I just did the same test on my Canon 10D and got some disappointing results.
I used EF20mm/2,8 and EF24-70mm/2,8L lenses. Autofocus at 31 cm and sharpest
point is at 32-33cm. ;-(
But when I focus at a newspaper on a wall everything looks very sharp. Can
someone please explain this? Can I trust my camera or am I doing something
wrong when testing
Having just bought a 10D I have spent the last few weeks looking into this problem on various web resources. Firstly I must say that I'm sure there are 10D's with focus problems, and I certainly am not claiming to be any expert on the subject - haven't even tested my camera yet!
However, It makes sense to me that a Canon engineer who has worked on the design of an Auto focus system can come up with a design for a rig capable of testing the correct operation of that system. Anyone else may be able to come up with a design which will test the system, BUT they may be testing the limits of the system, rather that it's correct operation. As a result, a camera could be returned to Canon with reported back focusing problems; Canon can't reproduce the fault, but adjust the camera to front focus, to keep the customer happy - the photographer now has a camera which appears to work on his home made test rig, but in the 'real world' is actually front focusing...
Here are a couple of thoughts I have found on the web:
What area is covered by the AF sensor? It's probably not exactly the same area displayed in the viewfinder, so it's possible that the camera is finding something to focus on, outside the focus point shown in the viewfinder.
What can the AF system detect? A thin black line on a printed scale maybe an obvious target, but could the AF be focusing on the grain of the paper in front of this point?
Is close focusing the ideal way to test AF? Does the system stop focusing when the image is sharp, or 'hunt' to find the optimum POF, taking into account the greater DOF behind the subject?
Are we testing our lenses and AF systems in a way we never could before? With no enlarger optics, etc., we automatically look at a hi res image 100% on a monitor. Consider how close you are to the monitor and how large the complete image would be. Would a 35mm camera's AF receive such close scrutiny?
I should be testing my camera this afternoon, with a very amateurish home-made test target <g>, but if I find any problems I'll re-think the test conditions before rushing the camera off to Canon.
Again, I do not claim to be any kind of expert, but I'm curious to know more about this problem, any thoughts welcome. (Also any good designs for a test rig<g>)
Regards, William Davies.
=============================================================== GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE
