The Falacy of FF Cameras having Flatter DOF

[Donald]

Perspective is something I find I start losing when I spend too much time on photo forums.

I haven't read this thread, because just from the title you can tell what some of the discussion might be like, which is why I love this post.

[ajohnw]

I haven't read this thread, because just from the title you can tell what some of the discussion might be like, which is why I love this post.

If you did read it all and follow some links you might find that change a few things and it is a fallacy. Not sure where flatter came from though. Focal plains are hopefully reasonably flat and blur is just another flat image that doesn't coincide with the sensor.

What I've found interesting is that all of the shots I've seen posted and the links to compare the effect make the crop camera shots look better 'cause the full frame has too much blur which is distracting.

John
-

[Dave Humphries]

George, John,

When Neil Van Niekerk says "With a zoom, the perspective does not change - you are merely enlarging the image" he is absolutely correct and the statement is 100% true - if you think otherwise, please read on to understand properly.

Philip's example shots show this and he has described you how you can replicate this test for yourselves.

Yes, all three shots do show the same perspective!
Look at the sizes of all the foreground elements (trees, bins, et al) they do not change.

I will define perspective here as; the apparent difference in the sizes of (say) two objects in relation to each other when one is closer to the viewer than the other.
e.g. a foreground tree and a building somewhat behind it.

If you do not move and zoom only (or crop in PP, or magnify when viewing after the fact) the result is the same; the image gets bigger, but 'how far up' the top of the tree reaches on the building remains unchanged - ergo; perspective is unchanged.

Philip's diagram of what happens if you do move shooting position is also useful, although another couple of lines on it would have helped - e.g. extend the 'view line' from top edge of tree to show how far up the building it hits.
These new lines would clearly show a change of perspective between the two shooting positions; the person standing nearer the tree (and building) cannot see as much of the building because the view is blocked by the tree, which appears taller to him/her.

Hope that solves the issue, Dave

PS
Unfortunately John, the article "How to Achieve Blurred Backgrounds in Portraits" by Darlene Hildebrandt you linked to back in post #24 is badly written (although some parts are valid); she has muddled her facts and is helping to perpetuate the confusions for you and many others; since 2012 sadly.
Basically, the crucial thing she overlooks is the effect of changing camera to subject distance!
I may (another day) trawl her comments to see if anyone else has raised this and if not, let her know.
UPDATE: Nah, I just read some comments and her responses - and she can't see the problem with what she's written

[george013]

George, John,

When Neil Van Niekerk says "With a zoom, the perspective does not change – you are merely enlarging the image" he is absolutely correct and the statement is 100% true - if you think otherwise, please read on to understand properly.

Philip's example shots show this and he has described you how you can replicate this test for yourselves.

Yes, all three shots do show the same perspective!
Look at the sizes of all the foreground elements (trees, bins, et al) they do not change.

I will define perspective here as; the apparent difference in the sizes of (say) two objects in relation to each other when one is closer to the viewer than the other.
e.g. a foreground tree and a building somewhat behind it.

If you do not move and zoom only (or crop in PP, or magnify when viewing after the fact) the result is the same; the image gets bigger, but 'how far up' the top of the tree reaches on the building remains unchanged - ergo; perspective is unchanged.

Philip's diagram of what happens if you do move shooting position is also useful, although another couple of lines on it would have helped - e.g. extend the 'view line' from top edge of tree to show how far up the building it hits.
These new lines would clearly show a change of perspective between the two shooting positions; the person standing nearer the tree (and building) cannot see as much of the building because the view is blocked by the tree, which appears taller to him/her.

Hope that solves the issue, Dave

PS
Unfortunately John, the article "How to Achieve Blurred Backgrounds in Portraits" by Darlene Hildebrandt you linked to back in post #24 is badly written (although some parts are valid); she has muddled her facts and is helping to perpetuate the confusions for you and many others; since 2012 sadly.
Basically, the crucial thing she overlooks is the effect of changing camera to subject distance!
I may (another day) trawl her comments to see if anyone else has raised this and if not, let her know.
UPDATE: Nah, I just read some comments and her responses - and she can't see the problem with what she's written

Reading and thinking it all over it's the same as point of view. Without any relation to something as a third dimension.

George

[ajohnw]

I don't really have any problems at all with dof or blur Dave. I stick to my definition of perspective because that is what I was taught and it can be a useful factor to remember. I do think that there really is some confusion related to blur and crop, As I have shown with the links.

The link you mention I posted does show the basics of blur - the factors that can control it. It's not just the format of the camera. There are other factors. I posted it as a hint. Not taken so I posted links to the blur calculator with certain settings. I do have some doubts about the validity of the calculator but it does show trends.

One of the really silly aspects of this topic is that smaller formats can not show the same degree of blur as full format. It is a fallacy providing certain factors are changed when the smaller format is used. There are other variable not related directly to the crop factor - lens focal length, the aperture of the lens that is actually used and the distance from which the shot is taken. In these cases the framing can be exactly the same as was used on a FF camera. Ok there will be minor differences but in real terms few would spot the differences.

I was sucked in by crop factors though. I bought the 45mm when really I should have bought the 75mm. Maybe if there was a little more clarity about on this subject I might have bought the 75mm. Probably not really as I wanted the 45mm to obtain an eye perspective which it will more or less give. I don't think there is any such thing as an exact one. One pro at least seems to prefer the 60mm lens for full figure plus scenery work. Perhaps it fits in with his studio sets and the 75mm doesn't.

John
-

[george013]

I haven't read this thread, because just from the title you can tell what some of the discussion might be like, which is why I love this post.

I know what you mean. But did you know that perspective and point of view are the same? I mentioned it in post 26 and nobody confirmed that they are the same. And still not.

George

[mikesan]

Seems to me that this thread has forked in two different directions. It began with a question of the effect of sensor size on DOF, and then diverged to a discussion of perspective. Two really different subjects, and it's a pity that they have been conflated in this one thread.

As to perspective; I am a firm believer in the dictum that perspective is determined solely by the distance between the photographer (or the camera or lens nodal point or sensor; take your pick) and the subject being photographed. In particular it is independent of lens focal length. I have remarked, in another thread, that this dictum has often been the subject of controvery. I attribute this to the fact that those who dissent have a different concept of perspective (or how it is represented in a two dimentional image) and generally they fail to present a convincing counterargument.

George writes: You don't define perspective. You subscribe a feeling for what is perspective for you.

It's hard to argue with that.

Phillip's post (#35) provides (for me) a definitive answer and if George believes that the two images do not have an identical perspective, I should very much like to know how he reaches this conclusion.


Let me now turn to the subject of sensor size and DOF. I spent my entire career in science (bio-medical) which, I hasten to add, has not provided me with much indepth knowledge of optics. However I have learned that the most difficult task, when attempting to learn by experimentation, is the construction of a proper control to compare the outcomes. We often use the phrase "all other things being equal". Therein lies the problem; unless one is very meticulous, they are often not equal; the devil is in the details.

Bill (post #2) describes a simple example:

What I have stated often is that if we make two images using two different format cameras: then for the same FRAMING of any Subject and for any given same APERTURE USED, the image made with a 135 Format Camera (aka "full frame") camera will have a smaller Depth of Field than that which is made with an APS-C Format Camera.

I am not really sure of what Bill believes because, later in the same post, he writes:

a full frame camera has a shallower DoF than an APS-C Camera" . . . which I believe is clearly wrong and that point is good to clear up.

To follow Bill's scenario, SAME FRAMING needs to be exactly defined. I will assume this means a given object (in the plane of focus) will occupy the same area of the frame (otherwise stated as image magnification). It does not imply that the frames will encompass an equal angle of view. He fails to stipulate the focal lengths of the lenses used on each camera. I will also assume that one adjusts the acceptable CoC proportionate to the sensor size.
I will ignore any effect of difference in the pixel density or pitch between sensors, simply because I am not well informed on this subject.

Thinking through this excersize it is apparent that, in order to achieve the same FRAMING, one has two choices:

(a) Use the same focal length lens on both cameras; this ensures the framing must be identical. The only difference in the final images is loss of some portion of the scene due to cropping by the APS-C sensor, but all parts that are captured must have the same DOF as that of the FF camera.

(b) The most common solution is to equip the APS-C camera with a lens having a shorter focal length than that used on the FF camera in order to achieve the same field of view on both. When this is done one finds that to achieve the same framing (as defined above) one must move closer (shorten the object distance) with the APS-C camera. We must now consider two factors, the object distance and the lens focal length, both of which may affect DOF.

Decreasing the object distance (as we all know from experience) will narrow the DOF. The required decrease in object distance is directly proportional to the focal length of the lenses used and the decrease in DOF is very closely approximated by the same fraction.

We know that altering the focal length may have an effect on DOF. We often hear that a wide angle lens will provide a greater DOF than a tele lens. Beware! this is not always true. In this case we are concerned with only a single application: we wish to keep the framing (image magnification) constant. Under this condition the focal length will have a negligible or insignificant effect on DOF when the object distance is smaller than the hyperfocal distance. With longer object distances the shorter focal length lens will provide more DOF . Try it on your choice of DOF calculator. To complete the picture (for the truly pedantic) a shorter focal length lens will have more rear DOF, while the longer lens will have more front DOF.

So the bottom line is (according to my analysis; and I welcome any comments or corrections) under the conditions of Bill's example (as ammended by my assumptions); under most conditions the FF camera will deliver a greater (wider if you prefer) DOF than that of the APS-C camera.

For those interested in reading one more website on this subject, I can highly recommend the following. Most interesting is his discussion of the distinction between DOF and background blur.

http://toothwalker.org/optics/dof.html

[NorthernFocus]

Sorry I haven't gotten back to this thread, folks. I haven't had a chance to give it proper attention but have followed along and it's been interesting. Forgive me if the following is poorly articulated but it's late.

Most of the commentary thus far has IMO validated my purpose for the OP. Those of us who are.... technically oriented tend to get ourselves wrapped around the axle with formulae etc. And on the topic of DOF, my argument is that, while useful, the accepted methods don't really reflect current technology and practice and therefore can be very misleading to the average user.

When I posed the question regarding if cropping an image affects DOF...

Sure it does. DOF is defined for a final printed photo (image size vs viewing distance)...

Precisely. The classic definition which isn't highly meaningful for the average user nowadays who 1) likely crops their images substantially and 2) likely doesn't print them but rather reduces them for digital display.

... and it is difficult to assess by looking at 100% magnification on the screen.

Yet that is exactly what we do when critically analyzing focus when DOF plays an important roll.

...It does not matter if you used a crop sensor with an FX lens, cropped the image in PP or printed a 2 meter poster and cut out a small part of it with a pair of scissors - cropping always leads to blur becoming more visible.

If I understand this argument you're simply saying that making more detail/pixels visible increases DOF. Which sounds suspiciously like a matter of resolution that has nothing to do with optics. After all pixels in the original capture don't change when the image is cropped. Viewed at pixel level, objects x feet in front of and y feet behind the focal plane are blurred(aka lose contrast along edges). That blurring is due to optical performance and influenced by sensor resolution. The size of the capture medium is not relevant.

Regarding the classic definition of DOF involving an 8x10 print, etc, when comparing different size sensors the definition blows up from the start. The comparison is supposed to be with a subject framed identically (aka with the same FOV) on the two capture media with "all things being equal". Which is an impossibility. In order to produce the same FOV with two different sensor sizes either a)range or b)focal length have to be different. Or c) the image from the larger sensor has to be cropped. And in the case of c, and all other things truly being equal, the images will be the same. Other than any differences in resolution of the final images being compared.

Which brings me back to my OP slightly re-stated. In practical application, abandoning the "classic" method of analysis, it is a fallacy, aka misleading, to generically state that a FF sensor is capable of flatter DOF. Conditionally, it can be an accurate statement. But most of us tend to omit the conditions that should accompany the comment.

FYI, what started me down this mental rabbit hole is the fact that I've been doing a lot of shooting lately with the D810 which can shoot in 1.5x and 1.2x crop modes. On any given "walkabout" photo outing it is likely that I'll end up shooting FF, 1.2x, and 1.5x modes. It sets one to wonder whether any given mode is (creatively)advantageous for a given subject...

[Stagecoach]

FYI, what started me down this mental rabbit hole is the fact that I've been doing a lot of shooting lately with the D810 which can shoot in 1.5x and 1.2x crop modes. On any given "walkabout" photo outing it is likely that I'll end up shooting FF, 1.2x, and 1.5x modes.

It sets one to wonder whether any given mode is (creatively)advantageous for a given subject...

I can only think that the advantage (creatively) of using the 1.2/1.5x modes are any advantages you get by having to physically move farther from the subject to achieve the same framing you would have got with FF.

[george013]

Sorry I haven't gotten back to this thread, folks. I haven't had a chance to give it proper attention but have followed along and it's been interesting. Forgive me if the following is poorly articulated but it's late.

Most of the commentary thus far has IMO validated my purpose for the OP. Those of us who are.... technically oriented tend to get ourselves wrapped around the axle with formulae etc. And on the topic of DOF, my argument is that, while useful, the accepted methods don't really reflect current technology and practice and therefore can be very misleading to the average user.

When I posed the question regarding if cropping an image affects DOF...


Precisely. The classic definition which isn't highly meaningful for the average user nowadays who 1) likely crops their images substantially and 2) likely doesn't print them but rather reduces them for digital display.


Yet that is exactly what we do when critically analyzing focus when DOF plays an important roll.


If I understand this argument you're simply saying that making more detail/pixels visible increases DOF. Which sounds suspiciously like a matter of resolution that has nothing to do with optics. After all pixels in the original capture don't change when the image is cropped. Viewed at pixel level, objects x feet in front of and y feet behind the focal plane are blurred(aka lose contrast along edges). That blurring is due to optical performance and influenced by sensor resolution. The size of the capture medium is not relevant.

Regarding the classic definition of DOF involving an 8x10 print, etc, when comparing different size sensors the definition blows up from the start. The comparison is supposed to be with a subject framed identically (aka with the same FOV) on the two capture media with "all things being equal". Which is an impossibility. In order to produce the same FOV with two different sensor sizes either a)range or b)focal length have to be different. Or c) the image from the larger sensor has to be cropped. And in the case of c, and all other things truly being equal, the images will be the same. Other than any differences in resolution of the final images being compared.

Which brings me back to my OP slightly re-stated. In practical application, abandoning the "classic" method of analysis, it is a fallacy, aka misleading, to generically state that a FF sensor is capable of flatter DOF. Conditionally, it can be an accurate statement. But most of us tend to omit the conditions that should accompany the comment.

FYI, what started me down this mental rabbit hole is the fact that I've been doing a lot of shooting lately with the D810 which can shoot in 1.5x and 1.2x crop modes. On any given "walkabout" photo outing it is likely that I'll end up shooting FF, 1.2x, and 1.5x modes. It sets one to wonder whether any given mode is (creatively)advantageous for a given subject...

You missed the basics of the sharpness/dof theory.
That's that the human eye experiences a circle of 0.25mm as a point. All the other things are calculations based on that aspect. Basic it should work with monitors too. But pay attention to the pixelsize. It equals the 0.25mm. In my office the screen is 19", res. 1280x1024. Pixelsize 0.29.
And when working for the net, the only monitor you've under control is yours.
If you think you should use another coc, you can fill that in in the dofcalculator.

Your last alinea, I have a zoom lens for that. Why playing with the crop factor in your camera if you can do that in pp?

George

[dem]

And on the topic of DOF, my argument is that, while useful, the accepted methods don't really reflect current technology and practice and therefore can be very misleading to the average user.

I agree that trying to combine the classic definition of DOF with pixel peeping and pixel count is less than straightforward and probably unnecessary.

If I understand this argument you're simply saying that making more detail/pixels visible increases DOF.

No, when you magnify the image, you magnify the blur. That means that the areas of sharp focus shrink: DOF decreases.

That blurring is due to optical performance and influenced by sensor resolution.

The blurring of out of focus areas has very little to do with sensor resolution, lens quality or diffraction softening. It comes from simple geometrical optics on which all DOF calculators are based.

Regarding the classic definition of DOF involving an 8x10 print, etc, when comparing different size sensors the definition blows up from the start. The comparison is supposed to be with a subject framed identically (aka with the same FOV) on the two capture media with "all things being equal". Which is an impossibility.

It is not impossible. In fact that is exactly what "equivalence principle" is. This is probably the most comprehensive article on the topic I have seen:
http://www.josephjamesphotography.com/equivalence/

Which brings me back to my OP slightly re-stated. In practical application, abandoning the "classic" method of analysis, it is a fallacy, aka misleading, to generically state that a FF sensor is capable of flatter DOF.

Quite the opposite. When people switch from small sensor cameras to DSLRs, the common complaint is often "Why my photos are so blurry?". With typical focal lengths and apertures available you are more likely to accidentally blur background on a DSLR, either crop or full frame, than a typical P&S or a smartphone. So size matters. It probably does not matter that much when you go from a full frame to say 1.5 crop sensor.

[ajohnw]

I like that post Dan but I think one DOF aspect needs expanding on.

One of the problems with this area is comparing like with like. My impression is that it gets bogged down with what's on the sensor. The only meaningful like with like is the final result when it's viewed. In that case in order to produce a more or less identical image as a FF camera the crop camera must have a greater dof because it will need more enlargement to reach what ever size the final image is.

So in my case with M 4/3 I need blur levels that are approximately 1/2 what a FF camera can produce for the same framing. I played with the settings of the calculator again because it can show that I can't match a FF camera precisely. This time the blur for the same size final image gets greater than FF at some particular distance. Prior to that it's very similar. The previous one was a touch less and and then very similar. Difference under 1/2 stop. In this case the focal lengths are very similar but as Grahame mentioned for the same framing the camera would have to be further away - so what. It might or might not be a problem.

http://howmuchblur.com/#compare-1x-8...m-wide-subject

Here is what happens if the M 4/3 lens is changed to 45mm so similar to the 85mm FF lens. Bearing in mind that for the same final image size the blur does need to be less and is roughly 1/2 that of the FF.

http://howmuchblur.com/#compare-1x-8...m-wide-subject

Actually if I could put 42.5mm in the calculator the blur would be 1/2. Not surprising really as that's how the sums work.

On perspective I had a number of lectures from Royal Society people years ago on the nuts and bolts of cameras and the effects lens focal lengths and and apertures etc. I can't really be expected to accept any other definition other than the one I gave but if others want to call that something else or see perspective as something different that's fine by me. I don't see that the term has any bearing on the other factor which is framing - it just has an effect on the apparent depth that will be seen in the final image. Distortions relating to an eye view too in some cases.

John
-

[pnodrog]

A word of caution for some of the above arguments - DOF and resolving power/resolution are not the same thing.....

[ajohnw]

I don't think I have mentioned resolving power Paul. Maybe some one else did.

It's another can of worms based on popular views. Reality is a touch different.

John
-

[William W]

. . . Those of us who are.... technically oriented tend to get ourselves wrapped around the axle with formulae etc. And on the topic of DOF, my argument is that, while useful, the accepted methods don't really reflect current technology and practice and therefore can be very misleading to the average user.

I think that there is a lot of misinformation on the www and part of the reason for that misinformation is poor quality inaccurate regurgitation.

. . . The comparison is supposed to be with a subject framed identically (aka with the same FOV) on the two capture media with "all things being equal". Which is an impossibility.

I think that you are missing the point of the rationale of why DoF Calculations were ‘invented’. I think that you are applying and or thinking about DoF calculators and Charts the wrong way about. The words in bold are very important:

The purpose of using a 10x8 print as a ‘standard’ is twofold:
1. to allow viable comparisons of DoF within one camera format, for example using different lenses and different apertures and to provide a rough guide as to what will be in “acceptable focus" for a particular shot.

2. to allow viable comparisons of DoF between camera formats using a suitable ‘appropriate’ different CoC to be applied to the different formats and to provide a rough guide as to what will be in “acceptable focus" for a particular shot.

The point is – without a “ ‘standard’ (non cropped) 10 x 8 print held at Arm’s distance” we would not have a starting point to backwards reference from what the eye can resolve and hence make COMPARISONS of what DoF will be in a shot and have a ROUGH GUIDE of what DoF we will get in a shot.

DoF Charts are NOT a definitive reckoning: they are for comparative and to be used as 'a guide' purposes.

BUT DoF Charts must have a “standard” as their base so that the comparisons and the guide have meaning.

This is an important point to grasp: DoF Charts are used at the front end of Photography to get a feel for the DoF that will be in the shot . . . this is often very difficult to explain in the classroom face to face so please re read this and I hope I have explained it well enough.

*

Which brings me back to my OP slightly re-stated. In practical application, abandoning the "classic" method of analysis, it is a fallacy, aka misleading, to generically state that a FF sensor is capable of flatter DOF. Conditionally, it can be an accurate statement. But most of us tend to omit the conditions that should accompany the comment.

Well . . . in my view DoF charts and calculators are NOT really for ‘analysis’ - that’s the whole point.

These tables are to be used as “guide” to use before making the shot.

So I expect that we agree that if you are indeed using DoF Charts for ‘analysis’ then that is probably misleading, but I would just say that is using DoF Charts inappropriately or not as they designed purpose.

For example - if you are out in the field and are considering a shot and you consult the DoF chart to provide a guide as to how much DoF you will have using F/4 and F/8, then firstly, if you do this often enough, you know how the DoF distance relates to your “Screen View” – or you could have (like Ted) made up a DoF calculator specifically predicated on a “Screen View” and not the ’standard’ 10x8 print. Secondly, Irrespective of the figures used to indicate the DoF you will have a COMPARISON of how the difference of the DoF at F/4 and F/8.

Furthermore – if you as a general rule crop away most of your image real-estate, then you know how that affects the resultant DoF. . . you don’t need a DoF chart or calculator to tell you that.

***

This brings me back to my initial response:

“I have stated often is that if we make two images using two different format cameras: then for the same FRAMING of any Subject and for any given same APERTURE USED, the image made with a 135 Format Camera (aka "full frame") camera will have a smaller Depth of Field than that which is made with an APS-C Format Camera.”

Here I will expanded upon why this is important fact to some Photographers:

. . . Let me now turn to the subject of sensor size and DOF. . . unless one is very meticulous, they are often not equal; the devil is in the details.
Bill (post #2) describes a simple example:

What I have stated often is that if we make two images using two different format cameras: then for the same FRAMING of any Subject and for any given same APERTURE USED, the image made with a 135 Format Camera (aka "full frame") camera will have a smaller Depth of Field than that which is made with an APS-C Format Camera.

I am not really sure of what Bill believes because, later in the same post, he writes:

a full frame camera has a shallower DoF than an APS-C Camera" . . . which I believe is clearly wrong and that point is good to clear up.

To follow Bill's scenario, SAME FRAMING needs to be exactly defined. I will assume this means a given object (in the plane of [sharp] focus) will occupy the same area of the frame.

Yes. That is one way of describing what is meant by the word “framing” (In caps to signify a technical term). However, no disrespect meant, I think that “same framing” is self descriptive. Firstly, for clarity FRAMING is what we do in the camera; CROPPING is what we do in the darkroom or with the Post Production Computer. So then for example if I take a two photos of a person’s head and in each the chin touches the bottom of the frame (in the viewfinder) and the top of the head touches to top of the frame (in the viewfinder) of both an APS-C and 135 Format camera and both viewfinders are 100%, then that is “SAME FRAMING”

*

It does not imply that the frames will encompass an equal angle of view.

Correct.

*

He fails to stipulate the focal lengths of the lenses used on each camera.

Correct.

*

I will also assume that one adjusts the acceptable CoC proportionate to the sensor size.

Sort of irrelevant, as there’s no need to ‘adjust’ CoC, as the statement was to precede a practical test: but if you want to pop the figures into a DoF calculator to get a theoretical answer, then yes, adjust the CoC to suit the Sensor size, but be careful to use the ‘acceptable’ CoC for each sensor based upon the same standard – for example “10 x 8 full crop print at arms distance”

*

I will ignore any effect of difference in the pixel density or pitch between sensors, simply because I am not well informed on this subject.

Good idea anyway.

*

Thinking through this exercise it is apparent that, in order to achieve the same FRAMING, one has two choices:
(a) Use the same focal length lens on both cameras; this ensures the framing must be identical. The only difference in the final images is loss of some portion of the scene due to cropping by the APS-C sensor, but all parts that are captured must have the same DOF as that of the FF camera.

I am not sure what you mean, but . . . If we use the SAME Focal Length Lens on each camera then to achieve the SAME FRAMING of any Subject the APS-C Camera must be situated farther from the Subject than where the 135 Format Camera is situated. We do not want to CROP, but rather we want the same FRAMING. (see above for the difference between “Cropping” and “Framing” as Technical Words in Photography).

*

(b) The most common solution is to equip the APS-C camera with a lens having a shorter focal length than that used on the FF camera in order to achieve the same field of view on both. When this is done one finds that to achieve the same framing (as defined above) one must move closer (shorten the object distance) with the APS-C camera. We must now consider two factors, the object distance and the lens focal length, both of which may affect DOF.

Yes. Different lenses were used in the sample image that I posted.

But No. We do not necessarily need to change the Subject Distance.

Choosing, for example, an 85mm lens on a 5D and a 50mm lens on a 50D, then means that the SUBJECT DISTANCE can be almost exactly the same to produce the SAME FRAMING – hence that’s why the example photo of two people taking a photo of each other .

*

Decreasing the object distance (as we all know from experience) will narrow the DOF. The required decrease in object distance is directly proportional to the focal length of the lenses used and the decrease in DOF is very closely approximated by the same fraction.
We know that altering the focal length may have an effect on DOF. We often hear that a wide angle lens will provide a greater DOF than a tele lens. Beware! this is not always true. In this case we are concerned with only a single application: we wish to keep the framing (image magnification) constant. Under this condition the focal length will have a negligible or insignificant effect on DOF when the object distance is smaller than the hyperfocal distance. With longer object distances the shorter focal length lens will provide more DOF . Try it on your choice of DOF calculator. To complete the picture (for the truly pedantic) a shorter focal length lens will have more rear DOF, while the longer lens will have more front DOF.
So the bottom line is (according to my analysis; and I welcome any comments or corrections) under the conditions of Bill's example (as ammended by my assumptions); under most conditions the FF camera will deliver a greater (wider if you prefer) DOF than that of the APS-C camera.

This is where I think that you go off onto a tangent and begin introduce new (and unnecessary) things to the statement that I originally made.

***

In point form, I’ll shall now re-write what I mean and this is the exact exercise that I give Students to do:

We will make two pictures of the same Subject using two different format Cameras: one APS-C Format and the other 135 Format

1. keep the SAME FRAMING
2. keep the SAME APERTURE
3. make a picture with and APS-C Camera
4. make a picture with a 135 format Camera (aka ‘full frame’)
5. don’t worry about and for all practical purposes it does NOT matter about the Focal Length of the Lens that you use

Then note: the picture that you made with a 135 Format camera will have a SHALLOWER DEPTH of FIELD (i.e. LESS DoF) than the picture made with the APS-Camera.

*

If you choose to not do the practical exercise but rather a theoretical answer, then you can pump the figures into a DoF calculator – remembering to use ‘appropriately acceptable’ figures for the CoC of the different Camera Formats. But I encourage you to do the PRACTICAL experiment as outlined above.

A mathematical proof can be calculated from first principles. I choose not to write that here.

***

I think the factual results of that experiment (or if you want, the mathematical facts if you want to go through the theory), has been misconstrued on the www to something like “One of these conventional wisdoms that I was recently pondering is the "fact" often stated that full frame cameras have flatter DOF(or APS-C sensors have wider DOF if you please).”

and that in the main was the reason why I responded to the Opening Post.

***

USEFULNESS and APPLICATIONS:

One useful application of this information is for those of us who want to make photos with very shallow DoF (for example some genres of PORTRAITURE). Using a 135 Format Camera will in general terms make that task much easier.

Note also that this is NOT a modern phenomenon which is restricted to contrasting the applications of APS-C and “full frame” Digital Cameras.

A similar situation exists when we compare Medium Format with 135 Format: referring again to “Equivalence in Photography” note that the ‘difference’ between the Aperture which is required to attain the same DoF when moving from 135 Format to 645 Format is about 1⅓Stops and to a 6x6 camera is about 1⅔Stops and a 6x7 camera is about 2 Stops. In Film Days, if we wanted to have a very shallow DoF for a shot, we would usually opt to use a 135 Format Camera and NOT Medium Format Cameras.

***

(This is the crux of all of this – if you only read one bit please read this bit)

The sole reason for 135 Format being ‘special’ in regard to that Format being able to easily produce IMAGES with very shallow DEPTH of FIELD is because of the very fast LENSES that are available for 135 Format cameras.

As examples:
– an F/0.95 was available for Rangefinder 135 Format Cameras many years ago
– more recently an F/1.0 lens is attainable for use with Canon DSLRS and F/1.4 and F/1.8 lenses are easily and relatively cheaply available.

On the other hand:
- (as far as I know) there is only one F/1.8 Lens for APS-C, the next fastest is F/2.
- 645/6x6 Formats Cameras, (I think) F/2.8 is still the fastest and most of those this lenses are expensive
- 6x7 I am not sure, but the fastest 6x7 lens that I have is F/4.

So – with that information, it is easy to understand that if you want really shallow DoF then (for example) buying a used EOS 5D and an EF 50mm F/1.8 MkII will provide that result for you at a really cheap cost. On the other hand to get anywhere near the same shallow DoF using an APS-C Camera, you will need a Lens Speed of about F/1.1.

And on the third hand to get anywhere near the same shallow DoF using an 645 Camera, you will need a Lens Speed of F/2.8, which does exist, but is usually an expensive Lens and if you have a 645 Kit, then you probably have an F/1.4 or F/1.2 or F/1.0 Lens for your 135 Kit and that will cream the F/2.8 Lens used on your 645 Camera, anyway.

Moving up the camera Formats - through 6x6; 6x7; 6x9 and onto 5x4, 10 x8 ... All those Camera Formats just cannot compete with 135 Format’s ability to attain very shallow DoF . . . but not because of the format size per se, but because of the Lenses which are available to use simply are not fast enough.

Similarly, as we travel down through the smaller Camera Formats beyond APS-C through P&S and Bridge cameras and Phone cameras ... it is simply impossible to get lenses fast enough to create the really shallow DoF we can get using a fast lens on a 135 Format camera.

***

For clarity what I am NOT stating is: “it is impossible to get shallow DoF with APS-C Format." Of cause it is possible to do that.

A substantial part of the opening post was concerned with misinformation about 135 Format (aka “Full Frame”) and how these cameras have a shallower DoF – what I am stating is, that using a 135 Format camera it is generally both easier and cheaper to make very shallow DoF images and also attain a greater range of shallow DoF images than using any other Camera Format.

WW

Footnotes and ‘by the way’ information:

1. The Technical Class of Camera for 135 Format is “Miniature Format:” and the Technical Class of camera for APS-C Format is “Sub-Miniature Format” – the term “Full Frame” is a modern iteration and as far as I have researched is attributable to a leading Digital Camera Manufacturer and was first used in advertising material as hype to differentiate and market the release of ... “the first ‘full frame’ DSLR” . . . which begs the question is a ‘blad H6D-50C “a Fuller Frame” and is a ‘blad H6D-100C a “more Fuller Frame” . . . etc ?)

*

2. For most practical everyday Photography purposes and certainly for the purposes of this conversation it doesn’t matter about the Focal length of the Lens which is used. This is because of the Axiom of DoF which, in general/simple terms, states that the DoF remains the same if the APERTURE and the CAMERA FORMAT remain the same. (This Axiom has been explained in other commentaries here at CiC in much detail). Therefore when comparing (for example) a 17 to 55 F/2.8 lens used on APS-C to a 24 to 70 F/2.8 Lens used on 135 Format – it doesn’t matter what FL that is used it is always impossible to make a shallower DoF for any given framing using the APS-C camera.

*

EXAMPLES:



Made with a 5D Series Camera and an 85mm used at Aperture F/1.8.

It is possible to make a facsimile with an APS-C Format Camera but the Aperture required would need to be about F/1.1 and whilst Canon did an EF 50F/1.0L (and I have used that lens) ... I really do get more bang for my buck using the EF 85/1.8 on a 5D Series than ‘investing’ a truck load of $ in Canon's 50 F/1.0L lens to use on an APS-C Camera.

*



Made with a 5DSeries Camera and a 135mm used at Aperture F/2.

It is possible to make a facsimile with an APS-C Format Camera, but the Aperture required would need to be about F/1.2 and whilst Canon make an EF 85F/1.2LMkII ... I get more bang for my buck using the EF 135/2 on a 5D Series than ‘investing’ a truck load of $ in an F/1.2 lens to use on an APS-C Camera.

*



Made with a 5DSeries Camera and a 35mm used at about Aperture F/2

It is becoming impossible to make a facsimile with an APS-C Format Camera, the Aperture required would need to be about F/1.2 and I don’t think that there is there is a 20mm/1.2 lens about. The closest lens probably is an EF 24/1.4L MkII – that’s a nice lens, but, if we are looking for value for money, I’d get more bang for my buck using the EF 35/2 on a 5D Series than ‘investing’ a truck load of $ in an F/1.4 lens to use on APS-C .

*

Examples can go on and on . . . for example when we use an EF35/1.4 at F/1.4 on a 135 Format camera . . . it is impossible to make a facsimile shot using an APS-C camera. etc.

Agreed these facts are not important to many photographers - but they are important to some Photographers and probably explain some of the reasons why some Photographers opt to buy a used 5D rather than a new 7D.

All Images © AJ Group Pty Ltd Aust 1996~2016 WMW 1965~1996

[William W]

For what it is worth I think this video may be of some interest as it actually demonstrates some of the issues under discussion.
https://www.youtube.com/watch?v=f5zN6NVx-hY
Be patient, it gets to the most interesting stuff if you wait.

Yes. Definitely worth a look. I know that video. I suggest that one interprets it as a lay overview and not a textbook guideline.

Around 6’00” he gets to the nitty-gritty of “comparisons”.

Some warnings - personally, I am not keen on some of the words that he uses and some of his phraseology – for example he seems to interchange “image quality” to mean different things - and that's only one example.

And I am really not keen on his asking for Camera Manufacturers to “apply the crop factor to both the Focal Length and the Aperture because that is more accurate and more truthful”

My point of view is, if the term “crop factor” were never ever unleased, then there would be seriously MUCH less confusion.

For example, there was never much confusion or lack of truth when using an 80/2.8 on a 645 Format and then that compared and contrasted to using a 50/1.4 on a 135 Format Camera. But that horse has bolted and there is no way it will ever get put back in the barn.

WW

Hi Trev - trust you are good and smooth rowing.

[William W]

I concur in every aspect with Dem's post #51.

I underscore to his first response in that post that it is my view that DoF Calculators are probably now being used in a manner not for which they were designed .

In that Post #51, there is also the link to James Joseph's paper - I referred to that paper in Post #17.

I also concur with Dem: Joseph's paper is a most comprehensive work on this topic

WW

[William W]

. . . In photography, perspective depends on the position of the photographer relative to the different parts and objects in the scene. It is independent of focal length. The latter changes the angle of view and, therefore, the field of view recorded on the rectangle of the sensor, but the perspective remains the same when the shot is taken from the same position using different focal lengths. . . etc.

AND

George, John, When Neil Van Niekerk says "With a zoom, the perspective does not change - you are merely enlarging the image" he is absolutely correct and the statement is 100% true - if you think otherwise, please read on to understand properly. . . etc

I concur with both responses.

I also note that there have been several previous threads at CiC where there has been a request for a definition of the term “Perspective”.

Sometimes, subsequent to the correct technical definition having being given (as per in this thread by Philip and Dave) there is disagreement or argument about that correct technical definition.

I don’t see a lot of point in continually debating a definition which is established in respected Texts and Curricula.

WW

[NorthernFocus]

Bill, first let me apologize for initiating a thread that ultimately resulted in your feeling compelled to spend so much time to generate a thorough response. But... on the other hand, now I don't feel so bad about generating so much confused discussion on what we all intend to be a learning forum. Because ultimately it has ended well because your post is an excellent tutorial on the topic. You included the conditions/qualifications that are typically left out of the broad statements made on the topic, provided well articulated explanations and visual examples.

I think that there is a lot of misinformation on the www and part of the reason for that misinformation is poor quality inaccurate regurgitation.

Completely agree.

...I think that you are missing the point of the rationale of why DoF Calculations were ‘invented’...

Not at all. Fully aware of the value/use of standards and standard practices. In my professional life I served on engineering standards committees once or twice.

... I think that you are applying and or thinking about DoF calculators and Charts the wrong way about...

Not at all for my own use. But nowadays they seem to be frequently used for comparing camera formats rather than for understanding lenses.

Thanks for your thoughtful input to this thread and my apologies for any unintended confusion. Regardless of how my comments thus far have been articulated and/or interpreted, I'm in full concurrence with your recent post. And how could it be otherwise? You're correct

[William W]

Bill, first let me apologize for initiating a thread that ultimately resulted in your feeling compelled to spend so much time to generate a thorough response.

I understand what you mean, but, there was never any feeling of compulsion. I choose what I choose, my free choice entirely.

But... on the other hand, now I don't feel so bad about generating so much confused discussion on what we all intend to be a learning forum. Because ultimately it has ended well because your post is an excellent tutorial on the topic. You included the conditions/qualifications that are typically left out of the broad statements made on the topic, provided well articulated explanations and visual examples.

Thanks. Seriously Thank you. That’s usually the main purpose of my choices here.

(WW) I think that there is a lot of misinformation on the www and part of the reason for that misinformation is poor quality inaccurate regurgitation.

Completely agree.

Good – I am beginning to understand your opening post better.

(WW)...I think that you are missing the point of the rationale of why DoF Calculations were ‘invented’...

Not at all. Fully aware of the value/use of standards and standard practices. In my professional life I served on engineering standards committees once or twice.

Now I much better understand all your posts on this conversation.

(WW)... I think that you are applying and or thinking about DoF calculators and Charts the wrong way about...

Not at all for my own use. But nowadays they seem to be frequently used for comparing camera formats rather than for understanding lenses.

Excellent. I really understand what you meant throughout all this conversation

. . .Thanks for your thoughtful input to this thread and my apologies for any unintended confusion. Regardless of how my comments thus far have been articulated and/or interpreted, I'm in full concurrence with your recent post. . .

The ”unintended confusion” is sometimes inevitable with written word forums.
The old Roman face-to-face style is much simpler and quicker.
Thanks for hanging in and thank you especially for this summing up response – such is very valuable for the archival integrity and quality of the thread.

WW