Depth of Field - color vs B&W

[Bill_1856]

Should the CoC be more for color than B&W in calculating DOF?

[Jim Jones]

CoC depends on many things: image use, viewing distance, subject matter, viewer preference, etc. I would expect CoC to usually be more critical in B&W than in color.

[Brian Ellis]

CoC depends on many things: image use, viewing distance, subject matter, viewer preference, etc. I would expect CoC to usually be more critical in B&W than in color.

Why?

[cotdt]

for critically sharp 4x5 film, what CoC value should be use for calculations of DoF?

[Leonard Evens]

I've never thought about this, but off the top of my head, I would think the controlling factor would be the resolution of the film and not whether it was color or b/w. It is harder to make a high resolution color film than a high resolution b/w film, but since the market is much larger for the former, film manufacturers have done pretty well. As usual, it is a matter of balancing speed against resolution.

A general rule of thumb is that the coc diameter is the reciprocal of the resolution in lp/mm, and vice-versa. So a coc of about 0.1 mm for 4 x 5 in the camera corresponds to about 10 lp/mm. Since most films we use run at least 50 lp/mm, it would seem that the film would have a relatively minor effect. But if you add in the effect of the lens, it gets more complicated. If the lens is yielding about 50 lp/mm and you use a 50 lp/mm film, the result might be something like 35 lp/mm. If you were aiming for a coc of 0.1 recorded in the film, that means you would have to start with a coc of approximately 0.096 mm in your dof calculations. In other words, you would have to reduce the coc by about 5 percent. If, on the other hand, you were more restrictive and insisted on a coc of about 0.05 (20 lp/mm), you would have to use a coc of about 0.41 mm (about 10 percent less) in your DOF calculations.

Since DOF estimates are at best approximations in any case, such calculations show that you probably can ignore film resolution in most circumstances. The effect is so small that it would be completely swamped by other uncertainties, such as the effects of lens aberrations, diffraction, and user focusing error.

[N Dhananjay]

The human eye-brain combination processes information (and not pictures - there is no little humunculus inside our heads looking at pictures transmitted through the optic channels). Color is seen as information and so a lower resolution matters less, in a sense, because the color adds a source of information. I suspect this is one reason why an image made by dye clouds can seem as sharp as an image made up of silver particles. So, I would actually suspect that you can get away with a larger CoC in color than in B&W. Cheers, DJ

[Jim Jones]

for critically sharp 4x5 film, what CoC value should be use for calculations of DoF?

Again, it depends. One commonly cited figure for the most resolution we usually see in a print is about 10 lp/mm. Of course young eyes looking closely at a print can see more detail. Some of us oldsters see less. Perhaps experienced LF contact printers can sense much more. The necessary CoC in an enlarged print must be divided by at least the degree of enlargement for calculating negative parameters.

I had little grasp of DOF and other lens calculations until I wrestled for hours with the formulae in Rudolf Kingslake's Lenses in Photography, 1951, with pen and paper (before the days of calculators). I'm not as good at explaining as N Dhananjay, (thanks, DJ), so I won't try to give you a simple value for CoC. One size does not fit all.

[Bill_1856]

Thanks, DJ. That was my reasoning, also, in asking the question, as well as observation of large B&W and Color landscape prints. Bill.

[cotdt]

Again, it depends. One commonly cited figure for the most resolution we usually see in a print is about 10 lp/mm. Of course young eyes looking closely at a print can see more detail. Some of us oldsters see less. Perhaps experienced LF contact printers can sense much more. The necessary CoC in an enlarged print must be divided by at least the degree of enlargement for calculating negative parameters.

10 lp/mm seems a bit low to me. It would only allow a relatively small enlargement when viewed at 25cm, for which 5 lp/mm figure is cited. Shouldn't we aim for 20 lp/mm?

[Bruce Watson]

The human eye-brain combination processes information (and not pictures - there is no little humunculus inside our heads looking at pictures transmitted through the optic channels). Color is seen as information and so a lower resolution matters less, in a sense, because the color adds a source of information. I suspect this is one reason why an image made by dye clouds can seem as sharp as an image made up of silver particles. So, I would actually suspect that you can get away with a larger CoC in color than in B&W. Cheers, DJ

Almost. Not quite. An eye-brain system processes data. It processes the data to create information. The two are different concepts and not interchangeable. Sorry to be picky, but it's important to note the distinction.

Part of building an image from the visual data is to find the relationships between the data points. For example, the eye-brain has to process a few points to decide that it's looking at a straight line. It takes even more points to decide that the line is an irregular curve. Clearly, the more data you have the easier it is to build information -- to build an image.

Now think about picking a twig out of the scene with just luminance data. It will take X number of points for your eye-brain to decide that it's looking at a gray twig against a lighter gray sky.

What happens if every point also gives you hue and saturation data? Then the twig becomes brown and the sky becomes blue. And it takes fewer points for the eye-brain to decide it's a twig against the sky. The additional hue and saturation information makes it easier to decide what you are looking at.

What this means is that in general it's easier for a person to recognize a scene when color data is available. Put another way, a person can generally recognize a scene from fewer data points when the scene is in color.

But the OP's question isn't about what it takes to build an image from image data. It's about COC and DOF. And what's in focus and what's out of focus is primarily a luminance thing. The rods in the eye deliver that luminance data, and the ability to see fine detail is in large part a measure of the spacing of the rods in the retina.

This leads one to think that what seems to be in focus, or out of focus, is the same in color or B&W. And in fact this is easy enough to verify. Construct an experiment and find out. All it takes is a scene with some depth, a camera on a tripod, some color and B&W film, and a willingness to make some exposures at varying f-stops, to make some prints at the same enlargement, and then to make some comparisons.

Sorry to ramble so, but it's Sunday night and I spent the afternoon sawing up a tree that the wind blew down, by hand (there's a reason, but don't ask). Anyway, I earned my ramble.