Depth of field

[Gavjenks]

(Feeling a little uncomfortable)

Yall are making it way too complicated.

Larger apertures and greater MAGNIFICATION = shallower DOF.

^ sentence explains it all, short of an equation.

As with other simple statements about DoF this might be best when accompanied by the condition that it doesn't really apply when comparing between formats, unless both the magnification of the object at the film/sensor and the magnification from the film/sensor to the final image remain unchanged between formats. There are other conditions when it doesn't apply even for constant format size. In fact, it only really applies when the lenses being compared are focused significantly closer than their hyperfocal distance, as Ysarex mentions.

The simple version is not intended to provide quantitative exact comparisons between anything. Neither different formats, nor even any particular magnifications or apertures. It makes no guarantees about linear relationships, either.

The point is simply that if you are standing somewhere, in any given situation, and you want a shallower depth of field, you can get it by either:
A) Using a larger aperture than you have right now, or
B) Using a lesser magnification on the stuff you want blurred than you have right now (which can be accomplished by using a longer focal length and then framing the subject the same way, which makes background objects more magnified than with a smaller focal length, and thus blurrier when not in focus. Or can be accomplished by recomposing the shot and walking closer, with the subject now larger in the viewfinder than before, and perhaps even in other ways.)

Both will always be true when considered relative to your current settings. And for practical purposes, I think this is the smallest number of factors that you can worry about, while still covering all your bases in terms of possible things you can actually change in the field. I.e., if you think of everything in terms of magnification and aperture, you will be able to successfully make all the DOF practical decisions that are available to you and might actually help you get the shot (Things like sensor size and the final size of the print matter, but are not things you can usually change standing in the field composing your shot). Additionally, actually CHECKING the DOF using the DOF preview button is usually going to be much faster and more accurate than attempting to calculate equations. Thus, relative relationships are generally plenty good enough to get you going in the right direction in changing things, then the final decision can be made from double checking the results.

Obviously, intuition / years of experience are even better than calculation OR estimating then checking, but I can't very well instruct people to just use those. They have to come to you naturally.




And to be crystal clear about what I mean by "magnification:"
"Magnification" as I am using it means the distance that an object of given real world length (say, 6 feet) takes up in your viewfinder.
So a 6 foot tall man that takes up only 1/10th of your viewfinder is half as magnified as a 6 foot tall man that takes up 1/5th of your viewfinder.

This is, in my opinion, the simplest way to think about it in the field that actually pretty well approximates the mathematics behind circles of confusion and actual DOF.

Example: changing focal length
If you use a shorter focal length, then the objects in the background (usually the ones that are blurred) will get smaller in the viewfinder relative to their actual size. Thus less magnified, thus deeper DOF. This is true EVEN IF you move to reframe your main foreground subject the way it was before (in this case, the DOF on your subject won't significantly change, because it's magnification is almost identical, but the stuff in the background, more of which is now crammed into the same space, will become noticeably sharper).

Example: changing distance to subject
If you keep the same focal length, but walk halfway to your subject, then they will be twice as magnified as they were before, thus shallower DOF.

Again, do not assume linear relationships or hard and fast rules. These are relative approximations of a more complicated reality. But if you attempt to think about the actual calculations, you will just get frozen and miss the shot. Thus, rules of thumb.




In fact, if you want completely predictable, calculatable reality, you are in the wrong place when it comes to DOF anyway. DOF is a psychological construct that is rooted heavily on our own perception, and can never be fully calculated by any equation known today. Actual DOF of the viewer of a print will depend on many many factors that have nothing to do with your camera or the subject of the photo, and are completely out of your control or knowledge. Things like:
1) Whether the viewer is old or young (eyesight resolving ability)
2) How far a person chooses to stand away from your print when viewing it
3) Whether the viewer is drunk or sober
4) Whether the viewer has cataracts
5) The degree to which the viewer is paying attention to your photo versus glancing absentmindedly at it (the brain will fill in details more or less in different situations)
6) How LONG the person looks at your photo (more likely to fill in imaginary details if they look at it a shorter time)
7) Where the viewer chooses to focus in the image (only a tiny portion of your vision has maximum resolution, and they won't scan the entire image evenly, even if they look at it for 10 minutes). In other words, areas that draw the eye in your image require more sharpness than areas that don't
8) Whether the viewer is looking at the print from an angle or not.
9) Blah blah etc. etc.

 

[unpopular]

In fact, if you want completely predictable, calculatable reality, you are in the wrong place when it comes to DOF anyway. DOF is a psychological construct that is rooted heavily on our own perception, and can never be fully calculated by any equation known today.

Since we're forgoing optical physics entirely, you're both wrong.

It really depends at the rate which the lens is traveling - after all, lens speed is proportional to depth of field. At normal combined velocity of the earth's rotation and orbit, the 35mm will have less dof than the 50mm. Interestingly, photos taken from space have no discrepancy relative to focal length regardless of velocity of travel, as there is no virtual point for which the optical center is referenced, so the dof is fixed at a constant relative to aperture.

the reason why working distance appears to affect DOF is because the optical center shifts, and thus the reference point changes. This modifies the speed which the light inside the lens is traveling relative to outside.

 

[Gavjenks]

Over the top trolls are less fun than subtle trolls =(

 

[unpopular]

Dont forget your macro fluid!

you're not entirely wrong. But the issue isn't so much depth of field, but rather suitable sharpness. Of course the focal point is a two dimensional plane perpendicular to the optical axis which cannot be measured in terms of 'depth', however, we can absolutely determine the diameter of the circle of confusion at any working distance at the film plane of any given wavelength and at any back focus distance. While suitable sharpness varies perhaps from person to person, this variation is not nearly as significant as one might expect; for one thing, the photographs we take are two dimensional, and are not affected by individual eyesight, the camera system records the image regardless of bad eyesight.

Depth of field is also relative to the gradient from the distance of highest resolve to lowest. I am not even sure if this gradient varies from design to design, or if it is fixed into the laws of physics - if it could be controlled, this may explain why some lenses have better separation in DOF, though I suspect this has more to do with edge contrast, resolution, ghosting, and other factors affecting general perceived sharpness.

I agree that depth of field is a bit of an odd concept, but I think it can be measured in terms of relative optical resolution. Absolute suitable sharpness naturally cannot be measured, at least not without knowing the viewer's optical resolution of his or her eye and controlling the distance by which he or she views the image, but because focus falls off immediately off the focal point, and because the focal point is infinitely thin (and their for theoretical in three dimensional space), the entire image is 'blurry' regardless of f-ratio - it is only a finite resolving power which gives photographs the sense of sharpness.

 

[Gavjenks]

But the issue isn't so much depth of field, but rather suitable sharpness.

Yes it is about DOF, because those are the same thing. DOF = range of distance in which you get suitable perceived sharpness in the final viewer's brain, in whatever medium the photo is published. Therefore it cannot, in fact, be measured in absence of detailed information about each and every viewer and their eyesight and how they will choose to look at the image, etc. None of which you have access to as a photographer.

You can approximate it with circle of confusion and the ASSUMPTION of certain standard viewers and viewing conditions, but this is not going to be a quantitatively accurate thing, because circle of confusion (which you can measure exactly) is only one small part of the issue. Most of it is still assumed values, which makes any DOF calculation a rule of thumb more so than anything rigorous.

And if you're already fudging 9/10 variables, you may as well fudge 10/10 variables, in my opinion, by using pure, relativistic rules of thumb like "larger aperture and more magnification = shallower DOF, relatively." In exchange, you are able to make much faster decisions and take more high quality shots in the field as opposed to wasting time calculating things with a level of precision that is just going to be obliterated later by all of your other imprecise assumptions.

It's like measuring out the flour for a cake with tweezers, but then pouring in the milk by gut feeling. Unless you measure EVERYTHING with tweezer precision, you were wasting your time being that precise with the flour.

 

[unpopular]

No, because it's relative from the sharpest portion to the least sharpest portion. If I take a photo with my relatively sharp GAF 50/1.4 verses my relatively soft G-Zuiko 50/1.4, at similar aperture, you wouldn't look at the Zuiko and say it has shallower DOF, you'd say it's overall softer. This is because DOF is the relative size of the CoF from the focal distance, what we perceive as the sharpest portion, to the least, the softest.

I get what you're saying, and there may be some influence by individual eyesight. But it's also a bit like saying that DOF is significantly impacted by camera resolution. It isn't.

 

[Gavjenks]

No, because it's relative from the sharpest portion to the least sharpest portion. If I take a photo with my relatively sharp GAF 50/1.4 verses my relatively soft G-Zuiko 50/1.4, at similar aperture, you wouldn't look at the Zuiko and say it has shallower DOF, you'd say it's overall softer. This is because DOF is the relative size of the CoF from the focal distance, what we perceive as the sharpest portion, to the least, the softest.

I get what you're saying, and there may be some influence by individual eyesight. But it's also a bit like saying that DOF is significantly impacted by camera resolution. It isn't.

I'm not following your logic, sorry. ALL of these things matter for DOF. Which is exactly my point in saying that it's fruitless to attempt to calculate it precisely.

DOF is indeed affected by lens quality. It does NOT affect the "circle of confusion," but it DOES affect depth of field. A bad lens will create a blurrier image of a point source of light at the same distance from the focal plane as a better lens. This is for reasons of spherical aberration, etc., usually, not for the same reason as the circle of confusion changes, but anything that has the end result of narrowing the range of distances of objects in the photo that are perceived as sharp enough contributes to DOF. With the practical result that if you have a bad lens, you should err conservatively on the side of more things being sharp (narrower apertures or less magnification)

Camera resolution also does indeed affect DOF, in some cases. Higher resolution = more reliable and uniform cutoff distances for your DOF. Lower resolution = more random and haphazard perceived sharpness. For instance, a super high resolution camera might have <1% deviation in perceived sharpness for every edge at a given distance from the focal plane. A very low res camera like a cell phone might have very noisy, variable 10% deviation in perceived sharpness between edges at the same distance. With lower res, it's mroe of a roll of the dice if any edge is going to seem sharp or not (based on whether it happens to fall on the boundary between pixels, which are few and far between). Thus, for practical purposes, when using a low res camera, you should err conservatively on the side of more things being sharp (narrower apertures or less magnification)

Whether your photo is going to be hung in a geriatric home or a youth hostel affects DOF, too... the same photo will have on average a greater DOF when hung in a youth hostel.

Edit: also all of the above things are very subjective and imprecise measurements. Thus, all the more reason not to bother with precision in measuring the circle of confusion, because if you aren't going to measure everything precisely, you may as well not measure anything precisely. Precision is a matter of the weakest link in an equation.

 

[unpopular]

DOF is indeed affected by lens quality. It does NOT affect the "circle of confusion," but it DOES affect depth of field.

if this is what you believe, then I am unsure I can even discuss this topic with you. Circle size determines what we call DOF and acceptable sharpness alike. Smaller circle size means greater resolution - smaller aperture means smaller circle size. If the circle size is smaller than the resolution of the recording medium, then it is at maximum absolute sharpness. The circle size will increase at a predictable and measurable gradient off-focus perpendicular to the focal plane. That gradient is depth of field. Sharpness doesn't really have anything to do with depth of field until you take into account resolution, which is again measurable.

Whether your photo is going to be hung in a geriatric home or a youth hostel affects DOF, too... the same photo will have on average a greater DOF when hung in a youth hostel.

This is absolutely incorrect, as the viewer with poor eyesight will view the image uniformly less sharp. I shouldn't have to reiterate that depth is relative to the sharpest perceived portion of the image. If the whole image is blurred, you don't get less DOF, you simply get a less sharp image.

 

[Gavjenks]

This is absolutely incorrect, as the viewer with poor eyesight will view the image uniformly less sharp. I shouldn't have to reiterate that depth is relative to the sharpest perceived portion of the image. If the whole image is blurred, you don't get less DOF, you simply get a less sharp image.

I'm operating under the assumption that the people aren't COMPLETELY blind. Just can't make out quite as fine of detail as a younger person.

If so, then the cutoff for "Acceptably sharp" is higher for the old person. The edges right at the plane of focus exactly will still look acceptably sharp to them, but ones a little further away might not, whereas they might to a younger person. Using made up "sharpness value" numbers:

For a young person
At point of focus: 100
A little further away: 90
further still: 75
Really far: 30

Now let's say the cutoff for acceptably sharp = 80. For the young person, the depth of field is out to the first distance away but no further. Now, imagine an older person, who UNIFORMLY (like you say) sees everything blurrier by 15 units of sharpness.

For an old person
At point of focus: 85
A little further away: 75
further still: 60
Really far: 15

Now, for the old person, the DOF doesn't make it out to even the first distance away. It is narrower than for the young viewer.

I shouldn't have to reiterate that depth is relative to the sharpest perceived portion of the image.

Citation? Because herein lies the rub.

If this were true, then your complaints would be valid (and would apply for the same reasons to the other sub-arguments, so I'm not going to bother addressing each of those individually. It all comes back to this). But I've never heard anybody describe DOF in such a way before. The only definition I'm aware of is "the range in the image of areas which are acceptably sharp to the viewer." Which implies absolute sharpness, not anything relative.

Where are you getting this relative business from?

 

[amolitor]

Does anyone actually use this stuff for anything? It's all very fun to muck around with the arithmetic, but there are so many variables for so little result I have never bothered.

I crank the aperture down enough until there's enough DoF for what I am doing, and then they press the button.

 

[unpopular]

This is absolutely incorrect, as the viewer with poor eyesight will view the image uniformly less sharp. I shouldn't have to reiterate that depth is relative to the sharpest perceived portion of the image. If the whole image is blurred, you don't get less DOF, you simply get a less sharp image.

I'm operating under the assumption that the people aren't COMPLETELY blind. Just can't make out quite as fine of detail as a younger person.

If so, then the cutoff for "Acceptably sharp" is higher for the old person. The edges right at the plane of focus exactly will still look acceptably sharp to them, but ones a little further away might not, whereas they might to a younger person. Using made up "sharpness value" numbers:

For a young person
At point of focus: 100
A little further away: 90
further still: 75
Really far: 30

Now let's say the cutoff for acceptably sharp = 80. For the young person, the depth of field is out to the first distance away but no further. Now, imagine an older person, who UNIFORMLY (like you say) sees everything blurrier by 15 units of sharpness.

For an old person
At point of focus: 85
A little further away: 75
further still: 60
Really far: 15

Now, for the old person, the DOF doesn't make it out to even the first distance away. It is narrower than for the young viewer.

This is such a clusterfutz of misconception, I am not even sure where to begin. First of all, the photographer's eye does not change the camera lens' depth of field (circle diameter), and the fact that wat is acceptable sharpness might differ at distance doesn't much matter when viewing a photograph at arms length, or even the focus screen at eyepoint for that matter. Simply because acceptable sharpness might differ from one person to the next does not influence the circle size of a lens!

The only definition I'm aware of is "the range in the image of areas which are acceptably sharp to the viewer." Which implies absolute sharpness, not anything relative.

Where are you getting this relative business from?

The word "range" not only implies relativity, the concept is absolutely intrinsic! But i would nontheless disagree with the definition, depth of field is not relative to the viewer, it is relative to the image. Anyone can look at any image and clearly determine if it is lacking depth of field, or if it is lacking overall sharpness.

Does anyone actually use this stuff for anything?

I think understanding the basics of technology and science is important to understanding the art and the technique. Photography is unique this way.

 

[amolitor]

Understanding the basics is fine, it's the descent into detail that strikes me as pointless.

I suspect that my background makes me among the most competent members of this forum to really dig in the depth of field, and I just can't be bothered. You have to set too many parameters down and fix them (e.g. print size, viewing distance) before you can make a calculation that's more meaningful than "smaller hole equals more depth, screw it down until it looks OK".

 

[unpopular]

I hear what you're saying. But I think knowing the difference between DOF and a soft lens is pretty important in selecting and understanding your equipment, likewise knowing that DOF in the human eye is irrelevant when viewing an image in the camera provided that you have enough to cover the eyepoint distance!

 

[Gavjenks]

This is such a clusterfutz of misconception, I am not even sure where to begin. First of all, the photographer's eye does not change the camera lens' depth of field (circle diameter), and the fact that wat is acceptable sharpness might differ at distance doesn't much matter when viewing a photograph at arms length, or even the focus screen at eyepoint for that matter. Simply because acceptable sharpness might differ from one person to the next does not influence the circle size of a lens!

That's why I never said anything about the circle size of the lens being affected by the viewer's visual acuity. I said the DOF was affected by the viewer's visual acuity.
Circle of confusion and DOF are not at all the same thing.
You can calculate circle of confusion with precision, and it has nothing to do with anything except distance, focal length, and aperture.
DOF is vastly more complicated, and is a psychological construct, which is influenced IN PART by circle of confusion, but also by potentially hundreds of other variables that might potentially have an effect on a mental state of mind, right down to what the viewer had for breakfast that day.

The word "range" not only implies relativity, the concept is absolutely intrinsic! But i would nontheless disagree with the definition, depth of field is not relative to the viewer, it is relative to the image. Anyone can look at any image and clearly determine if it is lacking depth of field, or if it is lacking overall sharpness.

Range does not imply relativity.

For instance, (100 yards - 300 yards) is a range. It is not relative at all. It references very specific absolute values. (400 yards - 600 yards) covers the same relative distance, but is NOT the same range as (100 yards - 300 yards).

And in the same way, DOF refers to an absolute range: the specific, absolute distance from the focal point in which a particular viewer is unable to visually distinguish a sharp point versus a blurry circle. That ability is not relative to anything else in the image (it is relative to the viewer himself, though). It's a specific number. Perhaps viewer #1, Sam, can just barely distinguish the difference for a point of light that was 3" away from the plane of focus in a print, whereas viewer #2, Samantha, can distinguish the difference only for a point of light that was 6" away from the plane of focus.If so, the DOF of the photo for Sam is 3" in that direction, and the DOF of the same photo for Samantha is 6" in that direction.

This is in no way relative to anything else in the image. In fact, you could cut out just one single piece of the image with a hole punch where the point of light is, and Sam and Samantha would each be able to tell you whether the point is within the DOF for them, without ever even seeing the rest of the photograph! The question is simply "does this look like a single point? Or a circle?" It has nothing to do with anything else in the picture and thus is not a relative question.

Does anyone actually use this stuff for anything?

No.

My whole original point (and still my point) in this thread is that DOF is so subjective that you cannot in fact use any quantitative equations to really tell you very much meaningful information about it.
The only thing that's really practically useful is to have some generic rules of thumb about which sorts of things tend to increase or decrease DOF, all other things equal.
This will then guide you in the field to fiddle with the correct variables and efficiently "guess and check" until you get results you like.
Then after many year,s you will just have an intuition for what to do without having to guess.

But attempting to crunch numbers and precalculate it to high precision is doomed to fail, because circle of confusion (the thing you can calculate) is only a tiny part of what contributes to DOF overall, and scientifically, precision is only generally as meaningful as the least precise part of an equation, which in this case is all the psychological factors of your viewer that you can't calculate well at all.




Conclusion: "Bigger aperture = less DOF" and "More magnification = less DOF" is about as good as you can get. Split that further into focal length or whatever other way of saying the same thing that you want, but at the end of the day, it's only going to come down to vague rules of thumb.

 

[Gavjenks]

In fact, let's go ahead and DO what I was talking about. Here is a point of light from a real life photograph:



Is this point of light within the DOF?

To answer this question, you do not need to know anything else about the photo. You don't need to know how tight of a crop this is. You don't need to know whether this was the sharpest point in the photo or not. You don't need to know anything about it.

All you need to know is "does this appear to me as a single one dimensional point? Or does it look like a circle with actual area to it?" If you perceive it as a point, then yes, this point is within the DOF of the overall image for you as a viewer. If a circle, then no, it is not within the DOF of the overall image for you as a viewer.



Note that your answer might change if you viewed this image from 35 feet away versus 1 foot away on your monitor. At some point, it will appear to be a dot, and at that point, it will actually be within the DOF of the image! Aperture didn't change, nor did focal length or anything like that. Only the viewing conditions changed, and yet DOF changed.

Similarly, if this were 35 feet away and you saw it as a point while you had your glasses on, it would be in the DOF, but if without your glasses, you might not be able to see it as a point, in which case it is not in the DOF anymore.





The circle of confusion in all of these situations would be exactly the same, 0.2765mm or whatever it is. But the DOF would not be the same.