In article <4bsrj5$isk@mars.hyperk.com>, glh@srv.net says...
>RADIGAN@ (John Radigan) wrote:
>>What percentage of the depth of field shown on a lense's markings is >>actually "sharp"? I would imagine an object at the hyperfocal >>distance will be the sharpest thing at the film plane. I would also >>suggest that objects in front and back of this first object will >>decrease in sharpness gradually as they approach the limits of the >>depth of field as marked on the lens.
>For a nonexistent "perfect" lens, a point (one of those vanishingly >small hypothetical ones) in the plane of focus would be rendered as
>a point on your film (or at least where the lens thinks the film ought >to be.) A point which is not in the plane of focus will be rendered
>as a 'blob' whose size grows continually larger as the point gets >farther from the plane of focus. The depth-of-field limits marked
>on a lens are based on the manufacturer's judgment about how large a >'blob' you are still willing to accept as a point. Unfortunately this >depends on a great many things, including the degree of magnification >of the final image. Hence the depth-of-field markings are at most a >rough guide. If you make full-frame prints no bigger than 5x7 or 8x10, >they are likely to be a reasonable guide; if you make 20x30 prints and >stand close to them, you will probably NOT agree that the range of
>distances marked on your lens depth-of-field scale are in "reasonable" >focus.

Great answer, Gary! To it, I would add that discounting one stop
of marked DOF is useful (the lens makers are somewhat generous in
what they would call sharp), and fudging the lens focus toward
infinity is useful (technically equal resolution of near and far
subjects will make the far subjects look less sharp than the near
ones [the scales of detail are different: most distant subjects are
finely detailed {texture, with a few simple shapes}, but the near subjects {often the same material, but larger}, have complicated
shapes, but often with less important texture]).
Hope this helps (rather than confuses...)