D200 expected focus accuracy

Hello, yesterday I decided to check my D200 and the Tamron 17-50 for focus

accuracy. I put the camera on a tripod, and a small box rather rich in detail

on a focusing rail 1 mt away, so that I could focus and then make a series of

shots precisely moving the subject +10 to -10 mm from the plan of focus

detected by the camera. Used the Central sensor, AF-C mode activated by my

thumb (my default setting) F2.8 for all the exposures, camera obviously did not

refocus between exposures. I repeated the test several times and also at a

different distance: 50 cm, and the properly focused image was always different

from the one focused by the camera, the real plane of focus was closer by 4 to

10 mm, not constant. Then I repeated the same with my 50 1.4 AFD, always at

f2.8, and in this case the real plan of focus in different series went from

behind to in front of the plan where the camera focused (!), by a similar

amount. To summarize, with the two lenses the camera is able to focus within

about 10 mm from the real plane of focus, at f2.8, 1 mt from the subject. What

do the experts say? Should I expect more precision? Or are these results within

expected tolerances? Needless to say with both lenses at f2.8 there is quite a

difference from the optimal image to one misfocused even by only 6 mm. Thank

you for your insights, Marco

Try the exact same thing using AF-S. In AF-C mode, the camera and the lens are working as team, in a somewhat predictive mode. The camera is telling the lens where it needs to be, and the lens is expected to follow those instructions - without a continual back-and-forth between the camera and the lens to refine the expected focus. If ALL of your shots, with all lenses, were similarly off, then you'd have the camera to talk about. If one lens is different than another, then it comes down to how well the lens is calibrated.

 

In AF-S mode, you should experience some different behavior (since the camera isn't in a mode that's all about speed). I'm sure that you'll get a more lucid explanation from someone else... but you might also try using manual focus, and using the camera's in-focus indicator (in the finder) to see if the AF sensors are actually off.

I find these sorts of tests fraught with uncertainty and difficult to even replicate. Take the camera/lens combo and aim it at someone's eyes and examine the shot. My D200 nails the focus in the real world tests, not so great in the "attempts to quantify" tests. As Matt says, for the type of shot you were attempting, AF-S is the preferred focusing mode over AF-C.

Thank you Matt and Anthony. I decided to make this test as I can't get reliable focus in the real shots, for example focusing at someone's eyes as suggested. It's entirely possible that the problem is in the way I use the camera. With static subjects I usually leave the camera on AF-C as I think the camera should not find movement, focus and quit thinking. I will try the test with AF-S, let's hope its the right way. I am not a very experienced AF user, always used MF gear until recently...

If you get the same results in AF-S using Focus priority (a2 in the menu), then the camera should be sent to Nikon for servicing.