Depth of field: Full-frame lens vs. APS-C lens, both on APS-C sensor

[CameraClicker]

Many of the on-line calculators have a problem. They suggest the same prime lens on different bodies FF and APS-C for instance will have different DOF if the camera and subject do not move and aperture is the same.

The major factors influencing depth of field are focal length, aperture and subject distance but there will indeed be a different DoF for different sensor sizes because the circle of confusion is a function of film or sensor size. The difference will not be great and will tend to slightly lessen the increased DoF due the increased subject to lens distance when shooting with a crop sensor.
So I would have serious reservations about any DoF program that DID give the same DoF with the same lens, f stop and distance setting but different sensor sizes.

This is an example of what I was referencing. Photo site size may indeed have an effect on DOF. But, I have a full frame body and an APS-C body that has virtually the same size photo sites. I have another APS-C body where there are over twice the photo sites, so I expect those photo sites are much smaller. When I did the test, I used the newer APS-C body with more photo sites. Differences in DOF were less that errors in focusing, so not worth worrying about. I decided to do the test when it was pointed out that one of the on-line calculators was giving results that were exactly opposite what another photographer and engineer was saying should be expected.
In electronics, 10% is considered to be "close enough" for most applications. Sometimes that is not true, and you can get components, usually military grade, that are more expensive and have tighter tolerances. The DOF difference was well under 10%, so not worth worrying about.

 

[Derrel]

Despite an egregiously erroneous statement above, the smaller the sensor or film size, the greater the depth of field at each picture angle. According to a published article written by Herbert Keppler, editor of Modern Photography magazine, KODAK, in fact, literally INVENTED the very small "Kodak disc film" format, as a way to achieve hyperfocal depth of field with a lens that had its focus permanently set to about three feet. Using the then-popular 126 cartridge and 110 cartridge cameras, many snap-shooters were taking out of focus close-up shots. In order to achieve exceptionally deep, expansive depth of field, Kodak engineers determined that the only solution was to MAKE an ALL-NEW, INCREDIBLY SMALL film format! And so they did. They named it the "disc format". It's important to understand this point: Kodak created an all-new, ultra-small film format and cameras to go with it, in order to create a camera that offered huge depth of field from very close, to Infinity. Despite the small size of 110 film, there was not enough depth of field at close distances with fixed-focus lenses, so the solution was to follow the laws of optics, and decrease the film size and decrease the lens length, to get a normal picture angle of view AND expansive, deep depth of field. "Science".

110 film used 13mm x 17mm frames and was introduced by Kodak in 1972. Disc format was introduced a decade later, in 1982, and used 8mm x 11mm frames. The simple practical upshot is that if one wants to achieve tremendous depth of field, where things from very close to very far away are rendered in good focus, the easiest way to do that is to use a camera with a very small sensor, with its commensurately short focal length lenses. This is why smart phones, with their exceedingly short focal length lenses and tiny sensors, are able to create images with exceptionally deep depth of field. For social photography, there is a significant advantage to using a SMALL-sensor digital camera, which will allow the photographer to shoot close-in, social photography scenes that have deep depth of field, so that focusing super-critically, yet still suffering from inadequately deep focus, becomes less an issue than it was with say, a medium-format rollfilm camera, or a 35mm or "135 film format" camera, etc..

Medium format 6x6 cm rollfilm cameras with their so-called normal lens, which is 75mm or 80mm in length, have relatively shallow depth of field with their normal lens. Even the 65mm semi-wide-angle lens on 6x6cm film format has fairly limited depth of field. By contrast, an iPhone 5 has tremendous depth of field with its semi-wide-angle lens--even when the lens is at a wide aperture like say, f/2.4.

Let's get this clear: when making pictures that have the same angle of view, the smaller the film or sensor format, the MORE depth of field there is at equivalent apertures. Do the experiment yourself using your smartphone and your d-slr. If you have ever owned and shot a medium-format rollfilm camera you will KNOW, from actual experience that its "normal lens" creates shallow depth of field images at ranges inside of 40 feet, and that if you want to get more, actual, real, workable depth of field, that you want to move down to a camera that shoots to a smaller-sized capture format.

 

[CameraClicker]

Despite an egregiously erroneous statement above, the smaller the sensor or film size, the greater the depth of field at each picture angle. According to a published article written by Herbert Keppler, editor of Modern Photography magazine, KODAK, in fact, literally INVENTED the very small "Kodak disc film" format, as a way to achieve hyperfocal depth of field with a lens that had its focus permanently set to about three feet. Using the then-popular 126 cartridge and 110 cartridge cameras, many snap-shooters were taking out of focus close-up shots. In order to achieve exceptionally deep, expansive depth of foeld, KOdak determined that the only solution was to MAKE THE FILM FORMAT INCREDIBLY SMALL!!!! and so they did.

110 film used 13mm x 17mm frames and was introduced by Kodak in 1972. Disc format was introduced a decade later, in 1982, and used 8mm x 11mm frames. The simple practical upshot is that if one wants to achieve tremendous depth of field, where things from very close to very far away are rendered in good focus, the easiest way to do that is to use a camera with avery small sensor, with its commensurately short focal length lens. This is why smart phones, with their tiny lenses and tiny sensors, are able to create images with exceptionally deep depth of field. For social photography, there is a significant advantage to using a SMALL-sensor digital camera, which will allow the photographer to shoot close-up scenes that have deep depth of field, so that focusing super-critically becomes less an issue than it was with say, a medium-format rollfilm camera, or a 35mm or "135 film format" camera, etc..

Medium format 6x6 cm rollfilm cameras with their normal lens length, which is 75mm or 80mm, have relatively shallow depth of field with their normal lens. Even the 64mm semi-wide-angle lens on 6x6cm film format has fairly limited depth of field. By contrast, an iPhone 5 has tremendous depth of field with its semi-wide-angle lens--even when the lens is at a wide aperture like say, f/2.5.

Let's get this clear: when making pictures that have the same angle of view, the smaller the film or sensor format, the MORE depth of field there is at equivalent apertures. Do the experiment yourself using your smartphone and your d-slr. If you have ever owned and shot a medium-format rollfilm camera you will KNOW, from actual experience that its "normal lens" creates shallow depth of field images at ranges inside of 40 feet, and that if you want to get more, actual, real, workable depth of field, that you want to move down to a camera that shoots to a smaller-sized capture format.

Yep! Exactly the deal with a cell phone camera, and many point & shoot cameras. Super tiny sensor, or film, super short lens. A very short hyperfocal distance, almost everything is always in focus. They are trying to get the angle of view you might get from 35 mm or 50 mm, on an SLR.

But if you hold every variable, except the sensor size, you can see the effect of just changing sensor size. No effective difference.

 

[WayneF]

But if you hold every variable, except the sensor size, you can see the effect of just changing sensor size. No effective difference.

You're making a joke? You cannot possibly hold every variable the same. If focal length and distance are the same, the lens can be the same, but the cropped view is still 1/3 smaller (FX to DX), not at all the same picture. To have any meaning, we do need to compare like pictures. We could crop the larger sensors image smaller so we could compare, I guess. You could be right then. LOL

Plus the smaller frame requires greater enlargement, which affects CoC, which affects DOF calculation. DOF is not an absolute from the lens, the overall concept simply depends on how you enlarge it to view it. The normal DOF calculation always assumed an 8x10 print viewed from 10 or 12 inches (standard conditions). It is about what we see. CoC is the dot the eye can see, when enlarged to the same print size. The smaller cropped sensor suffers that way.

The smaller cropped sensor is considered to have greater depth of field because we do NOT keep everything the same. We only keep the image the same. This same view requires either a shorter lens (more DOF), or standing back farther (more DOF).

 

[Derrel]

The smaller the capture format, the shorter the focal length needed for a given angle of view. The smaller the format, and the shorter the lens, the shorter the focus distance needed to achieve hyperfocal depth of field. A very good article on depth of field and sensor size was written a few years back by Bob Atkins.Depth of Field, Digital Photography and Crop Sensor Cameras - Bob Atkins Photography

This is probably the single best article on the subject I have seen. What's important to note, and Bob notes this, is that depth of field does not increase or decrease in a simple, linear manner! Look about 3/4 of the way down the page, to the chart that has the light aquamarine, lime green, and then lavender-colored segments!!!! DOF examples are broken down into three zones: Close-up; intermediate; and distances close to the hyperfocal distance.

His chart shows APS-C, FX or 35mm, and 6x9 cm film format, and charts out the depth of field for a 46-degree angular field of view, across all three formats, at close-up range (0.15 meters), intermediate distance of 2.0 meters, and at 5.0 meters.

Once again, this article is comparing three distinct sizes or "formats", ALL with the lens required to give a 46 degree angular view. It's a simple fact: moving to a smaller film or smaller digital sensor format increases depth of field at close-up range, at intermediate ranges, and at longer distances. And the important thing to note is this article is charting out the depth of field differences with all three camera formats using their size-appropriate "normal lens" length!

As Wayne mentioned, we need to compare similar pictures, made with different camera formats and different lens lengths. That is what this article helps people understand.

 

[CameraClicker]

But if you hold every variable, except the sensor size, you can see the effect of just changing sensor size. No effective difference.

You're making a joke? You cannot possibly hold every variable the same. If focal length and distance are the same, the lens can be the same, but the cropped view is still 1/3 smaller (FX to DX), not at all the same picture. To have any meaning, we do need to compare like pictures. We could crop the larger sensors image smaller so we could compare, I guess. You could be right then. LOL

Plus the smaller frame requires greater enlargement, which affects CoC, which affects DOF calculation. DOF is not an absolute from the lens, the overall concept simply depends on how you enlarge it to view it. The normal DOF calculation always assumed an 8x10 print viewed from 10 or 12 inches (standard conditions). It is about what we see. CoC is the dot the eye can see, when enlarged to the same print size. The smaller cropped sensor suffers that way.

The smaller cropped sensor is considered to have greater depth of field because we do NOT keep everything the same. We only keep the image the same. This same view requires either a shorter lens (more DOF), or standing back farther (more DOF).

No sir, not joking. Sure you can keep all the variables the same! Put the camera on a tripod. Shoot down a hallway. Use a stationary target. Use some basic measuring equipment. Light with strobes if necessary.

I spent a career doing engineering, mechanical, electrical and software. If you want to understand a system, keep everything the same except what you want to study. Then see the effect of that change. That is a pretty standard method of dealing with complicated systems. I'm sure you even use that method yourself sometimes when setting up strobe lighting? Set the first light, get it the way you want. Turn off that light, set the next light get it the way you want. Repeat for the other lights. Once they are all set, turn the lights back on and take your photo. Sound familiar?

When I got my first dSLR it was a 30D, and I was told the smaller sensor size would increase the effective length of the lens! At the same time, I was shown an Olympus that had a crop factor of 2, so it was supposed to make a 200 mm lens look like 400 mm! If you don't worry about framing and instead you look at each photo at 100%, you see what nonsense that is.

An 8 X 10 may make a degree of sense when dealing with portraits. It may also make some sense if you have a full frame Nikon and and APS-C Nikon and both have the same number of pixels. As you move to other brands and newer models, pixel counts are all over the map, as are photo site sizes. Cropping may also nullify the 8 X 10 argument. The point remains that while you may use an APS-C body differently than a full frame body, any change in DOF from the sensor itself will be extremely small or non-existent, while changes in DOF from moving closer or further away will be noticeable. Not everyone limits their photography to portraits and understanding what is really affecting your photos, and why, can be quite useful.

 

[WayneF]

No sir, not joking. Sure you can keep all the variables the same! Put the camera on a tripod. Shoot down a hallway. Use a stationary target. Use some basic measuring equipment. Light with strobes if necessary.

No, you changed the sensor size. That's a big deal, a new game then. You cannot just ignore it. You surely don't imagine tiny compact camera sensors are the equal of full frame DSLR sensors?

I spent a career doing engineering, mechanical, electrical and software. If you want to understand a system, keep everything the same except what you want to study. Then see the effect of that change. That is a pretty standard method of dealing with complicated systems. I'm sure you even use that method yourself sometimes when setting up strobe lighting? Set the first light, get it the way you want. Turn off that light, set the next light get it the way you want. Repeat for the other lights. Once they are all set, turn the lights back on and take your photo. Sound familiar?

Who didn't? But we do have a tendency to imagine our own solutions, imagine how it "ought" to work, making up how we would have done it, and then trying to tell everyone how it actually works that way. Not always true though, esp in an unknown field, but you can learn to break that habit. First step is to think back "how am I certain this is true?".

You're overlooking a major variable, sensor size. Just a little bit of of study can make it very clear, this stuff is all very well known. You just have not started learning yet. You are still imagining your own solutions, which are rather incomplete.

When I got my first dSLR it was a 30D, and I was told the smaller sensor size would increase the effective length of the lens! At the same time, I was shown an Olympus that had a crop factor of 2, so it was supposed to make a 200 mm lens look like 400 mm! If you don't worry about framing and instead you look at each photo at 100%, you see what nonsense that is.

Of course the lens cannot and does not change by just putting it on some camera body. No one thinks the lens changes.

But the cropped sensor does crop that lens view, it only sees a smaller cropped view. With the same lens at same distance, the cropped sensor only sees a narrow view, same view as the uncropped sensor would see with a lens longer by the crop factor.... 1.5x or 1.6x longer focal length. This view comparison is so that users familiar with 35 mm film (It was very popular for more than 50 years), know what the contraption is going to do. This crop factor is called Equivalent Focal Length, and it describes the cropped view, as relative to the uncropped, only important because it compares to the old familiar 35mm film sensor size. Then old timer 35mm film users know what a 100 mm lens does on a cropped sensor... the cropped view it sees mimics a 1.5x or 1.6x longer lens on the uncropped sensor. And it does mimic a telephoto view (if as seen on uncropped). Any cropping (any image, any time later in a photo editor) shows this same telephoto effect when enlarged back to same size. It's all well known.

Different film sizes worked exactly the same way too - It was all well understood way back then too. However it was uncommon to be able to use the SAME lens on different film sizes (we could at least try in darkroom enlargers). But it is why film sizes normal lens focal length varied with their film diagonal dimension. Equivalent focal length concept. Same thing. Same thing with DOF too. Smaller film size has greater DOF (but less effective resolution when enlarged more).. Remember Ansel Adams 1930s F/64 Group? This was trying to get sharper images on the 8x10 film size. This has all been well known for many decades. Digital is about the only thing new. Newcomers have to get up to speed with the facts first though.

An 8 X 10 may make a degree of sense when dealing with portraits. It may also make some sense if you have a full frame Nikon and and APS-C Nikon and both have the same number of pixels. As you move to other brands and newer models, pixel counts are all over the map, as are photo site sizes. Cropping may also nullify the 8 X 10 argument. The point remains that while you may use an APS-C body differently than a full frame body, any change in DOF from the sensor itself will be extremely small or non-existent, while changes in DOF from moving closer or further away will be noticeable. Not everyone limits their photography to portraits and understanding what is really affecting your photos, and why, can be quite useful.

It makes full sense to anyone that learns the actual facts instead of making up their own notions. It is NOT at all about pixels. Pixels merely attempt to digitally reproduce the analog lens image. The best job pixel resolution can do is to reproduce the lens image resolution well (pixels cannot make it better, or even as good - even if we ignore Bayer, there is Nyquist, etc). The overall concept is about the analog lens image projected onto the sensor, which we hope to copy well. Cropping makes that source image smaller, which then requires more enlargement to appear equal again, and we all know enlargement is a resolution disadvantage. Variable CoC with sensor size accounts for that, and affects DOF.

And DOF is certainly Not an inherent property just of the lens as you imagine. Focal length, aperture, and subject distance are factors, as is sensor size and the subsequent enlargement necessary to see it. Cropped sensors require more enlargement to compare at same size. DOF details have been known for at least 100 years.
It is really not a Day One subject for the newbies however.

The issue is that you're simply believing your own wrong assumptions (imagining DOF must only be about the lens). It just proves you have done no study. You really need to do a bit of study about how things work if you're going to try to explain how things work. It becomes pretty funny publicly otherwise. This stuff is all very widely known.

 

[Derrel]

Dude--sensor size, or film size, plays a HUUUUUUUUUUUUUUGE role in the lenses used, and the depth of field possible. Anybody who does NOT understand that smaller sensors have deeper depth of field than larger and larger sensors or larger and larger film sizes really does NOT understand the way cameras and their lenses actually function.

How about this: read Wayne's post carefully. Then, spend a few minutes and check out the FACTS, and then please, go through and kill-file your erroneous nonsensical posts made in this thread. It's no wonder people never can learn the basics when so much B.S. is put forth with so much earnestness and sincerity.

Background blur and its relationship to sensor size: Digital Photography Review

Please avail yourself of the very clear, simple chart: Tabelle_2.png

 

[theraven871]

I have read so many articles on this.
YES, there IS a difference. I'll simplify.

Your crop factor (on APS-C) applies not only to the focal length, but also to the aperture as well.
Thus, lets for argument sake say we want to purchase the equivalent of a 24-70mm f2.8 on a CROP sensor body.
Most people will tell you the 17-50mm f2.8 is identical. This is because a 1.5x crop on 17-50 is roughly 24-70mm.
What they DON'T tell you is that you also need to multiply the aperture.

So, for this example, a 17-50mm f2.8 is NOT equal to a 24-70mm f2.8.
When all the math is done, a 17-50 f2.8 is very close to a 24-70mm f4.

Of course, this math applies to all crop sensor lenses.
A 70-200mm f2.8 lens on a crop sensor body becomes a 105-300mm f4.
A 50mm F1.8 becomes a 75mm f2.8

And so on....
You just simply need to multiply BOTH your focal range AND aperture by your crop factor
(Crop factor is 1.5x for Nikon and 1.6x for Canon).

 

[Braineack]

are you suggesting the exposure changes between the two?

 

[Derrel]

are you suggesting the exposure changes between the two?

I think he's suggesting the well-known "equivalence" concept that the smaller sensor's greater depth of field requires a wider aperture at each and every equivalent field of view, to get equal defocused background or equal depth of field. This is pretty well known and accepted.

 

[Braineack]

okay. the difference is 65% less (DX to FX) DOF then, which I did do the math for in post #15.

seems to be direct relationship in the difference between the CoC area and the difference in DOF.

 

[WayneF]

I have read so many articles on this.
YES, there IS a difference. I'll simplify.

Your crop factor (on APS-C) applies not only to the focal length, but also to the aperture as well.
Thus, lets for argument sake say we want to purchase the equivalent of a 24-70mm f2.8 on a CROP sensor body.
Most people will tell you the 17-50mm f2.8 is identical. This is because a 1.5x crop on 17-50 is roughly 24-70mm.
What they DON'T tell you is that you also need to multiply the aperture.

So, for this example, a 17-50mm f2.8 is NOT equal to a 24-70mm f2.8.
When all the math is done, a 17-50 f2.8 is very close to a 24-70mm f4.

Of course, this math applies to all crop sensor lenses.
A 70-200mm f2.8 lens on a crop sensor body becomes a 105-300mm f4.
A 50mm F1.8 becomes a 75mm f2.8

And so on....
You just simply need to multiply BOTH your focal range AND aperture by your crop factor
(Crop factor is 1.5x for Nikon and 1.6x for Canon).

Wording is very misleading. wrong as stated. The problem is this means to speak only of DOF, but it never mentions DOF to specify that single meaning. It seriously needs more careful wording.

f/2.8 aperture is f/2.8 aperture, exposure-wise on any camera, of any size or crop, or uncropped. There is not any "equivalent" f/2.8. It either is, or it isn't.

However apparent DOF does increase on smaller cropped sensors. A 16mm lens does have a greater DOF than 24mm. To compare at the same field of view, the smaller sensor either has to use a longer lens (greater DOF) or if with same lens, they have to stand back farther (greater DOF), to see the same view, to allow any comparison. But it requires lens extension or a teleconverter to change f/2.8 to be f/4.

 

[theraven871]

are you suggesting the exposure changes between the two?

I think he's suggesting the well-known "equivalence" concept that the smaller sensor's greater depth of field requires a wider aperture at each and every equivalent field of view, to get equal defocused background or equal depth of field. This is pretty well known and accepted.

Exactly. My point was that a Full Frame lens on an APS-C sensor will produce less "depth of field" than with a full frame sensor.
This is because the Aperture is not the same on different sensors (using Full Frame as a reference point).

Where things start to get even more confusing is the focal range has also changed.
If you were focused on a subject 10 feet away at 50mm on a full frame
(then took that same lens on a crop sensor)
You will then need to step backwards to keep the same composition. I don't know the math off the top of my head, but then you may be somewhere near 12 feet away.
Not only did the aperture change from the crop sensor, but now since the distance to the subject changes, the DoF also changes.

So, if I'm looking at this correctly, their are TWO factors which will create a wider depth of field on a cropped sensor.
The fact that your aperture will multiply based on your cropped ratio (1.5x or 1.6x) AND that you need to change the distance to your focus point.

 

[astroNikon]

So to be equivalent
does the 17-50 need to be a f/1.4
to be equivalent to a 24-70/2.8 ?