JD750 Loc: SoCal

It all scales linearly.DOF doesn't magically appear as a part of the enlarging process. DOF is a function of the optics and it is inherent in the recorded image.

Let's try one more time to do a thought experiment. So take a camera, shoot a pic, and print it out. Now take a cutting tool crop out the center of that print, to APS-C dimensions. Do you think the DOF in your cropped print will magically change because you cropped it? That is the same exact thing as if you used that same lens on an APS-C camera.

DWU2 Loc: Phoenix Arizona area

I'm thinking jd750 was right: 10 pages.

bsprague Loc: Lacey, WA, USA

7 might be enough.

Borrowing from another thread, I either like an image or I don't. Sensor size has nothing to do with it.

Ysarex Loc: St. Louis

It's not magic it's human vision. Cropping a print does change DOF -- see illustration below. The fellow wearing a hat is within the DOF of the image on your left and outside the DOF of the image on your right. Depth of field is defined and measured in the final print.

One of the best DOF calculators can be found at this site: https://www.photopills.com/articles/depth-of-field-guide Also located there is correct information like this:

"What affects depth of field?
Depth of field depends on aperture, focus distance, focal length and circle of confusion (CoC). The latter depends on camera sensor size, final image print size, image viewing distance and viewer's visual acuity."

Still waiting for you to explain why every DOF calculator disagrees with you. Until you do you're proven wrong.

Return to this post: https://www.uglyhedgehog.com/t-776133-6.html#13955718
Note the formula as well as the included table for CoC values based on film/sensor size.

Explain how changing the value for c in the equation does not alter the result of the calculation (your claim). Provide some examples please. Until you do you're proven wrong.

Depth of field is defined and measured in the final print.
Here's another excellent DOF calculator that is supplied with correct explanation. Read the supplied commentary. It clearly states you're wrong and explains why: https://www.cambridgeincolour.com/tutorials/dof-calculator.htm
Depth of field is defined and measured in the final print.

JD750 Loc: SoCal

I know that this is true and I need to give it some thought and provide a reply.

Something is going on here that's subtle. So let's quit calling each other wrong (I have been guilty as well) and try to figure out what we are seeing differently. Henceforth I will leave the emotion out and concentrate on the subject.

If you crop an image and do not enlarge the image, like the example I gave with coping a printed image, obviously DOF cannot change. Would you agree?

BUT if I enlarge that cropped image, you say DOF changes. This could be the case. That is because resolution is changing and also the enlargement is magnifying any small imperfections. So what looked sharp zoomed out might not look as sharp when zoomed in. Again, would you agree?

Given two sensors of different physical sizes FF and APS-C but equal numbers of pixels, say 20 MegPix then the APS-C image is going to be sharper than a FF Image zoomed (by 1.5x or 1.6x) to yield the same framing as the APS-C image. Again, would you agree?

Additionally the center of the lens tends to be sharpest, so the APS-C has an advantage there as well.

So zoomed in image or enlarged images, will have less resolution than the original, which will result in lower sharpness. But that lower sharpness is across the entire frame.

Here is where it gets tricky. The enlarged or cropped (zoomed in) image will have less sharpness, throughout the frame but will reach the absolute fuzziness limit quicker than the un-enlarged image, as you move away from the plane of sharpest focus. The enlarged (or cropped) image is starting out fuzzy, part way to the absolute limit, and has less to go to get to that limit.

But that is not changing the optical resolving power of the lens, it's just reaching that limit quicker as you move away from the plane of sharpest focus, because of the magnification.

I suspect that is why DOF calculators are showing the differences for different sensor sizes.

But has DOF really changed? If you define it as an absolute point of fuzziness, then yes. However If you remove resolution from the equation, then it should be the same for both images. And that's what we saw with my test images, where the FF zoomed in to the same framing as the APS-C image, was about the same number of pixels as the APS-C.

Ysarex Loc: St. Louis

No. Because you're methodology then is faulty. As such you can't refer to DOF as changing or not. Proper DOF measurement requires enlargement to a standard print size and viewing of the print from a standard viewing distance. You must do that -- it's not optional. If you remove one of the DOF measurement requirements then you are no longer correctly measuring DOF.

It is the case. There's no "could" about it.

Exactly. And that has been incorporated into a correct definition of and proper methodologies for measurement of DOF. It's built into the math.

Now I agree and there's no need to go further.

Long before you and I were born, during the 19th century, the engineers and scientists at Voigtlander, Zeiss, Kodak et. al. worked through all of this and figured it out. Conceivably it took some time and effort, some missteps and revisions, even some heated arguments until they were able to codify a good working definition and usable methods of measurement. Clearly their goal was to present to the photographic community a functional understanding. For example they used the variable f/stop in the equations they developed instead of the diameter of the lens entrance pupil. When you're using your camera do you know the diameter of the lens entrance pupil or do you know the f/stop? What they gave us then was a pragmatic and functional definition of DOF that best matches our usage experience. They made the call back then to define DOF in the final print and the math incorporates that decision. They made the call back then to incorporate measurements of human visual acuity into the math so that the math better reflects what we experience.

We're now long past the point of refinements that may have been missed and, as an industry, we have long since agreed on the best definition and measurement methods to understand the phenomenon and to manage that phenomenon in our photographs. It's not rocket science. Those engineers and scientists in the 19th century knew all they needed to know to get this right. They got it right.

Here's an excerpt from Ilford's The Manual of Photography first published in 1890. It makes two points for you to consider. This from my copy 9th edition:

To summarize, the permissible diameter of the circle of confusion must be reduced if subsequent enlargement is to take place. In practice values of C from 0.025 to 0.033 mm are used for the 24 × 36mm format to allow ×8 enlargement. Values of 0.06mm and 0.1 mm are used for medium and large formats respectively.

It is worth noting that in the early days of small format photography a criterion of 1/1000 of the focal length of the lens was used for C, giving 0.05 mm for the 50 mm lens, which was for a long time the standard lens used for the 24 × 36 mm format, and correspondingly different values for different focal lengths. This criterion was in use for many years and is still quoted by some sources. It gives values for depth of field which imply different degrees of
enlargement for different focal lengths. Some confusion has arisen on depth of field matters, especially in the provision of tables and scales, as well as comparisons between equivalent lenses from different sources. The idea of C=f/1000 is now deprecated, and instead the value of C is taken as constant for a given format, for the whole range of lenses.

The first paragraph notes what we already know now from this thread: enlargement from different sized film/sensors is a DOF determinant variable expressed in the value of the circle of confusion.

Note the discussion in the 2nd paragraph concerning the CoC value for 35mm, recognition of a discrepancy from the "early days" and then that subsequent resolution. In other words this has long been worked out. We've had plenty of time to get this right and refine it -- we put our best people on it over a century ago. They got it right.

JD750 Loc: SoCal

Thank you for the reply and the definition.

Ysarex Loc: St. Louis

Happy to be of help.

JD750 Loc: SoCal

I think it’s interesting to note we were approaching the problem from 2 different directions. The Physics/optical definition and derivations of DOF do not account for enlargement of the image after it is recorded. For two examples see:
https://www.opticsforhire.com/blog/understanding-depth-of-field/#:~:text=Depth%20of%20field%20is%20a,achieve%20the%20best%20possible%20design
https://physicsandphotography.wordpress.com/2014/10/02/depth-of-field/

However since his is Photography Forum, we should use the Photography definition of DOF.

Ysarex Loc: St. Louis

I don't think there legitimately are physics/optical definitions of DOF -- just some people who misuse the term. I think it's absolutely fine to understand and do the math that describes the image formed by the lens on the film/sensor and to do that relative to an understanding of DOF. But it is also important not to confuse DOF with the image formed by the lens on the film/sensor. It's a real common mistake that a lot of people make. I kept repeating through this thread; DOF is defined and measured in the final print. It is not defined at the film/sensor plane.

That first article is very short and sketchy. The video (power point) is rather awkward as the narrator keeps stopping to ask a muffled voice what things are. They present standard DOF equations in the video however their definition of the circle of confusion is concerning.

They identify the sources of the circle of confusion as:
1. diffraction
2. lens aberrations
3. sensor pixel size

The Wikipedia article on DOF and circle of confusion identifies the sources as:
1. human visual acuity
2. print viewing conditions
3. enlargement from the original image

Not even a one item overlap in those two lists. Checking another trusted source (PhotoPills -- one of best DOF calculators) we find: "The Circle of Confusion is just a number that represents the diameter or the maximum size that a blur spot, on the image captured by the camera sensor, will be seen as a point in the final image by a viewer for a given viewing conditions (print size, viewing distance and viewer’s visual acuity).

As a result, once you’ve decided the values of sensor size, max. print size, viewing distance and visual acuity, you can calculate the Circle of Confusion. By doing so, you’re establishing the convention of what is considered to be acceptably sharp in the image." Nearly 100% overlap with the Wiki article.

Let's try another trusted source (Cambridge in Colour -- one of best DOF calculators) and we find: "In order to calculate the depth of field, one needs to first decide on what will be considered acceptably sharp. More specifically, this is called the maximum circle of confusion (CoC), and is based on the camera sensor size (camera type), viewing distance and print size. The default is to say that features smaller than 0.01 inches are unnecessary, when viewed in an 8x10 inch print at a distance of 1 foot (~25 cm). However, people with 20-20 vision can see features 1/3 this size. This calculator therefore also has the ability to adjust for parameters such as viewing distance, print size and eyesight — thereby providing more control over what is "acceptably sharp." Again agreement with the Wiki article.

I think it's worth noting that diffraction, lens aberrations and pixel size can complicate or impinge on the lens forming a sharp image and as such they can affect final DOF rendition to a small degree -- I don't think they're the major players and certainly not the only players.

This second article needs some work. I got a terrific laugh toward the end when he used his formula to calculate the DOF for a landscape photo. His subject he noted was 1 km distant and he calculated the total DOF as 265 km. He doesn't realize how nonsensical that is? Given the camera, lens and f/stop he used he'd get the same DOF result if he used a subject distance that was less than 1% of the 1 km he did use. 1 km and 0.001737 km would both generate a far limit of DOF at infinity.

Racmanaz Loc: Sunny Tucson!

I guess lens manufacturers got this wrong all along ha? 😀 This photo alone proves your point.