Originally posted by dms I think some of the comments about DOF results (tables/programs based on a consistent set of assumptions) are "throwing the baby out w/ the bath water."
The basic idea is to have a "method" ("a model" or "set of assumptions") that attempts to account for all the important parameters--and if you find it is not adequate--you adjust the model. e.g., you use the DOF for 1 stop more open or 1 stop more closed.
The alternative is to not have a method and consider all the variables each time--or just give up and use nothing. The former is impossible for ordinary mortals and the latter means no control.
+1
This is the best response to the whole depth of field discussion.
Fundamentally, the real problem with all DOF calculators is that they they all tie back to the concept that the image being viewed is based upon blowing the full frame up to an 8x10 inch print, viewed from reading distance of about 1 foot and having a point if light represented by a circle of 1/100 of an inch.
That translates back to the "circle of confusion" being 20 microns on an APS-C sensor or 30 microns on full frame.
But as soon as we crop, pixel peep or what ever, that concept is lost, so depth of field calculations don't really matter all that much, with a little work as you suggest, you establish your own rules, based upon the size you enlarge to, the amount of cropping etc.
What people often forget, with respect to the whole argument about depth of infield and it's relationship to hokeh, is the relative magnification of the background, when you change the shooting distance with a shorter lens on APS-C format. This changes the relative magnification of foreground to background, so the out of focus elements are smaller on APS-C than with full frame, making the hokeh less pleasing because you see what appear to be more defined and smaller noobs as opposed to the gradual merging of larger less apparent shapes into each other.
The larger the format the smaller the difference in magnification of subject and background and therefor the more pleasing hokeh, and more apparent separation if foreground to background.
It is all about understanding your format, not playing with little calculators