nsbca wrote:
slin100 wrote:

A practical consequence of the effect of sensor size on DOF is that one should use the DOF markings on a lens approximately 1 1/3 stops wider than the actual aperture when using the lens on a 1.6x crop camera.


I don't buy that. That is the part that doesn't change.

It does and I explained why. You can also read about it at http://www.photo.net/learn/optics/dofdigital/.
[Edit: To be specific, look at the Addendum in the article above that indicates the need to use DOF scales for an aperture wider than the actual aperture when using a lens on a 1.6x crop camera.]

Now, would you care to explain why you don't buy it?

What a fire storm to a simple question. I am well aware that distance and lens impact DOF but that wasn't the question. Nor was the impact of distance, lens or aperture even a consideration when the question was framed because the question was not will the need to change focal lengths to accomadate different sensor sizes to produce the same picture result in different DOF? The answer to that is obviously yes. But that wasn't the question.
The question arose from the assertion that the sensor size itself impacts DOF. To me that made no sense but I often miss things so I thought I would ask.

To the poster who asked who suggested that view camera users use F64 and swings and tilts to get acceptable DOF... there is a difference between acceptable DOF and great DOF. Why do view camera users use F64 and S/T? Because they can! When done properly and for good purpose the DOF on a view camera far exceeds anything available on a 35mm or medium format but it is not because the film size is bigger, it is because the film and/or lens plane is altered to match the subject.

slin100 wrote:
nsbca wrote:
slin100 wrote:

A practical consequence of the effect of sensor size on DOF is that one should use the DOF markings on a lens approximately 1 1/3 stops wider than the actual aperture when using the lens on a 1.6x crop camera.


I don't buy that. That is the part that doesn't change.

It does and I explained why. You can also read about it at http://www.photo.net/learn/optics/dofdigital/.
[Edit: To be specific, look at the Addendum in the article above that indicates the need to use DOF scales for an aperture wider than the actual aperture when using a lens on a 1.6x crop camera.]

Now, would you care to explain why you don't buy it?



To get an "equivalent field of view" it does change. Other then that it does not.

This is from your link:

Bob Atkins:

If you use the same lens on a EOS 10D and a 35mm film body and crop the 35mm image to give the same view as the digital image, the depth of field is IDENTICAL

What your talking about is the magnification of circles of confusion (which would have been a good title for this thread if you dropped the word circles) and DOF effects and not the actual depth of field, the area that is in focus.

nsbca wrote:

This is from your link:

Bob Atkins:

If you use the same lens on a EOS 10D and a 35mm film body and crop the 35mm image to give the same view as the digital image, the depth of field is IDENTICAL


That's point #3 from Bob's article, and it doesn't apply to this situation. Point #2 from Bob's article applies:
Bob Atkins:
Using the same lens on a EOS 10D and a 35mm film body, the 10D image has 1.6x LESS depth of field than the 35mm image would have (but they would be different images of course since the field of view would be different)


Bob goes on to illustrate the practical application of this point (bold text is my emphasis):
Bob Atkins:
Now let's look at the case when this lens [Bob speaks of a 17mm lens from a 35mm film camera] is used on an EOS 10D. As the article indicates, if you use a given lens on a smaller format, depth of field is reduced and the angular coverage ("effective 35mm equivalent focal length) decreases. In the case of the 10D it's reduced by a factor of 1.6, so it gives the same angular coverage (field of view) as a 27mm lens on a full frame 35mm body. The circle of confusion value for the 10D is reduced by a factor of 1.6x and what this means in terms of DOF scales is that you need to use the markings for about 1 1/3 stops wider aperture in order to estimate the DOF. So with the lens set to f16, we need to look at the DOF scale markings about 1/2 way between f8 and f11 as shown below.


What your talking about is the magnification of circles of confusion (which would have been a good title for this thread if you dropped the word circles) and DOF effects and not the actual depth of field, the area that is in focus.



All kidding aside about dropping the word "circles", DOF and circle of confusion go hand in hand. There is no separating one from the other.

Hrow wrote:
The question arose from the assertion that the sensor size itself impacts DOF. To me that made no sense but I often miss things so I thought I would ask.

Who made that assertion?

The whole concept of DOF is predicated on producing a specific size print and viewing it from a standard viewing distance. The ability of, or lack thereof, the human eye to critically discern a circle as an out of focus disk is the essence of DOF. The concept of viewing points on a print to determine DOF (point is out of focus or in focus) is Circle of Confusion. The circle is the point and when out of focus enough that the eye can see it is out of focus it is a Circle of Confusion. CoC is part of the DOF formula:

Hyperfocal Distance: h = f2 / ac

f = focal length of lens
a = aperture diameter (f/stop number)
c = circle of confusion

Depth of Field, Near limit: hs / h + (s - f)

Depth of Field, Far limit: hs / h - (s - f)

h = hyperfocal distance
s = distance from camera to object
f = focal length of lens

This is from the American Cinematographers Manual, 8th edition, pages 698-699.

Every DOF calculator uses an assumed CoC that varies by size of film or chip, as does every manufacture when they make DOF charts and scales for lenses. The size of the CoC changes in all the calculators when you change the size of format of film or digital chip. Smaller chips have smaller CoC and when they are enlarged to the standard size print, the enlargement of small chips enlarges the CoC and decreases DOF.

Does the size of the film or chip effect DOF?

Yes, unless you live in a vacuum and don't make prints. As soon as you start enlarging the file the CoC is enlarged and small format film and chips are enlarged more than larger formats films and chips to produce the same size print.

This assumes you are not changing distance to subject, focal length of the lens or the aperture. In other words, all the variables in the formula for DOF remain constant, except the CoC. CoC changes with the degree of enlargement (how big is the print) and how far away are you viewing the print.

Exactly, Jeff. Thanks.

In a practical sense, all of the constants remaining the same is at times unlikely. Photographers want to make a persons head a certain size. If the head is too small or too large the image is not acceptable by normal standards. The photographer then changes one or two of the variables to make the head size "acceptable" and the DOF increases dramatically. This is best illustrated by looking at the lens on Point and Shoot camera and seeing the 6mm is the "35mm equivalent" to a 50mm lens. By inserting a 5 into the DOF formula instead of a 50 dramatically increases DOF even after accounting for the magnification change in CoC when printing.

Nicely said Jeff.
How does cropping effect the DOF?
IE, 8x10 print from a FF sensor. If we crop the image, then enlarge the resulting image back to 8x10, does the DOF increase or decrease? Decrease correct? Same effect as using the longer FL to get the same FOV?

Jeff, I agree with you. When it comes to DOF, focal length length trumps sensor size since the former has a quadratic effect and the latter has a linear effect. I have just been trying to point out that one shouldn't ignore the latter.

Cropping, whether done in Photoshop or an enlarger in a darkroom decreases DOF because the Circles of Confusion are made larger. This assumes you make the same standard size print and view from the same distance.

This same concept also results in confusion about digital cameras having a "Crop Factor" or a" Magnification Factor." The net effect of have a camera with a smaller sensor than a "Full frame" sensor is that the subject on the standard size print is larger with the crop than the full frame. Think of the concepts however you like, but if you make prints from your digital files then the subjects will be larger in the prints from crop cameras than full frame cameras, assuming the other variables remain constant (lens focal length, distance to subject etc.)

So in other words, DOF is NOT affected by sensor size since the crop of an image from a FF sensor made and stretched to match that of an image from 1.6x sensor is IDENTICAL.. now stop trying to show off and impress everyone with your (referring ot more than one person) vast knowledge, stop calling folks clowns and simply accept that you are WRONG. you can twist it and talk about the effects of magnification circles of confusion and the rotational axis of Jupiter but ,unfortunately, the original poster didnt ask about these things nor their impact. take a FF image. crop it
stretch it. and you get the same image.

now lets all blame Hrow for starting this mess.

and this statement couldnt be more true...
Hrow wrote:
...........There is a lot of good info on this site but there is also a small percentage that is just bulls**t.


Edited by EnCapture on Aug 16, 2005 at 06:14 AM GMT

EnCapture wrote:
the crop of an image from a FF sensor made and stretched to match that of an image from 1.6x sensor is IDENTICAL

No quibble with this statement (because the condition where the DoF is identical is where the sensors are the same size - the CROPPED FF and the 1.6), but I suspect the OP didn't envision only using a small part of a larger (e.g. FF) sensor when he asked the question

This was asked in the Nikon forum this week as well, here is a neat explantion from Molson

The “Circle of Confusion” is an arbitrary factor, used as an aid in selecting an appropriate depth of field to provide subjectively “acceptable” sharpness for a desired enlargement size. It is a term that has been adopted as an attempt to describe the amount of subjective “unsharpness” that is acceptable to the human eye in a photographic print. It has ABSOLUTELY NO BEARING WHATSOEVER on the physical depth of field properties of a lens or camera format. As someone rightly pointed out, you can't change physics, even if you pretend you can by posting contrary notions on your web site.

What the authors of the popular Depth of Field calculator web site haven’t realized yet, is that with digital capture, the degree of enlargement required to produce a given size image is NOT a function of sensor size, as it was with film. The degree of “enlargement” required to make a specific size of print from a digital capture is dependant on the NUMBER OF PIXELS in the sensor used to capture the image (hint: that’s how you can make bigger prints with your 6 megapixel camera than you could with your old 3 megapixel camera). The size of each pixel is irrelevant, just as the size of film grain is irrelevant in depth-of-field and circle-of-confusion calculations.

Let me illustrate this with a real-world example. Slightly LESS "enlargement" is required to make an 8x12 print from a Nikon D2X file than from a Canon EOS 1Ds file, because the Nikon image contains more pixels (4288 x 2848 pixels, versus the Canon’s 4064 x 2704 pixels). The less the original image needs to be enlarged to produce the desired final print size, the larger the acceptable Circle of Confusion can be before it degrades the subjective appearance of the print. Therefore, since the Nikon D2X images require some 5% less enlargement than EOS 1Ds files to produce an identical print size, the acceptable Circle of Confusion for the D2X is then 5% LARGER than it is for the EOS 1Ds. Using the logic presented by the “full frame fanatics”, this would mean that the D2X has LESS depth of field than the full-frame EOS 1Ds, even though the 1Ds has a larger imaging sensor.

I won't start to delve into why interpolation is not the same as traditional analog enlargement, which makes the old Circle of Confusion arguments even weaker...

Good grief, how do people manage to make this stuff seem so complicated.

It's just so simple. If you have a different sized sensor, you WILL use a different focal length, that is a direct consequence of the different sized sensor. You cannot move yourself forwards or backwards to get the same composition, a given composition can be taken from only one point, otherwise you change the relationship between the size of different objects at different distances from the camera.

So, you have two cameras, an FF and a 1.6 crop. You stand in the same place and take the same picture. You do this with a 50mm lens on the 1.6 crop camera, and an 80mm lens on the FF camera. To get the same DoF with the FF camera as the 1.6 crop camera you need an f-stop 1.6 times greater, e.g. f/3.2 on the FF camera compared to f/2 on the 1.6 crop camera. That's all there is to it, try it out with a DoF calculator, it's simple to prove.

EnCapture wrote:
So in other words, DOF is NOT affected by sensor size since the crop of an image from a FF sensor made and stretched to match that of an image from 1.6x sensor is IDENTICAL.. now stop trying to show off and impress everyone with your (referring ot more than one person) vast knowledge, stop calling folks clowns and simply accept that you are WRONG. you can twist it and talk about the effects of magnification circles of confusion and the rotational axis of Jupiter but ,unfortunately, the original poster didnt ask about these things nor their impact. take a FF image. crop it
stretch it. and you get the same image.

now lets all blame Hrow for starting this mess.

and this statement couldnt be more true...
Hrow wrote:
...........There is a lot of good info on this site but there is also a small percentage that is just bulls**t.


In what way is taking a full frame image and cropping it different to taking a shot with a cropped sensor
The size of the sensor/negative is important as it fixes the minimum amount of magnification required to produce a print of a given size. The amount of magnification effects the DOF therefore the size of the sensor effects DOF when all other factors are constant. QED.

I can't believe someone brought pixel pitch into this. So long as it's small enough, pixel pitch does not have any effect on DOF. The argument falls through if we switch to film (fully expecting someone to resort, "What's film?").

jmcfadden wrote:
(hint: that’s how you can make bigger prints with your 6 megapixel camera than you could with your old 3 megapixel camera).

John, I would submit that this particular argument is with regard to how well the printer can reproduce the part of the image that's in focus, once you specify a ppi for the print and a viewing distance etc. The number of MP (or pixel size) will no doubt influence the size at which someone might think of printing an image, so for digital and discussions of DoF, we still have to standardize the conditions if we want to establish meaningful comparisons. Once photosite (and their equivalent grain cluster) sizes are sufficently small, do we not end up with the same considerations as before to determine DoF - i.e. if the CoC is bigger than the photosites, then the photosites just record the CoC more faithfully and we'll still see it when we enlarge? The analog, errr, analogy is the development of finer and finer grained films, is it not?

Sorry I asked! Can we just move back to arguing about the merits of non-existent cameras?

bouch wrote:
Imagemaster wrote:
Neither sensor size nor film size affects DOF. Cameras and lenses do, but that was not the question.

WRONG

What does this mean? The format size doesn't affect DOF but the camera somehow does? How do you figure? Isn't format size sort of tied to the camera?

I can pretty easily demonstrate that my digicam has more DOF than my 1Ds. The cause of this according to you, is not the format size, but some other inherant propropty of the camera? Do you mind telling me what?

Why don't you just stay out of these discussions if you don't have clue?

You might read the aforementioned article on photo.net, in particular the part about circle of confusion. The photo.net article is pretty much right on the money.

Hrow wrote:
Sorry I asked! Can we just move back to arguing about the merits of non-existent cameras?

. Actually, you should really consider the merits of non-existent cameras, lens, and printers to get to the bottom of this whole depth of field thing.

stevei wrote:
It's just so simple. If you have a different sized sensor, you WILL use a different focal length, that is a direct consequence of the different sized sensor. You cannot move yourself forwards or backwards to get the same composition, a given composition can be taken from only one point, otherwise you change the relationship between the size of different objects at different distances from the camera.

So, you have two cameras, an FF and a 1.6 crop. You stand in the same place and take the same picture. You do this with a 50mm lens on the 1.6 crop camera, and an 80mm lens on the FF camera. To get the same DoF with the FF camera as the 1.6 crop camera you need an f-stop 1.6 times greater, e.g. f/3.2 on the FF camera compared to f/2 on the 1.6 crop camera. That's all there is to it, try it out with a DoF calculator, it's simple to prove.

Exactly, thanks Steve

I think John's analysis above is incorrect in practical terms although at extreme limits it may have some validity. Example, how much DOF does a 640x480 image have if I print it 40x50? - probably none because it will look so crappy that nothing appears to be in focus. The D2X & 1Ds (and any of the orher relevant digital SLRs) are high enough resolution that for reasonable sized prints the in focus area can always appear sharp.

Lets get this straight. DOF, Depth of Field refers to the area foreward of and behind the focus point that is still in focus.

At any given distance, any given focal length and any given aperture setting the area in focus, or Depth of Field will not change by merely changing the size of the sensor.

Circles of confusion are DOF effects, that is the aesthetics of the out of focus area, outside the the area in focus. These are dependant on aperture sttings and focal length, but have little to do with the size of the sensor other then the amount of up-sizing needed to compensate for the smaller sensor. That would be percieved DOF effects. Nothing more.


In other words blowing a print up from 5x7 to 8x10 or 16x20 no more changes the actual DOF of a lens/body set-up then does the size of the sensor.

nsbca wrote:


Circles of confusion are DOF effects, that is the aesthetics of the out of focus area, outside the the area in focus. These are dependant on aperture sttings and focal length, but have little to do with the size of the sensor other then the amount of up-sizing needed to compensate for the smaller sensor. That would be percieved DOF effects. Nothing more.


In other words blowing a print up from 5x7 to 8x10 or 16x20 no more changes the actual DOF of a lens/body set-up then does the size of the sensor.



"the aesthetics of the out of focus area" - no, that's bokeh, which we covered in the previous lesson. This lesson is about DoF. The DoF is calculated from (aka defined by) the circle of confusion, not vice versa. The circle of confusion is "chosen" for a particular size print viewed at a specified distance, based on how well the human eye can, on average, perceive detail/sharpness. So if you routinely plan to print at a different size than that for which the DoF was calculated, you should recalculate the DoF scales and not use the ones on the lens. Earlier parts of the thread and the links to explanatory articles *are* worth reading!

nsbca for president!!!!

Edited by EnCapture on Aug 16, 2005 at 06:15 AM GMT