Confused by circle of confusion

[Stewart Ethier]

I want to work out depth of field for various formats, focal lengths, and focussing distances. The formulas are straightforward enough, but they all depend on the diameter of the circle of confusion. According to Kodak's Large Format book, this diameter is about 1/1000 of the focal length of the normal lens for the format (and multiply the result by 0.6 for critical use). Now this seems rather crude to me, because it doesn't take into account how much the film will be enlarged in the print, or how far the viewer will stand from the print. Are there more sophisticated formulas that take these factors into account? If not, let me be more specific: I want the diameter of the circle of confusion for no enlargement (i.e., contact prints) viewed at the viewer's minimum focussing distance. My intuition says that it shouldn't depend on format, e.g, for 8 x 10 and 20 x 24, the figures should be the same. Does this make any sense?

[N Dhananjay]

Yes, techniaclly, it does not depend on format. Howeever, given that LF is enlarged less for a given print size, the demands on COC are less. The formulae are reasonably starightforward to translate across formats.

Let's say a COC of Xmm on the print is seen as sharp. This means depending on enlargement, you need X/m mm on the negative (where m is the extent of enlargement). Looking at it another way, if you have Xmm COC on the negative, the COC on the print is going to be mX mm.

For example, lets say you need 0.5mm COC on the print at the shortest viewing distance. If you have a 0.5mm COC on the negative and you enlarge the negative by a factor of 2, the COC is now 1mm, which means it will look blurred at the minimum viewing distance.

However, keep in mind that the larger the print, the farther away the viewer stands (unless he/she is learning spotting techniques). So if normal viewing distance is A and the larger print means he stands mA away, the mX mm COC on the print still looks like it is Xmm. So with the above example, though the COC on the print is 1mm, the viewer is standing at twice the distance away and the COC will look like its 0.5mm. This is actually a pretty good approximation because it is said viewers typically stand at a distance equal to the diagonal of the print. Even if viewers don't exactly stand a diagonal length away, they do move back for the larger pictures. However, if it is the example of contact prints of the same subject from 8x10 and 20x24 negatives and they are both viewed at a distance of 10 cm, both prints will require the same COC on the print. But why would you look at prints like that?

[Stewart Ethier]

Thanks for your reply, which sounds exactly right.

"But why would you look at prints like that?" i.e., from a distance of 10 cm.

There is currently an exhibit at the Salt Lake City Art Center consisting of contact prints (5 x 7 to 12 x 20) and 20 x 24 Polaroid prints. After looking at the images from a respectful distance, I find myself going in close to examine the details. This is especially true of the 20 x 24 Polaroids, some of which are head shots more than twice life size. This got me to wondering what apertures are needed to keep the eyes (at least) within the depth of

[Bruce Gavin]

Here is a DOF calculator on the Net that uses different COC values by film size:

http://www10.geocities.com/SiliconValley/7116/jv_dofc.html

[John H. Henderson]

Yeah...the theory goes something like "The viewer will always step back to a distance equal to the diagonal of the print, so CoC on the negative doesn't have to be scaled to match different print sizes, but has to be scaled for the total enlargement, so 35mm, because it must be enlarged more for a given print size, must have a smaller CoC than 4x5"."

This greatly bugged me when I read it after buying 4x5" equipment. This theory begs the question, "If I am going to allow greater degradation for a larger negative, why bother going to large format?"

Of course, my concern assumes that the lens and film resolution in a given format are better than the CoC allowed.

It still bugs me. But if I calculate my own DoF charts, I can use whatever CoC I want, and I will probably lean toward the smaller ones, like that used for 35mm, and end up with smaller apertures and longer exposure times, or I might as well go out and get a 6x9cm view camera or a 35mm with shift lens.

[Alan Gibson]

Don't forget that the simple theories around CofC are concerned with what is 'acceptable'. You might find that what is 'acceptable' to you doesn't quite match any particular theory.

My experience is that the better the equipment and matrials, the smaller the acceptable CofC (of the final print). Put it another way, a slightly out of focus print from a cheap 35mm P&S can be more acceptable than an equally out-of-focus shot from a better SLR, which in turn looks better than a good lens on a 5x4. Of course, when he cameras are in focus, the order is reversed.

[Sean Billy Bob Boy yates]

Of course, there's no reason to restrict yourself to a circel, either. You could use the dodecahedron of confusion just as well.