Doug Kerr
Well-known member
In the film era, and in the early parts of the digital era, depth of field reckoning was generally based on this principle: we would consider blurring due to imperfect focus "acceptable" only to the point where the blurring could be expected to be perceived by a human viewer with "normal" eyesight (in a certain viewing context, perhaps a certain print size viewed from a certain distance).
I call this approach "outlook A".
This might typically be pursued by the (somewhat empirical) selection of a COCDL value (the metric we use to quantify the limit of "acceptable" blurring) of 1/1400 of the image diagonal size at the focal plane.
In modern times, it has become common to deprecate this outlook in favor of this: we will consider blurring due to imperfect focus "acceptable" only to the point where the blurring could be expected to noticeably degrade the camera's resolution potential.
I call this approach "outlook B".
This might typically be pursued by the (somewhat empirical) selection of a COCDL value equal to the sensel pitch of the camera's sensor.
Outlook B leads to a interesting paradox.
Imagine that I am planning a photographic task, a shot of a garden scene in which important objects were two statues that would be at 9 and 11 feet from the chosen camera position.
I have two cameras I might use, both with 36 × 24 mm sensors. The older one, mostly used for "backup", has a sensel layout of 3000 × 2000, and thus a sensel pitch of 0.012 mm.
The older one, used now for most of my "serious" work, has a sensel layout of 4500 × 3000, and thus a sensel pitch of 0.008 mm.
I assume the use of an 80-mm lens with either camera.
I will assume that for some reason, I think in terms of the same desirable aperture with either camera, f/5.6.
I assume that of course I will use my newer camera, and bang the numbers into my handy depth of field calculator to be certain that the important objects at distances 9 and 11 meters will both have not over the "acceptable" blurring. Well, poo! It won't work. The calculated depth of field is too small.
Just for kicks, I do the same thing for my older camera (assuming the same setup), again determining if the important objects at distances 9 and 11 meters will both have not over the "acceptable" blurring.
Son of a gun! It will work!
Guess I had better use the older camera! Glad I kept it.
Best regards,
Doug
I call this approach "outlook A".
This might typically be pursued by the (somewhat empirical) selection of a COCDL value (the metric we use to quantify the limit of "acceptable" blurring) of 1/1400 of the image diagonal size at the focal plane.
In modern times, it has become common to deprecate this outlook in favor of this: we will consider blurring due to imperfect focus "acceptable" only to the point where the blurring could be expected to noticeably degrade the camera's resolution potential.
I call this approach "outlook B".
This might typically be pursued by the (somewhat empirical) selection of a COCDL value equal to the sensel pitch of the camera's sensor.
Outlook B leads to a interesting paradox.
Imagine that I am planning a photographic task, a shot of a garden scene in which important objects were two statues that would be at 9 and 11 feet from the chosen camera position.
I have two cameras I might use, both with 36 × 24 mm sensors. The older one, mostly used for "backup", has a sensel layout of 3000 × 2000, and thus a sensel pitch of 0.012 mm.
The older one, used now for most of my "serious" work, has a sensel layout of 4500 × 3000, and thus a sensel pitch of 0.008 mm.
I assume the use of an 80-mm lens with either camera.
I will assume that for some reason, I think in terms of the same desirable aperture with either camera, f/5.6.
I assume that of course I will use my newer camera, and bang the numbers into my handy depth of field calculator to be certain that the important objects at distances 9 and 11 meters will both have not over the "acceptable" blurring. Well, poo! It won't work. The calculated depth of field is too small.
Just for kicks, I do the same thing for my older camera (assuming the same setup), again determining if the important objects at distances 9 and 11 meters will both have not over the "acceptable" blurring.
Son of a gun! It will work!
Guess I had better use the older camera! Glad I kept it.
Best regards,
Doug
