Pages: 1 [2]   Go Down

Author Topic: DoF, sensor size, and pixel pitch  (Read 17766 times)

gkramer

  • Newbie
  • *
  • Offline Offline
  • Posts: 27
DoF, sensor size, and pixel pitch
« Reply #20 on: January 30, 2007, 10:52:16 pm »

Quote
...at a quick check, I see nothing in your previous posts stating that all comparisons are done for the case of 300PPI prints. Also, I do not... see a good basis for an (unstated?) assumption of a fixed 300PPI...[a href=\"index.php?act=findpost&pid=98384\"][{POST_SNAPBACK}][/a]
Do your homework.

It's possible that I screwed up the math, or somehow set the problem up in the wrong way; I which case I would be grateful to be shown the error of my ways. But as your "quick check" remark suggests, you haven't made any real effort to understand the argument(s), and frankly, I don't think you grasp the issues.
Quote
I use "enlargement" in the well established photographic meaning of the ratio of the size of the print (or projected image) to the size of the image recorded in the focal plane of the camera. So regardless of film or sensor resolution, prints of the same degree of enlargement from the same format are the same size.
[a href=\"index.php?act=findpost&pid=98384\"][{POST_SNAPBACK}][/a]
I think this remark proves my point: sensor size (in physical dimensions) is utterly irrelevant; as I pointed out in my first post, a 6MP, 2000 x 3000-pixel image file is an 6MP, 2000 x 3000-pixel image file, "irrespective of whether it was captured by a 24mm x 36mm full-frame sensor, a 16mm x 24mm APS-C sensor, or a tiny, quarter-inch sensor from a pocket point-n-shoot." Stripped of its EXIF data and loaded into PhotoShop for editing, resizing, and printing, neither Photoshop nor the printer would have any idea of whether it came from a full-frame DSPR, or a tiny pocket point-n-shoot, or whatever; nor would it matter. Any sensible discussion of "enlargement" (ie, how big a print we want to make from the file) in the digital world must take account of that reality. Obviously you haven't grasped the point.
« Last Edit: January 31, 2007, 09:30:29 am by gkramer »
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #21 on: January 31, 2007, 10:46:25 am »

Quote
I think Bob Atkins would have made this statement in the context that both the 10D and 35mm film have approximately the same resolution. He really shouldn't have used the word identical. He probably meant identical with respect to most practical purposes, taking an average quality of film, and ignoring pixel-peeping concerns.

[a href=\"index.php?act=findpost&pid=98387\"][{POST_SNAPBACK}][/a]

Well, he did use the word identical.

You seem to dismiss Mr. Atkins statement with he didn't mean it, or should have said something different.

You did not address: "There was a query in October, 2001 on my Discussion Forum as to whether Depth of Field was calculated any differently for digital Vs. film. The answer is, no. There is no difference whosesoever. DOF doesn't care about the recording media type or size, ... "

Is this author also mistaken or over speaking?  What did he really mean to say?

I didn't know there were homework assignments for this forum.  Maybe you should check your own work, rather than demand others do it for you.  You seem to be alone in your thinking, but that doesn't mean you are necessarily wrong.  But for such a simple and well travelled topic, probably.
« Last Edit: January 31, 2007, 11:09:52 am by howiesmith »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
DoF, sensor size, and pixel pitch
« Reply #22 on: January 31, 2007, 03:46:49 pm »

Quote
you could equally plausibly (or implausibly) attack the C0C = 30 micron standard for 35mm photography--from which I presume you would conclude that manufacturers shouldn't engrave DoF scales on their lenses.
[a href=\"index.php?act=findpost&pid=98415\"][{POST_SNAPBACK}][/a]

On the contrary, the COC=30 micron standard for 35m format supports my point: it depends on the format size and not on film or sensor resolution.

The DOF scales on lenses are based on making prints of a certain reference size (5"x7"?), and viewing them from a certain standard distance (10"?) and calling something in-focus if the resulting CoC on the print in no larger than about 1/1500 of the viewing distance, on the basis that this is about the limits of what the human eye can resolve. The print CoC threshold is thus about 150 microns, and for 35mm format, the degree of enlargment needed to get a 5"x7" print is about 5x, so the CoC for DOF scales is about 1/5th of 150 microns, or 30 microns.

Note:
1) These DOF scales do depend of format size, since this determines the degree of enlargement needed to get the standard sized print. The DOF scale on medium format cameras are typically computed using a value larger than 30 microns: something like 1/1000th of the format diagonal length is common.
2) These DOF scale do not depend on the resolution of the film or sensor used.
3) These DOF scales are based on comparing prints of the same size regardless of film or sensor resolution whereas your 300PPI comparison involves comparing prints of different sizes from cameras with different pixel counts.
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #23 on: January 31, 2007, 04:02:23 pm »

Quote
On the contrary, the COC=30 micron standard for 35m format supports my point: it depends on the format size and not on film or sensor resolution.

The DOF scales on lenses are based on making prints of a certain reference size (5"x7"?), and viewing them from a certain standard distance (10"?) and calling something in-focus if the resulting CoC on the print in no larger than about 1/1500 of the viewing distance, on the basis that this is about the limits of what the human eye can resolve. The print CoC threshold is thus about 150 microns, and for 35mm format, the degree of enlargment needed to get a 5"x7" print is about 5x, so the CoC for DOF scales is about 1/5th of 150 microns, or 30 microns.

Note:
1) These DOF scales do depend of format size, since this determines the degree of enlargement needed to get the standard sized print. The DOF scale on medium format cameras are typically computed using a value larger than 30 microns: something like 1/1000th of the format diagonal length is common.
2) These DOF scale do not depend on the resolution of the film or sensor used.
3) These DOF scales are based on comparing prints of the same size regardless of film or sensor resolution whereas your 300PPI comparison involves comparing prints of different sizes from cameras with different pixel counts.
[a href=\"index.php?act=findpost&pid=98587\"][{POST_SNAPBACK}][/a]

1)  The standards are format dependant.  DoF is not.  One does not have to adopt any standard one does not want to.  I choose not to adopt a standard, but to allow myself to make any size print from any crop of any format.  It is easy math (add, subtract, multiply and divide).  Just use the same equations used to calculate the standard and use your own design.

2)  I agree with this one.

3)  Again, the scales do depend on comparing the same size prints under the same viewing conditions.  The factors used in those standards do not have to the same a photographer actually uses to design a print.  He can use his own print.  If I want a certain DoF of an 9x 11 print under certain conditions, from a cropped 35mm negative, I can do that.  If I want a "standard" print, I can make that too.  I don't need any standards, DoF scales, DoF tables or on-line calculators.  I don't even have to know the standards if I don't misuse and tool based on them.
« Last Edit: January 31, 2007, 04:43:07 pm by howiesmith »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
DoF, sensor size, and pixel pitch
« Reply #24 on: January 31, 2007, 05:39:31 pm »

Quote
1)  The standards are format dependant.  DoF is not.  One does not have to adopt any standard one does not want to. ...

2)  I agree with this one.

3)  Again, the scales do depend on comparing the same size prints under the same viewing conditions. ...
[a href=\"index.php?act=findpost&pid=98589\"][{POST_SNAPBACK}][/a]
Indeed! I think that Michael made the basic point in an essay on DOF somewhere on this site. To paraphrase

1) DOF scales on lenses, DOF tables and most DOF calculators give an "indication" of DOF under certain stated "reference" conditions, something like printing the entire image recorded by the camera and viewing the print from a distance a bit greater than the diagonal length of that image; what I will call "reference viewing distance".

2) From this guidance, a photographer planning for different viewing conditions can make adjustments. For example, if one plans on making a 16"x20" print that is likely to be viewed from as close as 15", the DOF will be only about half what the DOF scales indicate, and one would need to stop down to twice the f-stop suggested by such scales. Likewise cropping by 2x (using only half the height and width of the full image) and then viewing from "reference viewing distance" would give about half of the indicated DOF.

After all, it would be messy to mark lenses with DOF scales that cover all the possible viewing conditions for prints!
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #25 on: January 31, 2007, 06:24:33 pm »

Quote
Indeed! I think that Michael made the basic point in an essay on DOF somewhere on this site. To paraphrase

1) DOF scales on lenses, DOF tables and most DOF calculators give an "indication" of DOF under certain stated "reference" conditions, something like printing the entire image recorded by the camera and viewing the print from a distance a bit greater than the diagonal length of that image; what I will call "reference viewing distance".

2) From this guidance, a photographer planning for different viewing conditions can make adjustments. For example, if one plans on making a 16"x20" print that is likely to be viewed from as close as 15", the DOF will be only about half what the DOF scales indicate, and one would need to stop down to twice the f-stop suggested by such scales. Likewise cropping by 2x (using only half the height and width of the full image) and then viewing from "reference viewing distance" would give about half of the indicated DOF.

After all, it would be messy to mark lenses with DOF scales that cover all the possible viewing conditions for prints!
[a href=\"index.php?act=findpost&pid=98610\"][{POST_SNAPBACK}][/a]

I'm sorry.  I thought you were once argueing DoF was format dependant.  I must have been mistaking you for someone else, like Ray.

Yes, the photographer can use the standards as guidance.  But why bother to use the standard and adjust it, when all you need do is smiply calculate DoF for your own use?  Or the photographer can dumb down the whole thing and use tables, calcuators or engravings just like they were correct for his 20x24.

Yes, it would be messy to put all possible DoFs on a lens barrel.  So don't put any, which seems to be the current trend.  I say better none than misusing one.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
DoF, sensor size, and pixel pitch
« Reply #26 on: January 31, 2007, 06:26:20 pm »

Quote
Is this author also mistaken or over speaking?  What did he really mean to say?

I didn't know there were homework assignments for this forum.  Maybe you should check your own work, rather than demand others do it for you.  You seem to be alone in your thinking, but that doesn't mean you are necessarily wrong.  But for such a simple and well travelled topic, probably.
[a href=\"index.php?act=findpost&pid=98520\"][{POST_SNAPBACK}][/a]
 

I think Bob Atkins might be at least partially wrong on this issue, but I understand that anyone who has not made an attempt to work out this issue for himself, would probably prefer to take Bob Atkins word for it rather than mine.

That doesn't worry me because as a photographer, ultimately, it's the practical significance of such theories that matter. If my views differ from those of more recognised authorities, it's because I've done my homework in the form of practical experiments. If I'm not sure, for example, if the sharpness and detail of an image will be compromised by using f16 or f22 with a particular lens and camera, I'll do tests on real world images, as well as line charts. If I find, for example, that there's no significance resolution difference between f8 and f16 with one lens, and no significant difference between f22 and f11 with another lens, then no amount of theoretical pronouncements from recognised authorities, whether they have a PhD or not, will dissuade me from my view. All I could say in their defense is, they must have been referring to a different set of conditions.

It's quite clear to me that system resolution affects DoF outcomes at a practical level. I carried out the following experiment recently after purchasing a Canon 50/1.4 lens. I took a series of shots at both f8 and f1.4, focussing at various distances for each pair of shots. On examining the images, I discovered that the OoF parts of some of the f8 shots were just as sharp as the in-focus parts of the f1.4 shots. The difference in sharpness in these f8 shots, between OoF parts and the in-focus parts was sufficient to create the impression of a slightly shallow DoF.

Now it's clear to me that, whilst the F1.4 shots had significantly shallower DoF than the f8 shots, the DoF would not have been as shallow as the DoF calculators and formulas would imply because such formulas do not take into account lens quality.
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #27 on: January 31, 2007, 10:19:17 pm »

Quote
I think Bob Atkins might be at least partially wrong on this issue, but I understand that anyone who has not made an attempt to work out this issue for himself, would probably prefer to take Bob Atkins word for it rather than mine.

That doesn't worry me because as a photographer, ultimately, it's the practical significance of such theories that matter. If my views differ from those of more recognised authorities, it's because I've done my homework in the form of practical experiments. If I'm not sure, for example, if the sharpness and detail of an image will be compromised by using f16 or f22 with a particular lens and camera, I'll do tests on real world images, as well as line charts. If I find, for example, that there's no significance resolution difference between f8 and f16 with one lens, and no significant difference between f22 and f11 with another lens, then no amount of theoretical pronouncements from recognised authorities, whether they have a PhD or not, will dissuade me from my view. All I could say in their defense is, they must have been referring to a different set of conditions.

It's quite clear to me that system resolution affects DoF outcomes at a practical level. I carried out the following experiment recently after purchasing a Canon 50/1.4 lens. I took a series of shots at both f8 and f1.4, focussing at various distances for each pair of shots. On examining the images, I discovered that the OoF parts of some of the f8 shots were just as sharp as the in-focus parts of the f1.4 shots. The difference in sharpness in these f8 shots, between OoF parts and the in-focus parts was sufficient to create the impression of a slightly shallow DoF.

Now it's clear to me that, whilst the F1.4 shots had significantly shallower DoF than the f8 shots, the DoF would not have been as shallow as the DoF calculators and formulas would imply because such formulas do not take into account lens quality.
[a href=\"index.php?act=findpost&pid=98619\"][{POST_SNAPBACK}][/a]

So Ray, just where do you stand on DoF now.  I am having trouble figuring out your position.

Is it that no theory is any good and only your "tests" will tell?  Do you record test results for field use, or do you still go by the "if I think I need more , I stop down one more"?  I am impressed if you can remember that such and such a lens at f/whatever and focused at x feet gives a DoF on an A3 print viewed at y inches of whatever feet.  Actually, I really doubt both.  You seem to be constantly changing positions to keep your nose pointed upstream.

You seem to put a lot of faith in Albert Einstein, but relativity was just his theory.  He was unable to put it to much of a test (except e=mc^2 peraps).  Al just wasn't able to accelerate mass to nearly the speed of light to check some things.

Do you have ideas or are you simply contrary?
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
DoF, sensor size, and pixel pitch
« Reply #28 on: January 31, 2007, 10:39:59 pm »

Quote
So Ray, just where do you stand on DoF now.  I am having trouble figuring out your position.
[a href=\"index.php?act=findpost&pid=98653\"][{POST_SNAPBACK}][/a]


Pretty much the same as it has been for the last few years, Howie. If a print looks sharp all over, it has great DoF. The fuzzier the parts which are not in focus are, the shallower the DoF. I don't take slide rules, DoF calculators, tape measures or laser distance finders with me when I go out shooting. I rely upon experienced-based judgement for selection of aperture, shutter speed and point of focus. Sometimes I get it wrong.
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #29 on: February 01, 2007, 01:32:41 pm »

Quote
Pretty much the same as it has been for the last few years, Howie. If a print looks sharp all over, it has great DoF. The fuzzier the parts which are not in focus are, the shallower the DoF. I don't take slide rules, DoF calculators, tape measures or laser distance finders with me when I go out shooting. I rely upon experienced-based judgement for selection of aperture, shutter speed and point of focus. Sometimes I get it wrong.
[a href=\"index.php?act=findpost&pid=98656\"][{POST_SNAPBACK}][/a]

I agree DoF does depend to some degree on overall resolution.  Fuzzier prints make it harder to see where the edges of DoF are.  Afterall, DoF is not binary - in-focus or not in-focus.  CoC gradually gets larger as the DoF limit is reached.  A viewer does not see an abrupt change in focus.

And I also agree that judgement and experience with particular equipment plays a role.

Would you agree that simple DoF theory (limited to four factors - focal length, f/stop, focus distance and degree of enlargement) is at least a good place to start looking for the boundaries of DoF (near and far in-focus limits)?
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
DoF, sensor size, and pixel pitch
« Reply #30 on: February 01, 2007, 05:44:11 pm »

Quote
Would you agree that simple DoF theory (limited to four factors - focal length, f/stop, focus distance and degree of enlargement) is at least a good place to start looking for the boundaries of DoF (near and far in-focus limits)?
[a href=\"index.php?act=findpost&pid=98727\"][{POST_SNAPBACK}][/a]

Yes, I would, which is why I wouldn't object too much to Bob Atkins' use of the word 'identical' in comparing the methods of calculating DoF using a 10D and 35mm film. The formula is the same, the lenses are the same, the CoC chosen might be a bit smaller for the 10D, but over all system resolution is similar.

However, I think it would be true to say that the greater the disparity between the resolving power of any two systems compared, the greater the inaccuracy of any formula will be that excludes resolution considerations, whether it be the resolving power of the lenses used, or the resolving power of the film or sensor.
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #31 on: February 01, 2007, 07:08:18 pm »

Quote
Yes, I would, which is why I wouldn't object too much to Bob Atkins' use of the word 'identical' in comparing the methods of calculating DoF using a 10D and 35mm film. The formula is the same, the lenses are the same, the CoC chosen might be a bit smaller for the 10D, but over all system resolution is similar.

However, I think it would be true to say that the greater the disparity between the resolving power of any two systems compared, the greater the inaccuracy of any formula will be that excludes resolution considerations, whether it be the resolving power of the lenses used, or the resolving power of the film or sensor.
[a href=\"index.php?act=findpost&pid=98776\"][{POST_SNAPBACK}][/a]

OK.  I think Bob Atkins is more right than wrong, because system resolution changes are very small compared to the calculated DoF.
The overall resolution O of a camera ans sensor is:

1/O = 1/S + 1/L  

where S and L are the resolution of the sensor and lens respectively.  But when you make a print, which is usually the ultimate goal, the resolution of the print P must be added.  Then:

1/O = 1/S + 1/L + 1/P

So no matter how good the lens and/or sesnsor get, the overall resolution O will be no better than the print P.  And print resolution isn't all that good.
The resolution of the eye (the link between print and brain) should also be added on.  We have already taken it into account with CoC on the print.

If you are making comparisons of two files on your monitor, don't forget they can be mo larger than the final print, and viewed no closer.  And the monitor will likely be brighter (and easier to see) than a print.

From a Reichmann tutorial on DoF and focal length:

"There are those that will no doubt find fault with either my fundamental assertion or with this test. Yes, I know that there are some flaws with both, among them that I am not taking into account diffraction and other second order effects. But, from a practical point of view, what really counts to photographers working in the real world is what ends up on a print in front of them. The fine points of optical theory are one thing, prints hanging on the wall are another. My orientation as a photographer and as a teacher is toward the pragmatic." [Emphasis added]

It appears the tutorial agrees that lens and sensor resolution effects may be second order.  I have found that to be true when actually making prints instead of computerized "tests."
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
DoF, sensor size, and pixel pitch
« Reply #32 on: February 01, 2007, 09:01:03 pm »

Quote
If you are making comparisons of two files on your monitor, don't forget they can be mo larger than the final print, and viewed no closer.  And the monitor will likely be brighter (and easier to see) than a print.

[a href=\"index.php?act=findpost&pid=98786\"][{POST_SNAPBACK}][/a]

I don't forget this and often click the 'print size' button at the top of the Photoshop page. All detail, including noise in the shadows, tends to be revealed in my prints, although it has to be said that such noise (as well as shadow detail) is more apparent when the print is viewed in good lighting conditions.

The problem with DoF on the print is that the perception of it can change, not only with the viewing distance but also with the size of background objects in the composition. Consequently, if we take two shots with different lenses, keeping the main subject the same size on the sensor or film, the objects in the background in the shot taken with the wider angle lens, will appear to be further away. The background will be more extensive. The same background objects common to both prints will be smaller in the shot taken with the wider angle lens, and as a consequence they will appear sharper.

In other words, if we accept that the DoF formulas for prints only apply accurately when consistent with viewing distances of, say the diagonal of the print, whatever the size of the print, then it follows that a wide angle lens that reduces the magnification of the background will have a similar effect to viewing the print from another 'less wide angle' lens from a greater distance (main subject occupying the same space on the sensor in both cases). I wish I could think of a less convoluted way of expressing that   .

I'll try again. The assertion that all lenses produce the same DoF at the same f stop, provided the main subject is the same size on the sensor is not correct when viewing same size prints from the same distance.
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #33 on: February 01, 2007, 10:01:15 pm »

Quote
I don't forget this and often click the 'print size' button at the top of the Photoshop page. All detail, including noise in the shadows, tends to be revealed in my prints, although it has to be said that such noise (as well as shadow detail) is more apparent when the print is viewed in good lighting conditions.

The problem with DoF on the print is that the perception of it can change, not only with the viewing distance but also with the size of background objects in the composition. Consequently, if we take two shots with different lenses, keeping the main subject the same size on the sensor or film, the objects in the background in the shot taken with the wider angle lens, will appear to be further away. The background will be more extensive. The same background objects common to both prints will be smaller in the shot taken with the wider angle lens, and as a consequence they will appear sharper.

In other words, if we accept that the DoF formulas for prints only apply accurately when consistent with viewing distances of, say the diagonal of the print, whatever the size of the print, then it follows that a wide angle lens that reduces the magnification of the background will have a similar effect to viewing the print from another 'less wide angle' lens from a greater distance (main subject occupying the same space on the sensor in both cases). I wish I could think of a less convoluted way of expressing that   .

I'll try again. The assertion that all lenses produce the same DoF at the same f stop, provided the main subject is the same size on the sensor is not correct when viewing same size prints from the same distance.
[a href=\"index.php?act=findpost&pid=98800\"][{POST_SNAPBACK}][/a]

Maybe you need to take this up with Michael.

I made some assumptions and did some calcualtions.  With a focal fength of 50, a focus distance of 100, f/stop of 8 and a CoC of 0.1, I get DoF of about 7 feet, 4 feet in front and 3 feet behind the subject.  If I double the focal length to 100 and double the focus distance to 200 (I think the focused image will be the same size), I still get about 7 feet of DoF, 4 in front and 3 behind.  Just as expected.

Why should I care that an object not in focus is a different size on my two negatives?  The CoC is a point 4 feet in front of (or 3 feet behind) the critically focused subject that appears as a fuzzy disc 0.1 in diameter on both negatives.  Both negatives will be enlarged the same so all fuzzy discs (regardless of size, i.e., distance from the critically focused subject) will be enlarged exactly the same to make two equal sized prints.  Two images taken with two different focal length lenses at two different focus distances are just that - two different images.  The DoF is the same but the images are different.  I think the difference is called perspective - nothing to do with DoF.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
DoF, sensor size, and pixel pitch
« Reply #34 on: February 02, 2007, 12:42:57 am »

Quote
Why should I care that an object not in focus is a different size on my two negatives? 

You might as well ask, 'Why should I care if an object that is in focus is the same size on the negative or sensor?'

You could also ask, 'Why should I care if DoF appears to increase as I step back from the print?'

As I recall, Howie, you have maintained a position that choice of CoC is dependent on the viewer's perspective of the final print. If you intend making a postcard size print, you can afford to use a larger CoC. If you intend making a large print that can only be viewed from a significant distance (because perhaps there are obstacles preventing the viewer moving closer), then you can also afford to use a larger CoC. If you intend making a large print that is likely to be viewed close up, then you have to either choose a CoC which is much smaller or take the risk of people like me making comments like, 'Such a pity the foreground is out of focus'.  

Quote
Two images taken with two different focal length lenses at two different focus distances are just that - two different images.  The DoF is the same but the images are different.  I think the difference is called perspective - nothing to do with DoF.

Well, you can't have it both ways, Howie. If you agree that the perception of DoF on the print changes according to the viewer's perspective (distance from the print), then it seems rather illogical to maintain that changing the perspective of the camera does not change DoF.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
DoF, sensor size, and pixel pitch
« Reply #35 on: February 02, 2007, 02:29:19 am »

Quote
Well, you can't have it both ways, Howie. If you agree that the perception of DoF on the print changes according to the viewer's perspective (distance from the print), then it seems rather illogical to maintain that changing the perspective of the camera does not change DoF.
[a href=\"index.php?act=findpost&pid=98816\"][{POST_SNAPBACK}][/a]

Further to this interesting dilemma, there is a solution proposed by Charles Sydney Johnson in his recent article, Lens Equivalents. I quote the relevant passage below.

Quote
However, there is a caveat. The type of DoF computation depends on the way photographs will be viewed! In principle a photograph should always be viewed from its proper perspective point. That is to say, the angle subtended by the photograph at the eye should be the same as the field of view of the lens used to take the photograph. Therefore, photographs taken with a wide angle lens should be held close to the eye so as to fill much of the field of view while telephoto photographs should be held farther away. It is sometimes forgotten that perspective in a photograph depends only on the position of the lens relative to the subject (object).

So how will the photographs be viewed? In fact, prints are usually viewed from 10” to 12” regardless of the focal length of lens used to make the photograph. When mounted prints are viewed, observers typically stand about the same distance from all prints. Observers generally don’t know what focal length lens was used, and they simply react to the apparent distortions present when a wide angle photograph is viewed from a distance greater than the perspective point. Similarly there is an apparent flattening effect when telephoto photographs are viewed too close to the eye. Also, in photographic shows the audience remains seated at the same distance from all prints and from the projection screen. Under usual viewing conditions, it is appropriate to compute the DoF with a constant CoC in the image regardless of the focal length of the lens.

The first part of this quote, as I understand it, is basically saying if you use a wide angle lens from a closer distance, then, even though the resulting prints are the same size and the subject is the same size as in another print of the same subject taken with a longer focal length lens from a greater distance, the shot taken with the wider angle lens should be viewed from a closer distance in order to maintain the perception of equal DoF that is implied by the basic DoF formulas.

Let's flesh this out a bit with a concrete example. Let's take your example of 2 shots taken with a 50mm lens and a 100mm lens, the shot with the 50mm lens being taken from half the distance to the subject so that the subject is the same size on the sensor. Let's also assume that there is some significant background detail common to both shots, say a rather OoF house some distance behind the subject.

The fundamental DoF formulas are basically saying, in both shots the actual resolution of the house is the same. If you were to enlarge the house in both images on your monitor, so both houses were the same size, you would see the same amount of detail in both houses. I know because I've tried it. (In case anyone is confused, we're using the same f stop with both lenses and have focussed on the same subject in front of the house.)

If we make equal size prints of both scenes and view both prints from the same distance, say the diagonal of the print so we can appreciate the fine detail of the subject, say Howard, then we will find that the house in the 50mm shot appears sharper. It appears sharper because it is smaller on the print. The DoF therefore appears greater in the 50mm shot.

If we move closer to the print of the 50mm shot, closer than its diagonal, the subject, Howard, will appear to be less sharp and the more distant house will appear to be more sharp. At some appropriately different viewing distance, both prints will appear to have the same DoF and we shall all be able to sleep soundly knowing that our mathematically based DoF formulas are accurate.
« Last Edit: February 02, 2007, 02:36:17 am by Ray »
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #36 on: February 02, 2007, 09:28:29 am »

Quote
You might as well ask, 'Why should I care if an object that is in focus is the same size on the negative or sensor?'

You could also ask, 'Why should I care if DoF appears to increase as I step back from the print?'

As I recall, Howie, you have maintained a position that choice of CoC is dependent on the viewer's perspective of the final print. If you intend making a postcard size print, you can afford to use a larger CoC. If you intend making a large print that can only be viewed from a significant distance (because perhaps there are obstacles preventing the viewer moving closer), then you can also afford to use a larger CoC. If you intend making a large print that is likely to be viewed close up, then you have to either choose a CoC which is much smaller or take the risk of people like me making comments like, 'Such a pity the foreground is out of focus'.   
Well, you can't have it both ways, Howie. If you agree that the perception of DoF on the print changes according to the viewer's perspective (distance from the print), then it seems rather illogical to maintain that changing the perspective of the camera does not change DoF.
[a href=\"index.php?act=findpost&pid=98816\"][{POST_SNAPBACK}][/a]

'Why should I care if an object that is in focus is the same size on the negative or sensor?"

They are the same size simply because I designed them that way when choosing focal length and focus distance.

"Why should I care if DoF appears to increase as I step back from the print?"

Simply because the eye cannot see the CoC that did appear at the limit any longer and a larger CoC becomes the limit, increasing DoF.  DoF does not "appear" to increase, it does.

DoF does change as viewing distance changes.  Look at the assumption made in selecting CoC concerning the size disc that can be seen - viewing distance dependent.

All I can say about stuff on the negative is, once the image is made, the negative is fixed.  Changing camera position does change DoF deign, but that can be be accounted for in the deign process by changing the focal length lens.
Logged

howiesmith

  • Guest
DoF, sensor size, and pixel pitch
« Reply #37 on: February 02, 2007, 09:48:57 am »

Quote
Further to this interesting dilemma, there is a solution proposed by Charles Sydney Johnson in his recent article, Lens Equivalents. I quote the relevant passage below.
[a href=\"index.php?act=findpost&pid=98822\"][{POST_SNAPBACK}][/a]

Charles Johnson says prints should be viewed from the proper perspective distance.  We know it is not true that they are.  Otherwise all pages of a book would have to have images with the same perpsective.  They usually don't.

Images can be viewed in a book correctly if the images are designed correctly to be the proper size and viewed from a proper distance.

Perspective is only the focus distance and focal length - thinks that affect the size relationship of things on the negative.  These are both factors in DoF but not the only factors.  DoF on the print is based also on f/stop and CoC (enlargement of the negative).  CoC also depends on the viewer's conditions (distance being the most common).  

All I can offer now is if it works for you, Ray, do it.  I know what works for me, and I will continue doing it.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
DoF, sensor size, and pixel pitch
« Reply #38 on: February 02, 2007, 08:06:25 pm »

Quote
'Why should I care if an object that is in focus is the same size on the negative or sensor?"

They are the same size simply because I designed them that way when choosing focal length and focus distance.

So let's get this straight, Howie. Having designed that the main subject, in focus, is a certain size on your negative, you don't really give a stuff about the size of the objects in your image which are 'not in focus'. Is that right? Out of focus, out of mind, whether it's a little out-of-focus or a lot out-of-focus. Is that your position?
Logged
Pages: 1 [2]   Go Up