Depth of field (DOF) does NOT change with sensor size

[WayneF]

I think that is everyone!
It all sort of makes sense. Not sure I can regurgitate it yet.
The only two things I do not get, one why is it called "Circle of Confusion"?
And secondly, it is stated as fact in multiple places the formula, and specific values for CoC. I am perfectly willing to accept 0.3mm for full frame and 0.2mm for a standard 1.5 crop sensor. Even reading the Wiki page for some reason I cannot wrap my head around this explanation. It intuitively makes sense to me, but I just do not follow the "official" definition.

Tim

Tim, I think you refer to the page at Circle of confusion - Wikipedia

Circle of Confusion is the larger unfocused blob of an unfocused spot in a photo. You have surely seen unfocused areas before. That articles first image in top right page corner shows a diagram of it. An unfocused blob just spreads and grows larger.


About 3/4 way down that page are quotes of articles from 1829 and 1832, calling it Circle of Confusion. I don't know if that was first usage, but it already had that name then. That was a different time, not much was yet known about optics, and the unfocused blob was indeed a Circle of Confusion (there was no detail in it). That date was before cameras, that was about microscope and telescope lenses. Then the 1866 article after Daguerre had his early camera in 1840. At at the 1889 article, we had the first Kodak Brownie (one year before). Anyway, it is an early name for a very old concept. And sort of a quaint name from different times. But they were interested in calculating CoC size then.

The first text line (by that top diagram, about the diagram) is:

Diagram showing circles of confusion for point source too close, in focus, and too far
In optics, a circle of confusion is an optical spot caused by a cone of light rays from a lens not coming to a perfect focus when imaging a point source. It is also known as disk of confusion, circle of indistinctness, blur circle, or blur spot.

So CoC is simply the name of the larger blob diameter of a (tiny) unfocused point source, like the diagram shows. That should be clear enough. It is a spot of blur. If we are going to compute its size, we need a name for it.

Depth of Field calculators use the term CoC slightly differently for their purpose, as the name for the specification for the maximum allowable unfocused blob that is still too small for our eyes to see when enlarged to a usable viewable print (the 0.03 mm for full frame computations is a hypothetical dimension used for reference). And specifically, for standardization, the spec for DOF is to use an 8x10 inch print to compute DOF on. So it computes the unfocused blob diameters and the DOF distance limits are where the CoC blobs grow larger than this reference CoC, and thus becomes large enough to be visible blur to our eye when enlarged for viewing. If Depth of Field is said to be say 10 to 20 feet, and it was a full frame sensor (0.03 mm CoC) enlarged to 8x10 inch viewing size, then the CoC at exactly both 10 and 20 feet is 0.03 mm (regardless if there is any unfocused blob there or not). That 0.03 mm is on the sensor, it is larger on the enlarged print, at just the limit of our eyes capability there. And larger outside that range (blurred), and less inside it (not blurred). But the calculated DOF limits means the CoC calculation that did reach that maximum allowable amount (at 10 or 20 feet) could still call it sharp enough for our eye, when enlarged to a 8x10 inch print size.

The arguments (like Donny here) that the sensor size has no effect on the DOF simply don't know enough about DOF yet. DOF has been studied for more than 100 years, but Donny hasn't actually started yet. Sounds like he just found crap on the net that gave him the idea. The net is sort of a strange world, some of it is Not to be trusted. Any idiot can post anything. Some of it may be good entertainment.

Yes, it is certainly true that the sensor has no effect on the image that the lens creates. And that is what Donny is trying to say. However, DOF is a very different deal, because the sensor, and even most film sizes are simply too small to view any critical detail. We do not look at them, we always make larger prints, or show it large on a video monitor, or project a slide or movie on a large screen, so that we can see it big. Or use a 10x loupe if we do have to look at the negative. And we want that enlarged print to look good too, so we compute DOF on that enlarged size of blur.

Of course, that enlargement certainly enlarges blur too, and all that matters is how that enlargement looks. The Depth of Field formula computes the CoC blur enlarged to 8x10 inch print size. And a tiny sensor certainly requires much more enlargement to get to 8x10 inch size, so yes, sensor size ABSOLUTELY affects Depth of Field calculations. Sensor size does not affect the image already on the sensor, but it does of course affect the images that we enlarge from it.

Technically, the choice of specification for maximum allowable CoC is what adjusts DOF for enlargement size. For the 35 mm film frame, Sensor diagonal / 1442 = 0.03 mm (common in Japanese cameras, or we still see /1500 = 0.0288 mm sometimes). The Sensor diagonal size is what adjusts CoC in the DOF formula for the enlargement to 8x10 inch print size. A smaller sensor computes a smaller CoC, which is worse DOF due to the greater enlargement required. HOWEVER, the smaller sensor necessarily must use a much shorter lens on its cropped size, which works oppositely in the DOF formula. Not equally, a greater effect, but oppositely.

 

[Ysarex]

Depth of field remains the same with the same lens and with different crop sensors.

Not sure what you're trying to say with that sentence. Crop sensor typically refers to an APS sensor and different APS sensors are still APS sensors so they're the same size. So yep, DOF stays the same with the same lens on any APS size sensor.

What's happening is the crop sensor is just capturing a portion (let's say 1.5x or 2x) of what a full frame sensor would capture. So when you blow it up to match the FF,

You can't blow it up to match the FF. The FF image contains more information -- that info is not in the crop sensor image. If you use both cameras to take the same image the crop sensor image will have more DOF than the FF image. If you use the same lens from the same place on both cameras the FF image will have more DOF and will be a different image showing a wider FOV.

Joe

the depth of field is reduced. It's like changing a normal lens to a telephoto lens shooting from the same distance. The DOF is reduced. See my earlier post. Depth of field (DOF) does NOT change with sensor size

 

[petrochemist]

what are you trying to imply that i would not be able to follow?
really?

also this does not disprove that DOF does not change when you take the same shot with the same lens and same distance using both a crop sensor camera and then full frame or even medium format, ..
DOF does not change, the crop changes, but not the DOF the DOF is exactly the same a 50 mm lens craps out the same amount of DOF on both a crop sensor camera and a full frame camera, the only thing that changes is the crop of course, but the lens still works the same way as far as DOF goes.. Period & FACT!!!!!!!

Perhaps we need to go back to basics & define DOF?
Depth of field is the region of the subject that will look acceptably sharp in the image.
If we magnify the image a given degree of blur will be more noticeable, so how can the region that is acceptably sharp be the same???

Take any image & view it full screen unmagnified & access the limits of what looks focused to you. Magnify the same image 2 fold & view in the same way (effectively cropping the image) you WILL see more blur than originally the DOF is changed.

 

[Tim Tucker 2]

IMO photography is based on visuals, not charts and numbers.

Absolutely, and highlights the many problems with misconceptions and theories in threads such as these.

What we try to do as humans is *rationalise a logic* to explain how things work, but what generally happens that we create a framework or narrative of logic to satisfy our desire to have it understood. We then defend that narrative and refuse to *see* it in any other way, it becomes a loggerheads to and fro revolving around a simple "I'm right and you're wrong," *discussion*.

@donny1963, like many others, is basing that logic around an understanding of how cameras work. And just like many others there are blindingly obvious contradictions in the logical narrative that we create, leaps and assumptions that have no basis in actual observation but only in how we think it *should* work.

A lot of it fails because of one simple and obvious flaw:

We fail to look objectively and understand how we see through our own human eyes.

Instead we see what we understand and have decided is true, we glance at things to re-inforce our own pre-determined opinions.

In trying to create a logical narrative @donny1963 is following the logic of how a lens forms an image on a sensor. It completely ignores that we see images through human eyes with all the distortions of perception that comes from viewing with human eyes. It completely ignores over half the equation of how images are formed. Images are formed on a 2D output, then light reflected/generated from the image is focussed by our eyes and the image is perceived and interpreted through human vision and human understanding.

Images are not and never have been about how a camera sees but always have been based on an understanding of how we as humans see an react to what we see. The fundamental error here is not that the exact same lens behaves in the exact same way regardless of the sensor, but that it tries to describe the whole process of forming an image by how the lens works alone. It pays absolutely no attention to either understanding the subject or how we as humans view and interpret the finished image.

Here is an image to try and demonstrate this, it is exactly the same image and so formed by exactly the same lens on exactly the same sensor:

You may be fooled into thinking that the image is crisp and sharp when you glance, but actually there is very little that is either sharp or in focus.

Here it is again enlarged:

You can see that your entire assumptions of sharpness and therefore what is actually in focus change simply by altering the size of the finished image against the same viewing distance. Perception of sharpness is illusion and dependant on the assumptions we make when viewing the finished image with human eyes.

DOF is all about how we *perceive sharpness* in a finished image, it's about human perception and has far less to do with aperture and how a lens works than many photographers are prepared to admit.

To go one stage further and demonstrate that understanding the subject is more relevant than how the lens forms the image here is another one:

You may perceive that the near water is sharp, but then it remains sharp beyond the remains of the pier that is clearly out of focus. I know that many will make assumptions based on their understanding of photography in that they may assume some focus stacking. But that is an example of trying to quantify what you see into the mechanics you understand. The truth is that there is no focus stacking and the effect is simply due to smooth reflections having a higher acutance and therefore appearing sharper to the human eye.

What I'm trying to illustrate here is that how we interpret sharpness in images is mainly down to human perception. That by trying to force a complete understanding based upon only how a lens works we completely fail to understand how the illusion of DOF and sharpness really works in a finished image. The whole premise upon which DOF is based, human perception and not lens mechanics.

 

[BananaRepublic]

Is light a particle or a wave ?

 

[Ysarex]

what are you trying to imply that i would not be able to follow?
really?

also this does not disprove that DOF does not change when you take the same shot with the same lens and same distance using both a crop sensor camera and then full frame or even medium format, ..
DOF does not change, the crop changes, but not the DOF the DOF is exactly the same a 50 mm lens craps out the same amount of DOF on both a crop sensor camera and a full frame camera, the only thing that changes is the crop of course, but the lens still works the same way as far as DOF goes.. Period & FACT!!!!!!!

Perhaps we need to go back to basics & define DOF?
Depth of field is the region of the subject that will look acceptably sharp in the image.
If we magnify the image a given degree of blur will be more noticeable, so how can the region that is acceptably sharp be the same???

Take any image & view it full screen unmagnified & access the limits of what looks focused to you. Magnify the same image 2 fold & view in the same way (effectively cropping the image) you WILL see more blur than originally the DOF is changed.

Exactly! You had this right back at the beginning of the thread in post #5 -- now we're at post #65. DOF is defined on the print/final image at the final enlargement size and viewing distance. That's where blathering idiot in the video trips himself up. He's trying to define DOF as something that's happening at the film/sensor plane. DOF doesn't happen there -- it's not defined there. You can talk about what's happening with the projected image from the lens on the film/sensor but you can't call that DOF. That's not how DOF is defined.

Most DOF calculators and basic equations assume default values for print/image size and viewing distance but just because those are assumed doesn't mean they go away. The DOF calculator at Cambridge in Colour is instructive. It allows you to change print size and viewing distance. The definition of DOF requires those variables. Blathering idiot left them out of his video.

There is a reason why DOF is defined the way it is and there is a reason the variables used in the equations to calculate DOF are subject distance, focal length, f/stop, sensor size, print size and print viewing distance. We could do what this blathering YouTuber:
does and take f/stop out of the equation and use aperture (entrance pupil) instead. But which one of us while taking a photo knows the diameter of their lens aperture. Tim has hinted at or plain said this back through this thread: That's not helpful to photographers. Photographers are very familiar with lens focal length, f/stop and subject distance (the three variables they are likely to be able to change while taking a photo). DOF is complicated enough and it doesn't simplify the topic to force photographers to use uncommon variables like entrance pupil.

We can express DOF in different and possibly simpler terms but not in more helpful/useful terms.

DOF = magnification + f/stop. Yep I always know the magnification when I'm taking a photo -- not very helpful.

Put a tripod on the ground and don't move it. Mount any camera you like with any size film/sensor. In every photo without changing the subject position or tripod position set the exact same FOV and at the same physical aperture -- all cameras will render the same DOF as viewed in an 8x10 print. I didn't have to mention focal length, and I almost sound like the YouTuber I just posted a link to above, but I controlled for focal length by specifying a constant FOV (he missed that). This kind of stuff can be instructive. It's good to know that focal length and subject distance = magnification. But when we're out there taking photos this kind of stuff is not very helpful. Tim is spot on from back in post #2 with his comment about "equivalence priests," "I find that it does allow you a more complete understanding of the relationships between the fundamental controls of the camera, but is also used as a sort of *geek speak* by those who want to create the illusion that they understand photography and digital better than anybody else, rather than my preferred method of using understanding to create better images."

More than 100 years before any of us were born DOF was well understood by our industry. And those shoulders we stand on very wisely defined and described DOF in the most helpful terms they could so that we as working photographers can think about f/stop and not entrance pupil. So that we don't have to figure magnification but rather pay attention to our lens focal length and subject distance. And finally we can view the expected result in our finished photo.

Way back in post #27 Braineack put up what is arguably the most helpful post in the entire thread. He posted the two photos in which you can see the obvious -- hey look at that: changing sensor size changed the DOF. Look at any standard definition of DOF and you're likely to see the term "acceptably sharp." Think about that term -- "acceptably?" We use math equations to calculate DOF. Imagine a mathematician saying 2 + 2 = something acceptably close to 4. DOF as defined is a phenomenon of human visual perception. It defines an aspect of how we see photographs. It is not defined as what a lens craps out.

To control DOF I can't think of a more helpful way to explain it to photographers other than it's a function of subject distance, focal length and f/stop with sensor size and print enlargement/viewing distance functioning as fixed variables not normally accessible to change in the field.

Joe

 

[deeky]

DOF as defined is a phenomenon of human visual perception. It defines an aspect of how we see photographs. It is not defined as what a lens craps out.

This, right here. I've been watching this post, but I have to admit I just didn't have the brain energy left to read all of them in depth. This right here, though, I think is the issue of contention in all of this. This made it click for me more than the other charts and graphs and equations. And perhaps more importantly, it takes photography back to the art that it always has been rather than mere science.

Thank you for the clarification.

 

[markjwyatt]

I confess I haven't CLOSELY read through all of the above, but I don't recall a single mention of a "normal lens."

I wonder if that's what causing all the confusion.

Those of us who used a variety of film cameras (old guys), know the normal focal length of a lens changes with the format of the camera. AND since the vast majority of the camera market was not (and still isn't) the pros, most discussion about lenses refereed to 35mm cameras. Oddly, it still does... at least a bit. Thus the terms "full frame" and "crop sensor." These terms are relative to 35mm photography. One sensor is simply a different "format" than the other.

A normal lens is one that approximates what we usually see with our eyes. The focal length of a normal lens differs from format to format. I seldom hear the term used anymore.

A normal lens for a 35mm camera is something like 47mm (most cameras were equipped with a 50mm or 55mm. A normal lens for a square format 12o camera (6x6) is about 80mm. And a normal lens for a 4x5 camera is about 6" (150mm).

So could this be contributing to the debate? Hmmm.

-Pete

Peter- I agree with you. This is what is missing, and [I believe] is key to understanding this issue [see my note at the bottom].

If you compare normal lens on a 4/3 sensor with a given shot (say subject at 6', f5.6, 1/125 s) to a full frame with its normal lens (subject at 6', f5.6, 1/125 s)- I.E., The same picture, the 4/3 camera will have more depth of field. This is the case. Some say it does not matter. You can put on a longer lens and stand back. But you cannot always stand back that far. You may run into a wall, other objects may start interfering with the shot etc. Now this does not make the problem insurmountable. Maybe you want the greater depth of field. Maybe you can shoot the 4/3 at f2.8 and 1/500s. In general you just use whatever tool you have.

It is difficult or even impossible to get the same view, perspective, and DOF at the same distance with a full frame and a crop camera. If your shooting style is to be in close and get small DOF, with normal perspective relations then a larger sensor will help. If you want large DOF then crop may be an advantage. I bring this up at camera shops where they push 4/3, and they always say it is proven a myth. But they seem to ignore this point. In terms of effect on a given image, I do not think in terms of sensor size and 35mm equivalents, but rather am I shooting a normal, wide or long lens, and in those cases, whether it is full frame or crop, or even tiny sensors or iPhone it makes a difference in DOF.

Using the Cambridge DOF calculator (default parameters- 25 cm or ~10" viewing distance, 10 in. print, mfr standard vision);
Compare full frame and 4/3, same picture, same distance, same exposure and same aperture:
A Flexible Depth of Field Calculator

Focus 5m
aperture f/5.6
(4/3 crop factor = 2, assume same shutter speed both cameras, thus same exposure and aperture)

Full frame 50mm lens: 3.69 < DOF < 7.75 meters, FOV 14' 2.00"
4/3 crop 25 mm lens: 2.92 < DOF < 14.52 meters, FOV 14' 2.00"

(FOV: Depth of Field (DoF), Angle of View, and Equivalent Lens Calculator | Points in Focus Photography): note this calculator uses different parameters for DOF and the DOF numbers are a little different, but close and consistent.

Now the 4/3 also has a 2x factor on the aperture, so if we went to f/4 (and doubled the shutter speed)
4/3 crop 25mm lens: 3.31m< DOF < 10m a bit closer to the full frame, but still greater DOF, plus we seem to need faster lenses (and potentially shutters too) on the 4/3 to get close to equivalency.

NOTE: I posted this response to Peter, then read Joe's post above and questioned my conclusions. I then added the numbers and feel my conclusions are sound, but not consistent with Joe's. My conclusion would say that a "normal" lens on a 4/3 camera (taken as 25mm) with equivalent distance, focus, and aperture (and exposure overall if ISO were the same, but this is actually somewhat irrelevant) would have greater DOF than a "normal" lens on a full frame (taken as 50mm). The normal lenses have the same FOV, and I would suspect the same perspective (i.e, normal). Similar conclusions would apply for equivalent wide and long lenses. This is getting interesting and I am learning a lot. My conclusion was intuitive based on experience but not math or experiments, and Joe suggested it was wrong (and I certainly could be wrong).

 

[markjwyatt]

...
Put a tripod on the ground and don't move it. Mount any camera you like with any size film/sensor. In every photo without changing the subject position or tripod position set the exact same FOV and at the same physical aperture -- all cameras will render the same DOF as viewed in an 8x10 print...

Joe- Did I miss something? See above post (Depth of field (DOF) does NOT change with sensor size)

The only thing missing from the quote and my example is perhaps the same focus point, but I think that is implied and very important. Another issue you mentioned is magnification, but I suspect these are equivalent in this case?

 

[markjwyatt]

...

Here it is again enlarged:

View attachment 166140

...

I was always taught with portraits that the key is to have the catch light in the eyes sharp. Get that and the impression of sharpness is favorable. Lose that and you lose the image. In your image, the catch light is pretty sharp.

Here is a picture I took of my dog. I really did not get the catch lights perfectly sharp, and there is an impression of softness (I still like the image but would prefer the eyes to be sharper). If you look closely, the collar is pretty sharp.


Taffy Pose by Mark Wyatt, on Flickr

 

[Ysarex]

I confess I haven't CLOSELY read through all of the above, but I don't recall a single mention of a "normal lens."

I wonder if that's what causing all the confusion.

Those of us who used a variety of film cameras (old guys), know the normal focal length of a lens changes with the format of the camera. AND since the vast majority of the camera market was not (and still isn't) the pros, most discussion about lenses refereed to 35mm cameras. Oddly, it still does... at least a bit. Thus the terms "full frame" and "crop sensor." These terms are relative to 35mm photography. One sensor is simply a different "format" than the other.

A normal lens is one that approximates what we usually see with our eyes. The focal length of a normal lens differs from format to format. I seldom hear the term used anymore.

A normal lens for a 35mm camera is something like 47mm (most cameras were equipped with a 50mm or 55mm. A normal lens for a square format 12o camera (6x6) is about 80mm. And a normal lens for a 4x5 camera is about 6" (150mm).

So could this be contributing to the debate? Hmmm.

-Pete

Peter- I agree with you. This is what is missing, and is key to understanding this issue.

If you compare normal lens on a 4/3 sensor with a given shot (say subject at 6', f5.6, 1/125 s) to a full frame with its normal lens (subject at 6', f5.6, 1/125 s)- I.E., The same picture, the 4/3 camera will have more depth of field. This is the case. Some say it does not matter. You can put on a longer lens and stand back. But you cannot always stand back that far. You may run into a wall, other objects may start interfering with the shot etc. Now this does not make the problem insurmountable. Maybe you want the greater depth of field. Maybe you can shoot the 4/3 at f2.8 and 1/500s. In general you just use whatever tool you have.

It is difficult or even impossible to get the same view, perspective, and DOF at the same distance with a full frame and a crop camera. If your shooting style is to be in close and get small DOF, with normal perspective relations then a larger sensor will help. If you want large DOF then crop may be an advantage. I bring this up at camera shops where they push 4/3, and they always say it is proven a myth. But they seem to ignore this point. In terms of effect on a given image, I do not think in terms of sensor size and 35mm equivalents, but rather am I shooting a normal, wide or long lens, and in those cases, whether it is full frame or crop, or even tiny sensors or iPhone it makes a difference in DOF.

Using the Cambridge DOF calculator (default parameters- 25 cm or ~10" viewing distance, 10 in. print, mfr standard vision);
Compare full frame and 4/3, same picture, same distance, same exposure and same aperture:
A Flexible Depth of Field Calculator

Focus 5m
aperture f/5.6
(4/3 crop factor = 2, assume same shutter speed both cameras, thus same exposure and aperture)

Full frame 50mm lens: 3.69 < DOF < 7.75 meters, FOV 14' 2.00"
4/3 crop 25 mm lens: 2.92 < DOF < 14.52 meters, FOV 14' 2.00"

(FOV: Depth of Field (DoF), Angle of View, and Equivalent Lens Calculator | Points in Focus Photography): note this calculator uses different parameters for DOF and the DOF numbers are a little different, but close and consistent.

Now the 4/3 also has a 2x factor on the aperture, so if we went to f/4 (and doubled the shutter speed)
4/3 crop 25mm lens: 3.31m< DOF < 10m a bit closer to the full frame, but still greater DOF, plus we seem to need faster lenses (and potentially shutters too) on the 4/3 to get close.

NOTE: I posted this response to Peter, then read Joe's post above and questioned my conclusions. I then added the numbers and feel my conclusions are sound, but not consistent with Joe's.

Your conclusions are sound and consistent with mine.

My conclusion would say that a "normal" lens on a 4/3 camera (taken as 25mm) with equivalent distance, focus, and aperture (and exposure overall if ISO were the same, but this is actually irrelevant) would have greater DOF than a "normal" lens on a full frame (taken as 50mm).

I concur but we need to be clear what you mean when you say aperture. If you take the same photograph 4/3 and FF (same perspective, same FOV, same subject focus point) at the same f/stop for both lenses the 4/3 photo will have more (deeper) DOF. I've never said otherwise.

Joe

The normal lenses have the same FOV, and I would suspect the same perspective (i.e, normal). Similar conclusions would apply for equivalent wide and long lenses. This is getting interesting and I am learning a lot. My conclusion was intuitive based on experience but not math or experiments, and Joe suggested it was wrong (and I certainly could be wrong).

 

[markjwyatt]

at the same physical aperture

Ok, so by physical aperture you mean the same actual DIAMETER of the aperture, not the same f/stop number/ratio. What matters to photographers when shooting tends to be f/stop, not physical aperture, but I get what you are saying.

So, this still holds (and I think this is what matters most to photographers shooting pictures and puts the argument in the clearest context as suggested by Peter earlier):

My conclusion would say that a "normal" lens on a 4/3 camera (taken as 25mm) with equivalent distance, focus, and aperture (and exposure overall if ISO were the same, but this is actually irrelevant) would have greater DOF than a "normal" lens on a full frame (taken as 50mm).

The normal lenses have the same FOV, and I would suspect the same perspective (i.e, normal). Similar conclusions would apply for equivalent wide and long lenses.

 

[Ysarex]

...
Put a tripod on the ground and don't move it. Mount any camera you like with any size film/sensor. In every photo without changing the subject position or tripod position set the exact same FOV and at the same physical aperture -- all cameras will render the same DOF as viewed in an 8x10 print...

Joe- Did I miss something?

Yes, I think so. I used that as an illustration of the kind of presentations that are not helpful. I said same physical aperture not aperture and not f/stop. In that example each lens used must have the same entrance pupil. So if a 4/3 camera is one of the cameras used and has an entrance pupil of 10mm then if a medium format camera is used as well the lens would also have to have an entrance pupil of 10mm. The difference in f/stops there would be quite substantial.

Joe

See above post (Depth of field (DOF) does NOT change with sensor size)

The only thing missing from the quote and my example is perhaps the same focus point, but I think that is implied and very important. Another issue you mentioned is magnification, but I suspect these are equivalent in this case?

 

[Ysarex]

at the same physical aperture

Ok, so by physical aperture you mean the same actual DIAMETER of the aperture, not the same f/stop number/ratio.

Yes. You're getting it.

What matters to photographers when shooting tends to be f/stop, not physical aperture,

Yes! That was my point. The YouTube blathering idiots are not helping anyone with their usually misguide and misunderstood contortions.

but I get what you are saying.

So, this still holds (and I think this is what matters most to photographers shooting pictures and puts the argument in the clearest context as suggested by Peter earlier):

My conclusion would say that a "normal" lens on a 4/3 camera (taken as 25mm) with equivalent distance, focus, and aperture (and exposure overall if ISO were the same, but this is actually irrelevant) would have greater DOF than a "normal" lens on a full frame (taken as 50mm).

Absolutely! That's a fact. And that means sensor size is a determinant factor for DOF.

The normal lenses have the same FOV, and I would suspect the same perspective (i.e, normal). Similar conclusions would apply for equivalent wide and long lenses.

Do not make the error of assigning perspective to lens focal length. Perspective is a function of where you put the camera, period. Lens focal length does not determine perspective. So perspective means both photos taken from the same place. This is a critical factor. In order to compare same photos they must both be taken from the same place -- same perspective.

Joe

 

[markjwyatt]

Joe- I think we are entirely consistent. Sorry for not considering your "how not to do it" aspect! We actually use that approach in selling software, we say How Not to Do "X", then list three easy ways that people use to try and handle "X" which we then demonstrate not to be correct. We then follow by demonstrating how our software [or more generally the theoretically preferable approach] is the best way to handle "X".