>I have a few friends that adjust their AF Fine Tune throughout >the year based on the subjects and exepected distances. For >example, with some lenses, they adjust for the long side >during grizzly bear times in the spring. Then, when the >songbirds start moving through and the are shooting closer, >they readjust for the best results.
One thing I've been doing is saving the Focal results (the graphs) and then I will put the graphs together into one composite image with near, middle and far, and for zooms all the way out, and all the way in.
Putting these all one one page is a great way to look for the happy middle -- if you just draw a line down them you can see where on the falloff graph each extreme falls, and get a feel for the price you pay for any given compromise. You could also use this as a reference, e.g. if you were going out to shoot long, you could go to the distant/zoomed-in graph, and just adjust to that value; and then the reverse later.
I do not find myself changing as a shoot normally, as I already have too many variables I need to adjust. But on certain lenses for certain activities I might.
Here's an example (happens to be the D4 with 85/1.4G not a D800) but
Frankly it's a case one could argue should go back to Nikon, but I suspect it's not really fixable. It also has a bit of focus shift as well (which is a well known feature of it). Since I mostly shoot it closer, I just leave it set on the positive side, but if I were going to use it for something long while wide open I'd probably either use live view or tweak the AF fine tune, since now I know it's going to be off quite a bit.