Upload & Sell: Off
It's a pretty simple process - the goal is to eliminate fuzzy testing factors and poor test enviroments, find a CONSISTENT error in a camera body / lens combo, and use fine tune / microadjust to fix it.
1 - have an AF target of some kind (I previously had great sucess with a big piece of cardboard, folded in half so it would stand on its own, about 16x20 inches or so on the face, and I made a big 'bar code' on it with strips of black duct tape of varying width.)
2 - I stood this on the lawn (the grass will be very useful, just wait) in GOOD light - not too bright, not too dark
3 - I determined the test distance using that distance calculator that I linked to (or use the 50x focal length guideline, but really what you want to use is a very common distance that you would be using that lens at), and put the camera on a good tripod that distance from the target
4 - align camera to the same plane (sensor PARALLEL to cardboard) as the target, and put your focus point on the lower portion of the target, near where it is meeting the lawn but do not have any lawn inside the focus point, only target itself
5 - set lens aperture wide open, iso at base, single point focus (whatever focus point you want to test), and use self timer or cable release to shoot.
6 - take a few shots; before each, manually unfocus the lens and then use viewfinder AF (NOT live view) to focus on the target, then take a test shot. Unfocus, AF, shoot. Unfocus, AF, shoot
7 - review your test shots either in the LCD for gross adjustment or on the PC for fine detail. If your AF is working properly, the blades of grass in the same plane as the target should be in sharpest focus. If the camera is consistently focusing in front of the target, you have front focus (dial in + adjust). If in back, back focus (dial in - adjust). The blades of grass are just as good as the fancy computer generated test targets and focus rulers I think, they really show the plane of focus.
8 - dial in adjustment and repeat test until you can't make any further improvements, then move on to the next lens.
Typically it goes something like this for me - test/test/test - whoah - that's back focusing a lot. Dial in -15. test/test now it's front focusing a bit. Dial in -10. test/test/test still back focusing a bit. Dial in -11. test/test/test, Dial in -10 test/test/test Dial in -12 test/test/test, decide that -11 is the best of the 3, not so easy to tell anymore, but I'll leave it there. Stop testing, go out and shoot real world shots, enjoy my higher keeper rate.