Also... before you go through the bother to send this to Canon, make sure you've carefully tested it.
That means:
1) Camera on tripod (never hand-hold a camera for a focus test because if you sway or lean after it locks focus then you just invalidated the test.)
2) Pick a single AF point and make sure it's on your target. Don't let it use just any AF point. It'll gravitate to the AF point that can lock focus at the nearest focus distance (which could be mistaken for front-focus even though it's working exactly as intended).
3) Use a real focus test target - not a typical subject. You can use a commercial focus chart, or you can download one from the internet, print it out, follow the instructions, and do the test that way. Here's one I like: http://regex.info/blog/photo-tech/focus-chart
When testing, I like to run the focus all the way in (closest focus distance) so that the camera is forced to move the focus on the lens. Repeat a half dozen times and take notes of which shots started with focus all the way "in". THEN... manually run the focus all the way out (to infinity) and repeat that test a half-dozen times, while taking careful note of which frames were tested with the focus starting at infinity.
Import the images to your computer and analyze:
(a) is it missing focus ALWAYS or just sometimes?
(b) does it always miss focus in the same direction (e.g. always front focus? or random?)
(c) does it miss focus the same regardless of whether you started with focus "in" vs. with focus "out"? If focus misses different in one direction than in the other it could indicate some mechanical gear backlash (loose gears) in your lens.
Tim Campbell
5D III, 5D IV, 60Da
