DOF determined by focal length, aperture and...?

[Christoph13]

Advertisement (gone after registration)

Hi all

 

so what exactly other than focal length (FL) and aperture determines depth of field (DOF)?

Some people are quite vocal about these two parameters being the only two determinants, meaning that e.g. two different 50mm lenses at f/4 will always and necessarily have the same DOF.

 

Just to make sure we are all on the same page: I am not talking about perceived DOF. As we all know a very sharp lens dissipates more abruptly into background blur than a lens that is less sharp, thus giving the (false) impression that the DOF of the latter is larger.

 

So, when DOF is handled more objectively as diameter of circle of confusion, what parameter other than FL and aperture is at play here? I haven't done any meaningful tests but am inclined to think that there is more.

The stage is yours

 

Christoph

 

PS: Knowing this will not make anyone take better pictures, but I like to be educated nonetheless.

[andybarton]

Sensor size. Small sensor = more depth of field

[jaapv]

In my book DOF is an illusion in the eye of the beholder.

[Christoph13]

Thanks for reminding me, Andy, I was going to include that as well (although that, too, is a perceived effect, but let us discuss that elsewhere)

 

So, refined scenario: position of camera and subject fixed, only the lens is allowed to vary.

[Christoph13]

In my book DOF is an illusion in the eye of the beholder.

 

Not according to the definition above. You are entitled to your opinion of the artistic effect, of course, but that is not the point of my question (see PS)

[jaapv]

It is only due to the imperfect resolving power of your eye. In reality it does not exist.

[Christoph13]

Advertisement (gone after registration)

My question is about capturing images of real objects using optics not the objects themselves, isn't the whole forum about this as well? True, there is no DOF in real life but then again, if this was what it's all about, why are people especially of our denomination prone to spending 10k+ on a single lens with a particularly narrow DOF? Limitations can be rather artistically rewarding.

 

So, yes, you may be right but my question still stands.

[andybarton]

A question I have often asked myself, but have never been able to answer myself is, how large an area does the human eye (and brain) actually focus on, at any particular moment in time?

 

Look out of the window at a distant object, or at a printed page in front of you. How much are you actually really looking at?

[pico]

Just to make sure we are all on the same page: I am not talking about perceived DOF.

 

Without considering perceptual phenomena, then it is all mathematics and DOF is equal between lenses of the same FL, aperture, given subject distance and circle of confusion.

 

Andy: A question I have often asked myself, but have never been able to answer myself is, how large an area does the human eye (and brain) actually see, at any particular moment in time?

 

If I recall correctly, and ignoring the blind spot which is perceived only under special circumstances, our field of view is about fifteen degrees, however the eye is always vibrating, or moving and the brain is always filling in areas (persistence) so that we see about 95 degrees wide, 60 degrees up and 75 degrees down. Motion in the peripheral area is noticed/perceived but not really seen (in the sense that you cannot read text at the points of peripheral vision).

 

--

Pico - at the moment replacing the ground glass on an 8x10" camera. Damn shims fell out. I hate wood cameras!

[andybarton]

If I recall correctly, and ignoring the blind spot which is perceived only under special circumstances, our field of view is about fifteen degrees, however the eye is always vibrating, or moving so that we see about 95 degrees wide, 60 degrees up and 75 degrees down. Motion in the peripheral area is noticed/perceived but not really seen. While I have uncorrectable vision, my peripheral sense is profound. I do not know if it due to conditioning, physiology or adaptation to poor central vision.

 

Most of that is periferal vision. Look at the screen in front of you. How much are you really looking at - concentrating on? Individual letters? Pixels? Words?

[pico]

Most of that is periferal vision. Look at the screen in front of you. How much are you really looking at - concentrating on? Individual letters? Pixels? Words?

 

That is an issue of viewing distance, Andy. When you look out the window or are outside, the perceived angle of view is wider than what I see on my laptop monitor. Regarding the level of attention to, for example, a laptop screen or book, most of us read in chunks of words. With training we can make the chunks larger and read faster.

[01af]

Some people are quite vocal about these two parameters being the only two determinants ...

They are badly wrong.

 

 

... meaning that e. g. two different 50 mm lenses at f/4 will always and necessarily have the same DOF.

This is wrong on (at least) two levels.

 

As a first-order approximation, depth-of-field basically is determined by the maximum accepted diameter of the circle-of-confusion, the focus distance, the focal length, and the aperture. These are the parameters that typically go into DOF formulas and calculators.

 

A fifth parameter would be the pupil magnification of the lens. However that is (1) negligible at medium and long focus distances and will have a significant effect on DOF only in macro photography, and (2) usually unknown to the photographer anyway.

 

Another parameter is the design of the focus mechanism of the lens (focusing by extension, focusing by extension with floating elements, internal focusing, front-part focusing , rear-part focusing, or a combination thereof) and how exactly it affects the actual focal length of the lens at focus distances shorter than infinity.

 

Yet another parameter—or actually a highly complex bunch of parameters—is the spherical correction of the lens and the residual spherical aberrations at the various focus distances. These are known to affect bokeh but they also affect depth-of-field, albeit in a subtle way (usually).

 

Bottom line—even when max. COC, focus distance, and aperture are equal, two different 50 mm lenses on the same camera can yield different depths-of-field. There are many reasons for this phenomenon, among them actual focal lengths (which usually is not exactly 50.0 mm for a 50 mm lens and may vary with focus distance), pupil magnifications, and residual spherical aberrations. Most likely this list of reasons is not yet complete.

[pico]

Olaf, would you agree that the additional parameters you mentioned (those other than FL, aperture, given subject distance and circle of confusion) produce virtually unnoticeable differences on depth of field in modern, nominally corrected lenses? IOW, disregarding lenses such as the Rodenstock Imagon and others that intentionally whack certain corrections.

[pgk]

As a first-order approximation, depth-of-field basically is determined by the maximum accepted diameter of the circle-of-confusion...

And this circle of confusion thingy is a concept isn't it? In the real world there can very rarely be subjects which give rise to this mathematical entity.....

[Christoph13]

This is wrong on (at least) two levels.

 

As a first-order approximation, depth-of-field basically is determined by the maximum accepted diameter of the circle-of-confusion, the focus distance, the focal length, and the aperture. These are the parameters that typically go into DOF formulas and calculators.

 

Yep, and these are covered in the conditions above.

 

A fifth parameter would be the pupil magnification of the lens. However that is (1) negligible at medium and long focus distances and will have a significant effect on DOF only in macro photography, and (2) usually unknown to the photographer anyway.

 

Unless I get this wrong, at any given subject distance two lenses might be affected but in the same way.

 

 

Another parameter is the design of the focus mechanism of the lens (focusing by extension, focusing by extension with floating elements, internal focusing, front-part focusing , rear-part focusing, or a combination thereof) and how exactly it affects the actual focal length of the lens at focus distances shorter than infinity.

 

Very true. I did a lot of macro work and one of my lenses (Canon EF 100 2.8 macro) reduced its focal length when focussing, at 1:1 its effective FL was closer to 67mm if I remember correctly. An Olympus lens did not have this effect or only much, much weaker. As a result, it needed a stronger correction at higher magnification ratios BTW.

With regard to the question above we are in ambiguous terrain though. One realization might be that FL (even for a prime) should not be regarded a constant but that, all other things being equal, two lenses having the same FL (which might not coincide with the FL imprinted on their barrels) at any given magnification ratio should still have approximately the same DOF.

The distinction quickly becomes academic though, good point.

 

 

Yet another parameter—or actually a highly complex bunch of parameters—is the spherical correction of the lens and the residual spherical aberrations at the various focus distances. These are known to affect bokeh but they also affect depth-of-field, albeit in a subtle way (usually).

 

Bottom line—even when max. COC, focus distance, and aperture are equal, two different 50 mm lenses on the same camera can yield different depths-of-field. There are many reasons for this phenomenon, among them actual focal lengths (which usually is not exactly 50.0 mm for a 50 mm lens and may vary with focus distance), pupil magnifications, and residual spherical aberrations. Most likely this list of reasons is not yet complete.

 

This is something that will be different for different optical designs. And if I may add this point myself, the formulas above treat DOF the same irrespective of where in the frame we are looking (measuring). With curvature of field and all the complexities of different focusing mechanisms this may or may not be a good approximation. And also something that may affect different optical designs in different ways.

 

Many thanks Olaf for your exhaustive reply, much appreciated

 

Christoph

[pico]

And this circle of confusion thingy is a concept isn't it? In the real world there can very rarely be subjects which give rise to this mathematical entity.....

 

Circle-of-confusion (CoC) is real to the extent that the rule is applicable, a valid metric for determining acceptable focus or depth-of-field (DOF). There are considerations such as viewing distance or degree of enlargement which determine what CoC one chooses. (One thing we tend to forget is that the CoC 'spot' degrades in contrast (CoCs overlap) if the lens is stopped down to the extent that diffraction effects occur, however again it depends upon degree of enlargement or viewing distance.)

 

Those who shoot large format rarely worry about diffraction because our smallest aperture (largest numerical aperture) is usually quite large in absolute diameter, but 35mm photographers might have greater concern, again depending as mentioned above. I print 35mm rather small - rarely larger than 5x7" and I can still see degradation due to diffraction, but often I don't care if DOF is the critical issue.

 

What is confusing is using the term 'point' instead of CoC because a point has no dimension whatsoever - it is a location, however we usually know what a person means.

[01af]

Olaf, would you agree that the additional parameters you mentioned (those other than FL, aperture, given subject distance and circle of confusion) produce virtually unnoticeable differences on depth of field in modern, nominally corrected lenses?

Mostly, yes. But in some cases, the differences can indeed become (small but) noticable upon close inspection. For example, the depths-of-field of the Summilux-M 50 mm 1:1.4 Asph and Summarit-M 50 mm 1:2.5 are clearly different ... of course, in real life this becomes mostly apparent not so much in the DOF itself but in the degree of blur outside the limits of DOF. For the same background blur, the Summilux needs to be stopped down by half or two-thirds of an f-stop more than the Summarit.

 

 

A fifth parameter would be the pupil magnification of the lens.

Unless I get this wrong, at any given subject distance two lenses might be affected but in the same way.

Usually, two different lenses with the same (nominal) focal length and the same image circle coverage will have the same or similar pupil magnifications. But it is not always and necessarily so. For example, a Tri-Elmar-M 16-18-21 mm Asph, at 21 mm, has a significantly larger pupil magnification than an old Super-Angulon 21 mm for Leica M. Still, at distances like 0.7 m and longer, this hardly matters.

 

 

... but that, all other things being equal, two lenses having the same FL (which might not coincide with the FL imprinted on their barrels) at any given magnification ratio should still have approximately the same DOF.

This is true when depth-of-field is small in relation to the width and height of the field-of-view—i. e. in close-up photography and telephoto photography.

[SJP]

Olaf, good to see such clear explanation(s) from someone that understands all the gory details.

 

In this context I still have some remaining things that I am not sure about:

- the DoF scales on lens barrels appear to be identical across the range, also from datasheets. I think they just use the calculated DoF (indeed how would you measure DoF - not very easy to give a precise criterion). Is that true?

- there are some that seem to have "lost the faith" in DoF as a concept, probably because it is a blurry subject. I tend to think that the concept as such is firmly established, useful, and reasonably exact for 'standard' lens use (0.7-infinite, normal F/values). Macro, micro & f/22 then we have a different ball game.

 

Another stance, even if you don't like DoF/hyperfocal distance, and what have you, still understanding the concept is very useful - especially if you use a rangefinder.

[FlashGordonPhotography]

Sensor size. Small sensor = more depth of field

 

Actually, all else being equal, a smaller sensor has less DOF, not more. It sounds counter intuitive but it's not. The increase in DOF seen when using a smaller sensor is created by other factors (moving further away or shooting a wider focal length) in order to get the same framing. At rhe same distance, focal length and aperture a smaller sensor has less DOF. Grab your DOF calculator of choice and try it.

 

 

As far as I can remember the things that "may" affect DOF are focal length, focal distance, aperture, sensor size, display size, display distance and lens design. The circle of confusion is based on sensor size, display size and display distance. I may be wrong but I beleive the current COC calculations are still based on a 7x5 print viewed at 12 inches. That's not entirely relevant in the age of the 27 inch cinema display.

 

 

Gordon

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[pgk]

understanding the concept is very useful - especially if you use a rangefinder.

Absolutely. My point above was that theory is all very well but translating it into actual images is a less exact science. FWIW I would say that the advent of high MPixel sensors has meant a far higher degree of criticality in examining imagery. I am conscious that, whilst aperture choice remains an important factor, I am far more careful now in terms of my selection of my precise point of focus - the are in the image which I deem it is most important to be absolutely 'sharp'. This choice is based on both theory and experience, and is especially important if I anticipate that the image is to be used at large size.

 

I tend to find myself looking at the whole question of depth of field in another way these days (I think jaapv said something along the following lines in another thread); that there is a plane of 'sharpness' surrounded by progressively less sharpness. Its the point at which the less sharp area becomes 'unsharp' that we are concerned about and trying to define this by a simple mathematical formula is fraught with difficulties. I would also add into the equation that the subject matter too has an influence (low contrast edges will appear to 'soften' before hard contrast ones do), so DoF scales can only be considered, IMHO, as reasonable 'guidance' rather than anything else.

 

We seem to be tending to forget that photography is fundamentally visual and we generally interpret images with some degree of subjectivity (of course there are scientific requirements but even these still require visual interpretation).