Serial #'s and 10D focus

[greg_koni]

I know this has been beat to death but I have searched and not found

an answer. Is there a range of serial #'s of 10d's that are more

likely to have the focus problem? I am curious because I just ran

the test on my 10D that was purchased in 1/04. The picture was

taken with a 70-200 f2.8, wide open, at 200mm, on a tripod, without

flash. Focus point was the 16" mark. Did I screw up the test?

Does this body need an adjustment from Canon?<div></div>

[mark u]

I'd make the following comments: It's not quite clear from your picture alone what the conditions are - focus distance, angle at which the rule is to the sensor plane - is the image cropped?. Those data will define what part of the rule should be within the DOF. AF spec for the 10D is that the object focussed on should lie within the DOF - not necessarily be the centre of it.

 

Even with those points clarified, you should be aware that the AF sensors are much larger than the viewfinder indicators. Focus should be on the nearest part of the image under the AF sensor - an inherent problem for rule tests. Moreover, the AF sensors may well be offset from the locations indicated in the viewfinder. You can test for this by focussing on a narrow object against a distant background, incrementally panning. Don't forget to test in portrait orientation as well. Test all 7 focus points individually. You might find that their aim is offset to the right of the viewfinder indicators, in which case focus is more accurate than you imagine. You might as well test the accuracy of viewfinder alignment as well, by photographing a wall with post-it notes stuck in the corners of the image as seen through the viewfinder (it helps to have an assistant line up the post-it notes).

 

With those tests in hand, you are in a position to conduct better focus tests. Making sure that you align each AF sensor with your narrow object, use it to focus while having other objects in the image well away from the focus point at a range of staggered distances say at 1cm intervals (books are good for this) including the distance of your test object. Set the lens in MF to closer and further distances before using AF for the shots to test for repeatability. Also make an exposure where *you* assess focus manually rather than relying on the system to do it for you.

 

This should provide you with fairly full diagnostics to decide whether the lens and body need adjustment and to give details of the problems you have. You might also wish to repeat the final set of tests with your other lenses.

[james_castagno]

I think you would get a better idea of the focus point if the ruler was in front of you and stretched into the distance, that way if you shoot from a low angle the depth of field will be very narrow and the point of focus more obvious.

[greg_koni]

The tape in the picture was taken at approx 45* from the tape at 7'. I repeated the test using this test pattern

http://www.photo.net/learn/focustest/scale45.jpg. I shot 6 frames none of which were consistent. Since the Canon Service Center is near my fire station in Irivne, CA I dropped it off this morning. They requested that I leave both my lenses (24-70 f2.8 and 70-200 f2.8)so they could calibrate the body with them. They are also going to clean my sensor. Should have it back in 5-7 working days. I actually feel naked without my camera and a little nauseated about leaving all that equipment behind. Oh well, time will tell.