Sensor size and DoF

[gkramer]

"For example, let's say I'm moving down from large format and am not at all familiar with miniature cameras. Someone plays a trick on me and hands me a Nikon D2X renamed as a Canon 5D with some Nikkor lenses also renamed as Canon lenses of the same focal length. I know the camera is a 12mp DSLR. Does it not make any difference what the sensor size is when I choose f stops for a particular DoF effect?"

Yes, according to Wrotniak's M x A rule (http://www.wrotniak.net/photo/tech/dof.html), the DoF from an APC-sized sensor (with a magnification factor of M=1.6) at a given aperture A will be the same as that from a full-frame sensor at an aperture of M x A; in other words the full-frame Canon would have to use an aperture of f/12.8 to yield the same DoF as the D2X at f/8. This calculation takes account of the fact that the smaller sensor (or negative) will have to be enlarged 1.6 times more to yield the same final print size, and assumes that the images were taken with lenses of equivalent FoV (e.g., 100mmm for the D2X, 160mm for the full-frame Canon), and that the CoC (at the sensor) on which the DoF calculations are based is large relative to the pixel pitch (or film reslution), and compared to the Airey disk produced by lens diffraction (i.e. that we're shooting at a fairly large, non-diffraction limited aperture (f/8 is about the limit for the D2X, with its small pixel pitch).

[howiesmith]

Does it not make any difference what the sensor size is when I choose f stops for a particular DoF effect?
[a href=\"index.php?act=findpost&pid=80070\"][{POST_SNAPBACK}][/a]

It makes no difference.

Suppose there is a person in the view finder showing head to foot.  I want a waist up print, so I will need to crop.  Now does it matter whether I throw away a half inch of negative or 2 inches of negative? (I use the word negative to denote either a piece of film or a computer file.)  I think not.  It only matters how much I am going to magnify (enlarge) the person to the final print.  If the 1" tall person on the negative is made to be 20" from waist up on the print, I think the enlargement will be about 40X, regardless of original format.

You can aslo run through the optical equations knowing only the peron is 6' tall and is Y feet away, I am using an Xmm lens and I want half of that person to be 20" tall on the print.  I won't bore you here with those details.

You say "Who would crop that much?  Change lenses or move up."  Then you have changed focal lenght of focus distance.  Anothr problem, but format is still eliminated the same way.

As an aside, in another part of this site there are some images for critique.  On at least one, there are suggestions to crop off about half the submitted image.  If the same size print is now made from half the original file, doesn't that change the DoF assumptions?  And then we still don't know (or care) what the origianal format was.  It may already be severely cropped or shown full frame or something in between.

There is a very lovely image submitted there by pom (some fall trees).  If he now decides to make a 48" high print of that same file, will the DoF be the same or different than that in the submitted image?  I know I view that imake on my computer monitor at abpout 18".  The computer image is back lit.  Does a print (front lit) 48" high hanging on my hallway wall in dim light have the same DoF when viewed from a maximum of 30" away?  I haven't run the numbers (or better yet, viewed the print) but I seriously doubt it.

[Ray]

How can you expect to design prints if you don't understand your camera and lenses and how to use them?

I don't expect to be able to do that, but the issue we're discussing at the moment is the relevance of format to DoF calcualtions. If format is truly unattached to DoF considerations, then according to my simple understanding of logical processes, I don't need to know the format of my camera. I certainly don't need to know the format for compositions purposes using an SLR, because what I see through the viewfinder (or the EVF of a P&S) is what I get. If I have lots of experience using small cameras of various formats, I can make an educated guess as to the format of the camera. If I put a 50mm lens on a D2X with a 1.5x crop factor, which had been renamed as a FF Canon 5D, I might quickly sense that something was wrong and I'd probably be quite puzzled. (On the other hand, I probably wouldn't complain too much because all my prints would have a greater DoF than expected   ).

I use 4x5 and 6x6 film and can measure image sizes directly from the ground glass.  No need to know the pixel size or pitch because there aren't any.  If there were, I don't need to know thatanyway.  (I do need to know not to try to make CoC less than the pixel size though.)  Then it is a simple matter of saying that the 1 inch high tree on the ground glass will be 20 inches on the planned print.  A 20X enlargement.  Not hard to do and not determined by format.  Note that if I am using my 4x5 and think I am using my 6x6, I still get the same results - a 20X enlargement.

And would that 1" high tree on the ground glass be independent of format? What I'm trying to get from you, Howie, are the processes you would use to make a print of a particular size with a specified or designed DoF without having any knowledge of the format of the camera you used.

No need to know the pixel size or pitch because there aren't any.  If there were, I don't need to know thatanyway.  (I do need to know not to try to make CoC less than the pixel size though.)

Howie, in order to avoid making the CoC less than the pixel size, you need to know the pixel size or more precisely, pixel pitch.

[howiesmith]

I don't expect to be able to do that, but the issue we're discussing at the moment is the relevance of format to DoF calcualtions. If format is truly unattached to DoF considerations, then according to my simple understanding of logical processes, I don't need to know the format of my camera. I certainly don't need to know the format for compositions purposes using an SLR, because what I see through the viewfinder (or the EVF of a P&S) is what I get. If I have lots of experience using small cameras of various formats, I can make an educated guess as to the format of the camera. If I put a 50mm lens on a D2X with a 1.5x crop factor, which had been renamed as a FF Canon 5D, I might quickly sense that something was wrong and I'd probably be quite puzzled. (On the other hand, I probably wouldn't complain too much because all my prints would have a greater DoF than expected   ).
And would that 1" high tree on the ground glass be independent of format? What I'm trying to get from you, Howie, are the processes you would use to make a print of a particular size with a specified or designed DoF without having any knowledge of the format of the camera you used.
Howie, in order to avoid making the CoC less than the pixel size, you need to know the pixel size or more precisely, pixel pitch.
[a href=\"index.php?act=findpost&pid=80156\"][{POST_SNAPBACK}][/a]

Ray, you don't need to know the format of the camera.  You only need to be able to measure the image size in the viewer and relate that to image size in the camera.  That is format independant.

The 1" high tree on the round glass is format independant, but does depend on the lens focal length and the distance from the tree to the camera.  It will be 1" high with my 6x6 with a 150mm lens and 1" high with my 4x5 and a 150mm lens.  But once on the ground glass, I don't care anyway.  The tree is an inch high on the ground glass and on the film, for my cameras.  It would only matter to me when selecting a lens and/or a camera-to-tree distance.

That's right Ray, I never use the camera's format or film size for designing prints.  Just the degree of enleagement I want from a focused image.  I could crop each negative with a pair scissors, get the same print without ever knowing how much negative I cut off and threw away.

If you use digital and happen to use a CoC smaller than the pixel pitch, the results will simplt not be as expected.  The same can be said of film and a CoC that cannot be recorded on the film.  It's a problem that may surface.  And frankly, since I don't use digital, I don't worry about it or a solution.  If I get a digital camera, I'll try to figure it out then.

Because I don't own a pixel, I need not know anything about its size.  You may if you try to make CoC too small for your camera to resolve.  I have not yet run into a case where my CoC was smaller than the lens/film resolution.  In fact, I don't even know what that is for my lenses and film selections.

Finally, greater DoF can be a problem if the print design calls for less.  Huge DoF isn't always the goal.  I am speaking of controlling DoF.  If all you want is maximum DoF. just use f/64 (of the biggest number you got) and tiny CoC (maybe pixel pitch) to calculate the hyprtfocal distance and work with the DoF equations to get the effect you want.  Oh, I forgot.  You fuss with diffraction.  Then use f/whatever no diffraction and hope for the best.

[Ray]

You can aslo run through the optical equations knowing only the peron is 6' tall and is Y feet away, I am using an Xmm lens and I want half of that person to be 20" tall on the print.  I won't bore you here with those details.

Yes, I believe you can. If you have sufficient information about the subject being photographed and you have the means of measuring distances and angles and the means of measuring the size of objects in the viewfinder or on the ground glass of an LF camera, and you know the magnification relationship between the camera format and the image size in the viewfinder or ground glass, then you could also even dispense with knowledge of the focal length of the lens being used. You could work it out.

Could we then say that the focal length of the lens is unattached to DoF considerations?

As an aside, in another part of this site there are some images for critique.  On at least one, there are suggestions to crop off about half the submitted image.  If the same size print is now made from half the original file, doesn't that change the DoF assumptions?  And then we still don't know (or care) what the origianal format was.  It may already be severely cropped or shown full frame or something in between.

This is not relevant. All images without exception are cropped. Initially the image circle produced by the lens is cropped by the camera's format size and aspect ratio. The image might be further cropped in post processing, but that's not relevant to the discussion if it is assumed that the person doing the cropping knows the dimensions (format) of what's being cropped and knows the new dimensions (new format) of the resulting crop. I think that's a reasonable assumption, don't you?

[Ray]

Ray, you don't need to know the format of the camera.  You only need to be able to measure the image size in the viewer and relate that to image size in the camera.  That is format independant.
[a href=\"index.php?act=findpost&pid=80160\"][{POST_SNAPBACK}][/a]

Isn't the image size in the camera the format?

[howiesmith]

Isn't the image size in the camera the format?
[a href=\"index.php?act=findpost&pid=80163\"][{POST_SNAPBACK}][/a]

No.  Of course not.  Surely you know better.  Does the camera format change when I change lenses (or zoom) or move about?  I can increase the tree's size in the view finder by increasing the focal length (zoom in) or move closer.  All with the same camera and its fixed format.  I am not talking about the imaging size of the sensor or film.  I am talking about the focused image size of the subject.  It is focal length and distance dependant, but not camera format dependant.

If I use the same digital camera and look at the ficused image size in its view finder and tehn connect that same digital camera to a laptop computer, the tree size viewed will change but the camera format does not.  Now I need to know the relationship to the tree size on the computer screen to the tree size in the camera.

[gkramer]

"Ray, you don't need to know the format of the camera... I never use the camera's format or film size for designing prints.  Just the degree of enleagement I want from a focused image.  I could crop each negative with a pair scissors, get the same print without ever knowing how much negative I cut off and threw away..."

If we're comparing two images shot with the same lens at the same aperture, same subject-to camera distance, etc., then sensor size (or film size) doesn't matter, since the smaller (APC-sized) sensor amounts to a crop of the larger one, and DoF etc. within the cropped area of the image will be the same--though of course the smaller, cropped image (or cropped area on the negative) can't make as large a print to get the same DoF effects (viewed at the same distance, with the same eyes, etc.)

If, on the other hand, we compensate for the smaller sensor size by using a shorter focal-length lens to get the same FoV, then Wrotniak's "M x A rule" applies, which implies that the smaller sensor will have somwhat greater DoF for a given aperture (at the same final print size), or that the full-frame camera's longer lens will have to be stopped down more to get the same DoF effect on the final print.

All this assuming that the apertures and degree of enlargement in question are within the limits imposed by diffraction and the pixel pitch and/or resolution of the film.

[howiesmith]

If we're comparing two images shot with the same lens at the same aperture, same subject-to camera distance, etc., then sensor size (or film size) doesn't matter, since the smaller (APC-sized) sensor amounts to a crop of the larger one, and DoF etc. within the cropped area of the image will be the same--though of course the smaller, cropped image (or cropped area on the negative) can't make as large a print to get the same DoF effects (viewed at the same distance, with the same eyes, etc.)

[a href=\"index.php?act=findpost&pid=80167\"][{POST_SNAPBACK}][/a]

That is absolutely correct.  Format doesn't matter.  

If I reduce the focal length to get the same scene onto a smaller sensor, then the focused image size will be smaller and I will need greater magnification to the final print.  Format still doesn't matter, but DoF has changed because I have a different focal length lens and degree of enlargement of the focused image.

[Ray]

No.  Of course not.  Surely you know better.  Does the camera format change when I change lenses (or zoom) or move about?  I can increase the tree's size in the view finder by increasing the focal length (zoom in) or move closer.  All with the same camera and its fixed format.  I am not talking about the imaging size of the sensor or film.  I am talking about the focused image size of the subject.  It is focal length and distance dependant, but not camera format dependant.

Ray, you don't need to know the format of the camera.  You only need to be able to measure the image size in the viewer and relate that to image size in the camera.  That is format independant.

Howie,
One of us is totally confused. Let me re-phrase your statement immediately above. You measure the image size of the subject on the ground plate of your 4x5 camera. It's 1" high. You know that the relationship between the image on the film and the image on your focussing screen is 1:1. In other words, your ground glass viewer is the same size as the camera's format and the subject size on the film is also 1" high. You know, and have to know, the camera's format in order to know that relationship.

If you don't know the relationship between between the image in the viewer and the image in the camera, then the 1" measurement means nothing. Right?

[Anon E. Mouse]

Well, whatever your personal system, the answer to the poster's question is clearly DOF does change with format size.

[howiesmith]

Howie,
 You know, and have to know, the camera's format in order to know that relationship.

[a href=\"index.php?act=findpost&pid=80175\"][{POST_SNAPBACK}][/a]

No.  I don't nrrd to know that.  The focused image size is the same for a given focal length lrns and ficus distance.  Trgardless of format.

The example of the same camera with an LCD diplay on the back and a laptop connection have much different image sizes from the same camera format.  It is an electronic thing.  Nothing to do with format.  IF I know a half inch image on the LCD is an inch in the camera, and a 12 inch image on the laptop is an inch in the camera, how is camera format involved.  Same camera. two different image sizes on two different viewers.


All that I have been trying to say is that if you use two cameras of different formats but with the same focal length lens, same f/stop, same CoC and same focus distance, to make equal prints (same sizw focused image), DoF is the same for bothrints.  The only variable changed is camera format.  DoF does not change.  The logical conclusion is DoF is independant of format.  Format doesn't matter.  Change one and only one variable at a time to easily and correctly see the effect of the changing variable.  Don't confuse the issue by changing moe than on variable at a time.  

It isn't that hard.  I'm thinking you are either being obstenant or putting me on.

[howiesmith]

Well, whatever your personal system, the answer to the poster's question is clearly DOF does change with format size.
[a href=\"index.php?act=findpost&pid=80177\"][{POST_SNAPBACK}][/a]

Wrong.  Once more:

"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, ... ."  Emphasis added

[Ray]

It isn't that hard.  I'm thinking you are either being obstenant or putting me on.
[a href=\"index.php?act=findpost&pid=80179\"][{POST_SNAPBACK}][/a]

Well, I guess I am being a bit obstinate. I usually am when I think I'm right   .

There is a certain logic to your argument but I find it verging on the absurd. I agree, however, that one could dispense with the terminology of format and use information provided by the manufacturer, who does require knowledge of the camera's format, such as the relationship between the size of the image in the viewfinder and the image in the camera.

We could rename 'format' as being 'the maximum size crop the camera can make from the image circle of any appropriately designed lens attached to it.

You could then turn a blind eye to what that maximum size actually is on the grounds that you never use it, but I fail to see how you can make a formatless image. Whether you use the maximum format that the camera is capable of, to compose your images, or a smaller, self-imposed format, you still have to know the format.

For example, if I were continually dismayed by resolution fall-off and vignetting with my 5D (which I'm not, but I am occasionally) I could carefully draw a rectangle in the 5D viewfinder that corresponds with the FoV I would get from a 20D and paint the surrounding area black. I could then use the 5D as though it were a 20D which crops the full frame 35mm format, and all my DoF calculations and lens choices would be almost exactly the same as they would be if I were actually using a 20D, but not quite the same because by doing this foolish thing I'd have a lower resolving camera than the actual 20D, 5mp as opposed to 8mp, and the enlargement factor (or degree of interpolation required to make a print) would be greater for my modified 5D. To put it another way, the minimum size CoC I would ever use would be slightly smaller using the actual 20D, so in this sense pixel density, total pixel count and format all have a bearing on DoF.

[howiesmith]

Ray, I am assuming that "format" is the x by y dimensions of a sensor, be it film or digital.  I think this is commonly accepted.  I don't care what happens outside those dimensions.  It isn't recorded on the sensor.  

Using that definition of format, each time I crop a negative, be it film or digital, I change the x by y dimensions of what's left or its format.  I do not in anyway change what is on the new format.  As far as the information on the new format is concerned, nothing changed.  Not the subject size, exposure or, yes Ray, the DoF.  If I make an 8x10 print, then cut it in half, the characteristics of each half are not changed.  Only the print format has changed.  The DoF of each half is unchanged.  It is so simple as to approach absurdly so.

To more easily understand the effect of changing a variable (format, focal lenth, focus distance, f/stop, etc.) it is very useful to devise tests that change only that single variable while keeping all the other constant.  For instance, it is much easier to determine the effect of changing f/stop if f/stop is the only variable I change,  If I also change camera format and lens focal length and degree of enlargement, I just might, might get confused as to which changed variable is making the change in the outcome.

[Ray]

Wrong.  Once more:

"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, ... ."  Emphasis added
[a href=\"index.php?act=findpost&pid=80180\"][{POST_SNAPBACK}][/a]

DoF doesn't care about anything at all; neither format nor picture, nor if you are feeling well today. It's merely a conceptual tool which should be used, in fact has to be used, with format in mind to get the desired results.

Much of the time, as I wander about these ruins at Siem Reap, my camera is set on f11 in aperture priority mode. Sometimes I come across a bas relief in good condition and in interesting lighting, zoom in on it, forgetting my camera is set on f11, take the shot, realise the shutter speed was unnecessarily slow, stop up to f5.6 and retake the shot.

I've got 2 virtually identical shots. Hopefully the second shot will be marginally sharper, but maybe not, or perhaps the f11 shot will be sharper in the corners. The DoF formulas will tell me the f11 shot has greater DoF. But it hasn't. Wouldn't you agree? This is because DoF does not care about format, nor the composition, nor the picture. But I, as the photographer using the concepts of DoF, have to care about format.

[howiesmith]

Wouldn't you agree? ...  But I, as the photographer using the concepts of DoF, have to care about format.
[a href=\"index.php?act=findpost&pid=80213\"][{POST_SNAPBACK}][/a]

I have absolutely no idea what you are even trying to say, let alone agree with you.

If you think DoF is related to format, knock yourself out.  Keep on doing what you do.

Truth is an interesting thing, Ray.  It doesn't care whether you believe it or not.  It just goes on being true.  (I know, truth, like DoF, doesn't care about anything.)

[Ray]

I have absolutely no idea what you are even trying to say, let alone agree with you.
[a href=\"index.php?act=findpost&pid=80221\"][{POST_SNAPBACK}][/a]

So be it. I've done my best. Taking format into consideration works for me. I've got a glimmer of an understanding what you're talking about, I think. But I can't help associating your views with that of religious (photographic)  fundamentalism.

Ie. the basic DOF formulas do not include a format specification, therefore format has no bearing on the issue. The DoF calculators on the internet seem a little bit more progressive. To use them, you need to specify the format of your camera, by model name, to get an appropriate CoC in relation to an 8x10" print. As the great bard wrote, 'a rose is a rose by any other name'. (I've probably misquoted).

[howiesmith]

I know from hearing snippets of conversations that reduced sized sensors have different DoF characteristics than that of a FF sensor with same lens, f/stop, etc.  But I don't know why.  Could I persuade anyone to enlighten me, please?
[a href=\"index.php?act=findpost&pid=65745\"][{POST_SNAPBACK}][/a]

macgyver,

I think your problem may be "snippets" from less than experts who fail to provide all the assumption necessary to make there position plausible.  Luckily, if you really want to know, you need not take the words of a stranger or even one trusted expert.

My opinion is DoF is not dependant on format.  Surely another "snippet," but you need  not just agree.  In fact. please don't.  Some research will convince you.  Read "Understanding Depth of Field" on this site.  You may trust Michael Reichmann.

It is my understanding that smaller format cameras have greater DoF because they usually have shorter focal lenth lenses than their full frame brothers, and not because they are smaller format.  Hope this and all the previous posts, if you could manage to actually read them, help.

[Ray]

It is my understanding that smaller format cameras have greater DoF because they usually have shorter focal lenth lenses than their full frame brothers, and not because they are smaller format. 
[a href=\"index.php?act=findpost&pid=80258\"][{POST_SNAPBACK}][/a]

This is the sort of circular argument that doesn't get us anywhere. I'm using a short focal length lens because my camera is a small format. The greater DoF is due to the fact that the lens has a short focal length and has nothing to do with the format.

There's a self-evident truth here if you are referring to the DoF of the aerial image projected by identical lenses used with different formats However, if you are referring to a finished print of a given size, there's a lot more going on which influences the final perception of DoF as well as the manner in which it is calculated, specifically choice of CoC.

Let's take a concrete example and compare 2 currently available cameras of different format, the new 10mp Canon G7 P&S and the older 8mp EOS 30D DSLR, both of which can use 10mm lenses.

Your argument would be, if I've understood you, that in principle a 30D image cropped to the same size as the G7 sensor would exhibit the same DoF on equal size prints, both shots taken from the same position, using not the same lens, which is not possible, but different lenses of the same focal length and at the same f stop, say f4. Is that your position?

The 30D sensor is approx 340.5 square mm in area. The G7 sensor (7.18x5.32mm) is approx 38.2 square mm. If I crop the the 30D image to the same size as the G7 image, the resulting image is less than 1mp.

The G7 lens is certainly going to be a lot sharper at f4 than the EF-S 10-22mm at f4. The 10mp G7 image at the focal plane is going to be a lot sharper than the 1mp 30D image (actually only 900kp), yet the OoF parts of both images will probably be quite close in definition or degree of blurriness.

If I make A3 size prints from both images, experience tells me the G7 print will have noticeably less DoF than the print from the heavily interpolated 30D image which will probably beging to resemble an image from a pin-hole camera.