF. For "equivalent image" comparisons the change in distance more than counteracts this, but that's not the comparison you want for DoF scales.

And the dpreview guy's crack about the scales being useless for APS because they're calculated for FF is especially idiotic. Just shoot a stop or so down from the line you're reading on the lens. Duh.

Getting back to the more useful "equivalent image" case, it may just be f-stop and magnification, but for different fields of view you can only match the magnification at a single plane. So with, say, a 15 vs. a 500 (assuming same format), if you keep the subject plane magnification the same (by changing distance), the magnifications some distance ahead of or behind it will be different. For short DoF this is trivial, but stop down enough at non-macro distances and there will be a large difference in DoF.

And I really don't see the point in arguing over the standard DoF definition. Yeah, yeah, it's right about as often as a stopped watch. But they needed some sort of standard, this is at least a marked-on-the-lens starting point to correct from to get to appropriate numbers for your 16 x 20 enlargements or 4x5 contacts or whatever else it is you're doing that isn't a mediocre 8x10.

Roger Krueger said:

Getting back to the more useful "equivalent image" case, it may just be f-stop and magnification, but for different fields of view you can only match the magnification at a single plane. So with, say, a 15 vs. a 500 (assuming same format), if you keep the subject plane magnification the same (by changing distance), the magnifications some distance ahead of or behind it will be different. For short DoF this is trivial, but stop down enough at non-macro distances and there will be a large difference in DoF.

Click to expand...

Field of view is a function of focal length, frame size and distance.
If we assume we are dealing with one frame size only, frame size drops out of the equation as a constant, and field of view is a function of focal length and distance only.
Focal length and distance combined are magnification.

Now it is not true that DOF on film changes with field of view. When magnification and f-stop are constant, DOF will be too, no matter how large or small the frame.
Frame size enters the equation when you want to compare different frame sizes.

Let's keep the distance constant (for simplicity's sake). Then you need a shorter lens to fill a smaller frame the same way you fill a larger frame with a longer lens. That means magnification is smaller, DOF larger on the smaller format film.

When you want to compare same size images, from the same viewing distance (or smaller images from a shorter viewing distance, such that the apparent size is the same), magnification will be the same again, and so will DOF (there are a few side effects that will change the apparent DOF, but i'll ignore those for now).

Or (in short): the field of view is unimportant. Any difference you see when changing frame size is that in frame size.

There will not (!) be a large difference in DOF at all.


The thing that is correct is that unsharpness, the fore- and background blur, increases more rapidly when longer lenses are used. So there is a visible effect of focal length.
But that is outside the DOF. DOF itself is still the same, as long as magnification and f-stop are the same.

Roger Krueger said:

And I really don't see the point in arguing over the standard DoF definition.

Click to expand...

Very true.
The problem with the 'standard' definition we so often encounter in discussions of DOF however is that it is not that of DOF, but of hyperfocal distance.

Q.G. said:

Now it is not true that DOF on film changes with field of view. When magnification and f-stop are constant, DOF will be too, no matter how large or small the frame.

Click to expand...

Sorry, that's clearly incorrect for circumstances where the wider lens is even slightly near (or beyond) hyperfocal. Try it on dofmaster...at f22, 30mm @ 4ft. gives you 36.7 ft. of DoF vs. 300mm @ 40 ft. gives you 7.23 ft. of DoF. But at f2.8 it's only .91 ft. vs .89 ft.

Again, it is just f-stop and magnification, but it's magnification at the point whose focus is being considered, NOT magnification at the plane of best focus. Since you can't keep all of the magnifications the same for different FsoV you get different DoF results.

Roger Krueger said:

Again, it is just f-stop and magnification, but it's magnification at the point whose focus is being considered, NOT magnification at the plane of best focus. Since you can't keep all of the magnifications the same for different FsoV you get different DoF results.

Click to expand...

You really don't.

But i'm so tired of DOF, that i can no longer find the will to explain for at least 6 months.
My fault, i know. I shouldn't have let it tempt me today.

Q.G. said:

"Lots of choices. Pick one." "Define" (in the "pick one" sense). "[...] perceptible by a person"
And still "an objective scale"?

Doesn't work.

The second approach, based on arbitrarily chosen ("let us say") dimensions, does no better.

As you say: "The criteria may be arbitrary"
And there the whole things come falling down. DOF is not an objective thing.


A statement like "The second method is independent of factors such as magnification, viewing distance, photo size, etc." even cannot be a thing related in any way to DOF.
DOF, in essence, is dependent on factors such as magnification, [etc.]

Click to expand...

Technically correct, but not useful as a practical matter. OTO CoC are useful as a practical matter (caveat below), as in for a given format we all know what we are talking about, and therefore can actually "pick one". If someone picks another for the same criteria, one can extrapolate with ease and the subject becomes somewhat objective, at least enough for practical discussion.

I fully agree that none of the other things are related to DoF in the least, only to "apparent" DoF which is purely subjective.
Field of View is a function of format if other factors are the same, and if those factors are the same the actual properties of DoF will be identical. The subjective effect might be quite different however, say between 100mm on minox vs 8x10 inches. The minox frame will offer the same properties within a severely cropped area and appear as a compressed telephoto lens. The 8x10 will be very wide and offer the perspective of and focus characteristic of a very wide focal length. This is because the perspective of a focal length is relative to the format. Folks tend to forget this. To offer the same FoV between formats one must change the lens or distance and focus, and you are no longer comparing anything meaningful concerning DoF, not that it's overly meaningful the other way either.

The CoC for a given format ,focus, focal length, and stop allows one to calculate DoF in a practical way (there's that word again) I seldom calculate for stills because I can see it, but for motion picture where things are very fluid and judgments by eye can be very difficult I use a thing called a SamCine calculator. Basically a slide rule that allows you to calculate DoF for different format/focal/t-stop combinations using CoC as the objective criteria. Understanding the CoC implications for a chosen format is important to using the device. Where the CoC remains the same it will not have the same visual effect as a practical matter across formats, regardless of your subjectivity or objectivity, therefore CoC alone doesn't tell you anything. It's more than just about subjective/objective, it is about practical application, and there isn't some magic bullet formula that you can carry in your pocket that applies across the board. A only a working knowledge and experience can really inform. In other words, get out and shoot. A couple of exposures are worth more than all the words in this thread, mine included.

That's exactly it: a practical matter.

Forget all formulae and calculators, and use your eyes (is photography a visual medium?)

Select a subject. Decide what your picture is about.
Focus on that subject.

Then use the aperture to set how it relates visually to the rest of what is in the frame, judging not by table or scale, but by looking through the viewfinder to see how the sharp vs unsharp balance changes.

Yes, it is difficult when the groundglass gets darker and darker. But then, who promised it would be easy?

And to all those who say their subject stretches from the front lens to infinity, my response would be that such images are boring, and perhaps need not be produced in the first place. Just my view
But if you want to produce images without deciding what they are of anyway, just stop the lens down as far as it goes and be done with it.

But whatever you do, do not even think about DOF as a quantifiable entity. As something that can be discussed and decided upon separately, apart from the particular image your lens is projecting on your screen.
It quite simply is not.

It will always be either too much, too little, or just right.
But how much that is, is different every single time you create an image.
So just use your eyes and look.

Roger Krueger said:

Sorry, that's clearly incorrect for circumstances where the wider lens is even slightly near (or beyond) hyperfocal. Try it on dofmaster...at f22, 30mm @ 4ft. gives you 36.7 ft. of DoF vs. 300mm @ 40 ft. gives you 7.23 ft. of DoF. But at f2.8 it's only .91 ft. vs .89 ft.

Again, it is just f-stop and magnification, but it's magnification at the point whose focus is being considered, NOT magnification at the plane of best focus. Since you can't keep all of the magnifications the same for different FsoV you get different DoF results.

Click to expand...

You are absolutely wrong. Unless you live in a different universe, physics says that you are wrong.

I have posted in the past that while I was studying optics at Kodak, that I raised the question about why when I kept the image size and exposures settings the same, I could not get a better depth of field by changing lenses. The optical engineer started with the depth of field equation and made one substitution. Four lines later the focal length was in every term and then was divided out. The result is exactly as Q.G. stated.

When magnification and f-stop are constant, DOF will be too, no matter how large or small the frame.

Click to expand...

Steve