The Falacy of FF Cameras having Flatter DOF

[William W]

. . . It doesn't go further. . .

Ah! - So you are not only quoting only a PORTION of the Textbook - but you do not have ALL the text book to research so you can cite relevant quotes which would be applicable to this conversation.

That’s not really a good position if you want to mount a logical argument which is based upon citing quotes from a textbook.

*

OK, that stated, now let’s address your question -

Perspective
The term perspective is applied to the apparent
relationship between the position and size of objects
when seen from a specific viewpoint, in a scene
examined visually.

What is meant with the size of an object? My finger has a certain size,. Doesn't change when holding it before my eyes. But it does change "in a scene examined visually". A different experience holding my finger close or far. The relative size is changing. But relative to what?? To other subjects in that view. One can also say that the magnification is changing. And than we can say relative to the view we have.

I think that there might be a language issue – an issue of the nuance of meaning: it is not about the size of the object - it is about the apparent (apparent means “how it appears”) relationship of the sizes and positions of the objects in the scene.

In Paul’s Photos the apparent relationships of the sizes of the objects have NOT changed.

For example – in BOTH IMAGES:
> the little rock at the right hand side of the group of three rocks is about half the height of the big two rocks
> the little white outcrop of rock in front of the big brown rocks is about ¼ as wide as the big brown rock on the left
> the white water is about the same width as the big brown rock

All these are examples of “the apparent relationship between the position and size of objects
when seen from a specific viewpoint”

If we were to move the camera (i.e. “change the viewpoint"), then those relationships of size and position would change – but cropping does NOT change those relationships of size and position.

See here -



And now look at Paul's two photos - the RELATIONSHIP of the size and position of these objects in the scene are the same - within each photo.

WW

[ajohnw]

Well I have never mentioned "blur".

I did mention DoF, several times.

With reference to DoF it is indeed a FACT that: "people who use various sizes of crop camera can NOT produce the same results as a full frame camera can."

*

This fact is demonstrated in Post #55.

This fact is referenced in "Equivalence in Photography", James Joseph (op cit).

This fact can be demonstrated by using any reputable DoF calculator and that has also been mentioned in Post #55.

This fact can be proved by Mathematics, beginning at First Principles of the Calculation for DoF.

*

It is entirely your choice to refuse to acknowledge these facts - and that's fine.

Accordingly I do suggest that we call it quits in this conversation and I also suggest that maybe you might do the exercise that I outlined and also research the definition of "Perspective" from a few reputable sources, as I also suggested.


WW

So now blur and has no relationship to DOF or the size of final image ? I feel you are being a bit pedantic Bill because I am sure you are aware that crop cameras can produce extremely similar final images to those taken with full frame. All I have done is shown that this is still the case when the crop factor is used on lens focal lengths and images are enlarged to the same size.

Either that or the calculator linked to is completely and utterly wrong which I don't think it is other than it would be rather difficult to include variations in the quality of the lens used which will vary. The FF shot is also capable of being enlarged more than say an APS shot. That is mostly down to the fact that both use the same "quality" of lenses. In fact given the cash an F rather than FS lenses. Fortunately that doesn't apply comparing M 4/3 with APS sizes as the lenses can economically have higher resolutions. It probably doesn't apply to some FS lenses too but not to the same extent.

It is true in some cases that the framing can not be exactly identical.

I'll post the link again. Perhaps you could explain why the results would differ when the crop factor is accounted for and the other condition I mentioned is met in the final image. That effectively is crop factor based magnification as in some cases the framing can't be identical but feature in the shot can be.

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

The scale may need adjusting to 1cm to 60m but make the latter as long as you like. The utility does have a failing in that aspect but it's not relevant to the discussion.

John
-

[george013]

Ah! - So you are not only quoting only a PORTION of the Textbook - but you do not have ALL the text book to research so you can cite relevant quotes which would be applicable to this conversation.

That’s not really a good position if you want to mount a logical argument which is based upon citing quotes from a textbook.

*

OK, that stated, now let’s address your question -



I think that there might be a language issue – an issue of the nuance of meaning: it is not about the size of the object - it is about the apparent (apparent means “how it appears”) relationship of the sizes and positions of the objects in the scene.

In Paul’s Photos the apparent relationships of the sizes of the objects have NOT changed.

For example – in BOTH IMAGES:
> the little rock at the right hand side of the group of three rocks is about half the height of the big two rocks
> the little white outcrop of rock in front of the big brown rocks is about ¼ as wide as the big brown rock on the left
> the white water is about the same width as the big brown rock

All these are examples of “the apparent relationship between the position and size of objects
when seen from a specific viewpoint”

If we were to move the camera (i.e. “change the viewpoint"), then those relationships of size and position would change – but cropping does NOT change those relationships of size and position.

See here -



And now look at Paul's two photos - the RELATIONSHIP of the size and position of these objects in the scene are the same - within each photo.

WW

If you opened the link you knew.

It says "the term perspective is applied to the apparent relationship between the position and size of objects when seen from a specific viewpoint, in a scene examined visually. "
Not relative to each other, but to the frame.

Well, I ordered that book, 10th edition. If it is what I think it is, it's the book I was searching for.
Up to now it helped me to believe I'm not a fool.

George

[William W]

. . . I am sure you are aware that crop cameras can produce extremely similar final images to those taken with full frame.

Yes.

I clearly implied that in Post #55 here:

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.

However, my main point has not deviated throughout this entire conversation: APS-C cameras can NOT produced the same shallow DoF as a 135 Format camera can: that's not being pedantic, that's simply re-iterating the same fact.

***

So now blur and has no relationship to DOF or the size of final image ?

That’s an argumentative technique to imply a quote that someone stated: I have never stated that.

Yes background Blur and DoF have a relationship – but, I have not been discussing background blur.

***

I'll post the link again. Perhaps you could explain why the results would differ when the crop factor is accounted for and the other condition I mentioned is met in the final image. That effectively is crop factor based magnification as in some cases the framing can't be identical but feature in the shot can be.

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

The link posted is about Blur it seems to be a Blur Disk graph – it is NOT a Depth of Field Calculator.

Background Blur and DoF have a relationship, but I have no interest in accessing a Blur calculator to explain anything concerning the facts that I have stated about Depth of Field.

The homepage of the link provided, clearly indicates that “blur” and DoF, whilst being related are indeed DIFFERENT topics. Here my bold for emphasis:

“One of the most returning questions on photography forums is which out of two different lenses has the smallest depth of field (DOF). However, what many people actually mean when they ask this is which lens has the best ability to blur the background in their shots.”

I don’t mean and I have never stated that I mean “the best ability to blur the background” - I have only spoken about DoF – that’s not being pedantic that’s being both accurate and articulate.

WW

[William W]

. . . It says "the term perspective is applied to the apparent relationship between the position and size of objects when seen from a specific viewpoint, in a scene examined visually. "

Not relative to each other, but to the frame.

No George - you are missing the nuance of the meaning of the language.

The words "apparent relationship" means the relationship between the objects to each other inside the scene - exactly as I described previously.

Truly, I don't want to appear overpowering but English is my native tongue and I am trained and experienced in technical writing and journalism and I have produced a successful doctorial thesis – truly, that is what the words mean.

WW

I didn’t open the link you provided, because I could not open it – my computer hung when i tried.

I hope you enjoy the book.

[ajohnw]

Take no notice of that George your English is excellent and for some reason Bill is simply restating what you have already pointed out. It didn't need anything further.

Funny really as you native language might be English.

Anyway we now seem to be down to an effect that in real terms need not have any impact on the final image. The final image is what I assume interests most of us so I find that rather surprising.

John
-

[dem]

I think I've got it.

As we know, the CoC values used by most DoF calculators are 0.03 for FF and 0.02 for APS-C sensors.

To make the CoC work per pixel rather than for the resolution of 5 lines per mm in a 8"x10" print, you need to divide its typical value by a square root of (N Mpx / 4).

So for the D810 with a full frame 36 Mpx sensor, the adjustment factor is sqrt(36/4)=3 and the adjusted CoC is 0.03/3=0.01. If your DoF calculator doesn't allow changing CoC, just divide the DoF that it gives you by 3.

A 16 Mpx APS-C camera (same pixel density as the D810!) will have the adjustment factor of sqrt(16/4)=2 and the adjusted CoC value of 0.02/2=0.01.

Some cameras like Fuji X-Pro2 have a dual DoF indicator: the classical one for printing and the one for pixel peepers. The DoF bars are about 2.5 times shorter for the latter one. Let's check the adjustment factor for a 24 Mpx sensor: sqrt(24/4)=2.45.

This "adjustment factor" tells you by how much DoF shrinks if you are viewing the image on the computer screen at 100% magnification instead of printing 8"x10". The number 4 in the formula is 4 Mpx - the resolution that corresponds to classical definition of DoF. In other words, you are not supposed to be able to tell the difference between 4 Mpx and 24 Mpx 8"x10" printouts when viewed from a typical reading distance.

[pnodrog]

Dem I think you are on Dan's (NorthernFocus) wavelength. However the DOF calculation has never taken into account the resolution of the film/lens/sensor. It has always been assumed when viewing the 10"x8" print it was sufficient to not be a factor in our perception of image softness. Obviously in the days of film, a full plate camera would potentially offer far better print quality than a 6x6cm or 135mm format could but all were capable of producing a 10"x8" print of sufficient quality that the CoC was not influenced by print quality. Nothing has changed due to different/increasing sensor resolution.

A DOF based on 100% viewing maybe of limited interest to pixel peepers but as sensor resolution continues to improve a DOF based on 100% viewing will ultimately be completely irrelevant in producing an acceptable photograph to be viewed at normal magnifications.

[xpatUSA]

Funny really as your native language might be English.
John
-

As far as I know, George is a Nederlander . .

Having said that, his English is vastly better that my Dutch

[george013]

Dem I think you are on Dan's (NorthernFocus) wavelength. However the DOF calculation has never taken into account the resolution of the film/lens/sensor. It has always been assumed when viewing the 10"x8" print it was sufficient to not to be a factor in our perception of image softness. Obviously in the days of film, a full plate camera would potentially offer far better print quality than a 6x6cm or 135mm format could but all were capable of producing a 10"x8" print of sufficient quality that the CoC was not influenced by print quality. Nothing has changed due to different/increasing sensor resolution.

A DOF based on 100% viewing maybe of limited interest to pixel peepers but as sensor resolution continues to improve a DOF based on 100% viewing will ultimately be completely irrelevant in producing an acceptable photograph to be viewed at normal magnifications.

It's just simple that. The experience of sharpness on a certain distance and the back calculation to the size of the sensor. and then the calculation of the range that is within that result.

Viewing on a monitor is something else. This old 19" monitor has a resolution of 1280x1024 or 86 DPI. It's rather low. Viewing a picture of 4000x6000 will result in a metrical size of 177x118cm at a arm length distance and a resolution of 86DPI. About what sharpness are we talking?? In the "classical" approach of sharpness a magnification of around 8 was used for a FF. This example has a magnification of 49. That dot of 0.25mm that is the basic of the calculation is now 1.5mm For a 1.5 crop camera multiply with 1.5.
Viewing at screen size might be closer to viewing a standar print. But still at 86DPP for this monitor. The other one has a 108 DPI. Things are looking sharper on that one.

George

[Dave Humphries]

To my mind, DoF at 100% viewing on screen is irrelevant as a concept.

DoF is only relevant when viewing an entire composition.

If just part of an image is viewed at 100% (at the shooting resolution), DoF just doesn't work because. as a concept, it relies on there being a range of softness in fine detail which the human eye, at a typical viewing distance, assesses to be 'sharp'. If you're viewing a screen at 100%, that assessment process can't happen, all you will see is some areas sharper than others and a gradual transition between them.

An analogy might be trying to understand a book if the text can only be read through a hole in an overlaid card (mask) that allows you to see just one or two letters at a time - and you have to scan the hole along the lines of text.


Images for on line display should be downsized and then output sharpened for the resolution of display they are to be viewed on (or are most likely to be viewed on), this restores the detail sharpness lost when downsizing from say; 6000 x 4000 to 1920 x 1080 px.

This averaging during downsizing and sharpening then effectively 'overwrites' the capture DoF, which is very useful for macro shots, but not for say, portraits shot at f/2 or wider. I'm still considering this last aspect, it's not something I had considered before, it may affect how I process my wide aperture images in future.

[george013]

That makes it a downsizing question.

George

[William W]

. . . Images for on line display should be downsized and then output sharpened for the resolution of display they are to be viewed on (or are most likely to be viewed on), this restores the detail sharpness lost when downsizing from say; 6000 x 4000 to 1920 x 1080 px.

This averaging during downsizing and sharpening then effectively 'overwrites' the capture DoF, which is very useful for macro shots, but not for say, portraits shot at f/2 or less. I'm still considering this last aspect, it's not something I had considered before, it may affect how I process my wide aperture images.

Discerning Clients (and sharp eyed non clients too) do notice.

The vast swing over the past ten years to the great majority of photographic imagery being made for “on line” or “on screen” (i.e. at home) viewing has resulted in a substantial “wow” factor when a “big print” of excellent quality is produced. Especially if it is hung next to the “screen”.



This image looked very good on the Client’s u-beaut extra super-duper TV - they loved it: I had it hand printed on special paper to around 30” wide (about “life size”) and the print simply creams any screen display.

Apropos the impact of the shallow DoF in the print: - the print it looks at first glance that she is surging out of the frame and into the room that was not able to be achieved with any "on-screen" display of it.

WW

Image © AJ Group Pty Ltd Aust 1996~2016 WMW 1965~1996

[Dave Humphries]

This image looked very good on the Client’s u-beaut extra super-duper TV - they loved it: I had it hand printed on special paper to around 30” wide (about “life size”) and the print simply creams any screen display.

Apropos the impact of the shallow DoF in the print: - the print it looks at first glance that she is surging out of the frame and into the room that was not able to be achieved with any "on-screen" display of it.

I guess if future displays increase their dimensions and increase their ppi figures - to get much closer to the dpi of a good print, that may change, but we're probably a few years off these being cheap enough for them to become widespread. But it's easy enough to print bigger still and maintain a gap

[george013]

I guess if future displays increase their dimensions and increase their ppi figures - to get much closer to the dpi of a good print, that may change, but we're probably a few years off these being cheap enough for them to become widespread. But it's easy enough to print bigger still and maintain a gap

When I searched for a new monitor last year I focused also on the resolution in dpi. What I have now is 108 dpi and that works for me. Dell p2715q has a resolution of 163dpi. But what is the use? The smaller the pixel size, the less pixel peeping. It looks wonderful, that picture you're viewing. But you're the only one. So if you're shooting for the web, it's worthless to me.

George

[xpatUSA]

To my mind, DoF at 100% viewing on screen is irrelevant as a concept.

Dave, taken by itself, the statement is a little too sweeping, IMHO.

For example, I shoot at 2336x1568px raw and display captures on a 1920x1200px monitor at 100%, often cropping to 1920x1280px rather than downsizing, and my DOF calculator is set for my screen and my viewing distance.

Admittedly, I am probably unique here as regards the foregoing but I am usually viewing the whole image or major part thereof, as captured, at 100% on my screen.

Furthermore, 4K screens are here and 8K screens are coming. 8K is 7680x5120px which applies some serious meaning even to today's high MP cameras. Of course we could drift off into dot pitch and viewing distance, etc., but viewing at 100% is normal for me and will became more popular with others as time progresses, IMHO . .

Agree with everything else that you said - dare I say 100% ?

[NorthernFocus]

I guess if future displays increase their dimensions and increase their ppi figures...

I think the future is here, Dave. At least on this side of the pond. UHD TVs are rapidly pushing 1080p off the shelves and higher end laptops now have UHD screens. Just three or four years ago it was difficult to find a laptop with a 1080p screen. I just bought a new 5 inch smart phone with a 2560x1440 screen at less than half the cost of an iPhone. That's the same resolution as my 27in. monitor that I nearly took out a mortgage on four years ago. I also have a tablet of similar resolution that's a couple years old now(BTW very similar size to an 8x10 print). Personally I thought the high resolution screens in smaller sizes were a waste until I saw one. Images viewed on them are truly stunning (a much over used word).

I'm a convert. Ted may be the only one who agrees with me but IMO for viewing images prints can't compare with a high rez display. The only thing that comes close is a metal print with gallery style lighting.

[dem]

Dem I think you are on Dan's (NorthernFocus) wavelength. However the DOF calculation has never taken into account the resolution of the film/lens/sensor...

Exactly! That is what I have been saying in my previous posts in this thread and also said that it is unnecessary to redefine DOF in terms of sensor pixel size. I can see no practical benefits of doing this. But against my best judgement, I then went on and offered a simple solution of doing exactly that in post #87 for two reasons:

First, it can be done. There is no conceptual difference in defining DOF with respect to the resolution of a human eye or the resolution of a photographic sensor. In the case of classical definition of DOF, the size of the out-of-focus disc (that would have been a single point for anything located on the focus plane) on a 8"x10" print is compared against the typical resolving power of a human eye. We can bypass the printing altogether and simply compare the size of that out of focus disc on the actual sensor to the size of a single pixel. If the oof disc is larger than a single pixel, then some image softness can be detected when the image is viewed at 100% magnification, or 200%, it does not matter. The viewing distance and the dpi of the monitor are irrelevant, in this case, because the only thing we are trying to assess is whether the image of that point was defocused (not to say blurred) over more than one pixel or not.

Second, there is obviously some interest in what pixel density does to DOF (the correct answer is nothing). That is why this thread started and that is why Fuji introduced an alternative DOF indicator in their flagship cameras (I wish they didn't).

So my post #87 is not "the true theory of DOF" but an academic exercise that will hopefully convince people that we do not need a DOF definition based on sensor pixel density.

Example. Say you are shooting a portrait with a f/1.8 lens on D810 and the dofmaster calculator is telling you that for the chosen focal length and subject distance DOF is 15 cm. Now, the "pixel-based DOF" is going to be 15 cm / 3 = 5 cm as explained in post #87. This means that you are likely to spot some softening of the tip of the nose and the ears when you closely examine this image on the screen.

What is next?

Option 1. You want your photos to be pixel-perfect and from now on you only shoot portraits using aperture f/5.6 or smaller.

Option 2. You downsize the image for web or print it out and stop worrying about pixel density.

Ok, my last argument. What matters is what people see, be it on their computer screens or on prints, not what the sensor saw. There is no point in trying to please the sensor.

[xpatUSA]

Exactly! That is what I have been saying in my previous posts in this thread and also said that it is unnecessary to redefine DOF in terms of sensor pixel size. I can see no practical benefits of doing this. But against my best judgement, I then went on and offered a simple solution of doing exactly that in post #87 for two reasons:

First, it can be done. There is no conceptual difference in defining DOF with respect to the resolution of a human eye or the resolution of a photographic sensor. In the case of classical definition of DOF, the size of the out-of-focus disc (that would have been a single point for anything located on the focus plane) on a 8"x10" print is compared against the typical resolving power of a human eye. We can bypass the printing altogether and simply compare the size of that out of focus disc on the actual sensor to the size of a single pixel. If the oof disc is larger than a single pixel, then some image softness can be detected when the image is viewed at 100% magnification, or 200%, it does not matter. The viewing distance and the dpi of the monitor are irrelevant, in this case, because the only thing we are trying to assess is whether the image of that point was defocused (not to say blurred) over more than one pixel or not.

Second, there is obviously some interest in what pixel density does to DOF (the correct answer is nothing). That is why this thread started and that is why Fuji introduced an alternative DOF indicator in their flagship cameras (I wish they didn't).

So my post #87 is not "the true theory of DOF" but an academic exercise that will hopefully convince people that we do not need a DOF definition based on sensor pixel density.
<>
Ok, my last argument. What matters is what people see, be it on their computer screens or on prints, not what the sensor saw. There is no point in trying to please the sensor.

The well-respected Doug Kerr appears to agree:

http://dougkerr.net/Pumpkin/articles/Depth_of_Field.pdf

See page 14 and 15 under "THE CAMERA RESOLUTION OUTLOOK" . . .

[ajohnw]

The well-respected Doug Kerr appears to agree:

http://dougkerr.net/Pumpkin/articles/Depth_of_Field.pdf

See page 14 and 15 under "THE CAMERA RESOLUTION OUTLOOK" . . .

Thanks for that link Ted. I've been looking for what look to be the correct equations for some time. COCDL or what ever he calls it - they have to be the same irrespective of format ???? The correct amount of blur can do some interesting things to some photographs due to the way we see.

It would be rather difficult to make sense of actual pixel resolution of a camera sensor because of debayering which take 4 pixels that contain 3 colours and then use what might be called a kludge to resolve the image for any type of RGB display. Some people have assumed a 2 pixel pitch for resolution. Also I think 3 or more but I would have to get a book out to check. It might be 4. The larger one seems to work on microscopes but probably because by accident it accounts for the flaws these usually have.

There is a resolution limit via the lens and that for full frame and crop is circa 45 line pair per mm, if a great lens and used at it's best aperture. On a 50mm Milvus for instance that is at F4 at F1.4 it's about 32. Just as a comparison as it's interesting an Olymus 25mm F1.8's best it around 72 line pairs per mm at F4 and 65 at F1.8. Not so even across the field as the Zeiss as that drops 5 or so, 60 and 50 at the edge on Olympus. It's not the best lens resolution wise. I regularly get images that could be cropped considerably. I have on older full frame - yes it's very old same there but not really on more recent crop cameras. They will take some but no where near as much.

I've read that Calr Zeiss came up with the idea of using MTF50 resolution figures. Not sure if this is true or not, The idea is that beyond that point the contrast levels aren't suitable for photographs however really good quality lenses probably do give out useful detail past that point if the contrast is high enough.

John
-