AFMA Telephoto/Wide Discrepancy

[Eagle Eye]

Just purchased an EF 24-70 f/4L IS and ran it through Reikan FoCal on my 5d Mark II to get an AFMA setting. 24mm came in at +10 while 70mm came in at -2. This is the largest discrepancy between wide and tele of any of my lenses. I'm planning on parking it at 0 for the best result, but was wondering if this might indicate a problem with the lens. Anyone have similar results with a 24-70 f/4L?

 

[Loswr]

That's a wider range than I've seen at 50x focal length distance. At 25x, my 16-35L II was -4 and +7 on my 7D, but W and T were closer at 50x (3 units apart) and 3-4 units apart at both 25x and 50x distance on my 5DII and 1D X.

 

[Mt Spokane Photography]

My new 35mmL was +2 on m 5D MK II and +17 on my 1D MK III, I sent it to Canon with a explanation, and they adjusted it on their reference 1D MK III. After that, it was perfect.
You can exchange it or do the same.

 

[Eagle Eye]

Thanks, gents. My mistake: I failed to cover the viewfinder during the test. Actual result was -2 at 24mm and +2 at 70mm. Dialed it in at +1.

 

[AlanF]

The 5D III stores both the W and T values, which is a plus.

 

[skitron]

FWIW, the fastest and easiest way I've figured out to set AFMA is just set up on tripod, set aperture wide open then take a picture of a window screen at about a 45 degrees horizontal off perpendicular. Zoom the shot on the camera LCD all the way and verify where the vertical "focused strip" of screen is in relation to zooming the center of the shot. Adjust AFMA to get the "focused strip" in the center. I'll itterate between focusing from further out and focusing from closer in and then "average" the results to settle on the AFMA setting used. I have found this to be very accurate and quick...the texture and color of the screen really shows the focus zone prominently and the faster the lens the better it shows. LOL, maybe I'll get torched for offering such an unscientific and gadget free approach, but the shots afterwards pass the pixel peeping tests.