focus test - may 22, 2005

I ran another series of tests to try to hone in on the (suspected) focus problem on my 10D. Again, I shot the test page at 45 degress, as outlined in this article and further described here. I shot each of my lenses twice from 28mm (from the right of the target) and from 70mm (from the left of the target). I shot my sigma at 125mm from the left (as it couldn't achieve focus at shorter lengths. The first shot of each pair was after I let the camera achieve focus on its own. The second of each pair was after I moved the focus ring and let it refocus again.

I haven't had time to analyze these results too closely yet, but, from what I can gather, it appears that my 10D consistenly front-focuses somewhere around 1-2 cms.

sharpness test - may 20, 2005

As with the first set of test shots below, I shot these from about 8' (70mm)* with a Canon 10D mounted on a tripod, with the timer, at 400 ISO, in Av. I used the mirror lock function this time, however, which led to sharper images and more (though still not fully) consistent results. Also, I stopped-down to 8.0 rather than 6.7 for the fourth shot on each lens, making each increment a full-stop.

* except for canon 35mm (shot at 35mm from 4') and canon 75-300 (shot from 75mm, with tripod scooched back a bit)

---

focus test - may 20, 2005

My sharpness tests led to some very interesting results, some unexpected. In the test below, the Sigma 24-70 performs exceptionally well—better, I'd admit, than even my new L lens (at 1/3 the cost). Then, in test above, it performs exceptionally poorly until f/8. The huge jump in sharpness from 5.6 to 8 can't be explained by stopping-down one stop, especially when compared to the much better results wide-open below. In other words, there's a problem. This is not the sort of inconsistency that should occur in a relatively controlled experiment, with the tripod, body, subject, lighting, etc. identical throughout.

So, I conducted a couple focus tests, as outlined in this article. Shooting with my tamron 28-75 (at 70mm f/2.8) at 45 degrees to the test sheet creates some distrurbing results. As this image demonstrates, the camera is front-focussing — here, somewhere between 1-1.5 cms.

click to enlarge

The next image shot with my 35mm prime (at f/2.0) at 45 degrees also demonstrates (albeit less dramatically) the auto focus problems. This one shows rear-focussing by almost 2 cms.

click to enlarge

Together, these tests suggest that there may be something to the rumored problems with the 10Ds.

---

sharpness test - may 19, 2005

I shot these from about 8' (70mm)* with a Canon 10D mounted on a tripod, with the timer, at 400 ISO, in Av. I did not, however, use the mirror lock-up function. This contributed to some of the inconsistencies and unexpected results, especially on the lighter lenses (more prone to wobble).

lens	f/2.0	f/2.8	f/4.0	f/5.6	f/6.7
tamron 28-75 f/2.8	—	1/500	1/250	1/125	—
canon 24-70 f/2.8L	—	1/500	1/250	1/125	—
sigma 24-70 f/2.8	—	1/500	1/250	1/125	—
canon 28-135 f/3.5-5.6 IS	—	—	1/125 (f/4.5)		1/90
sigma 70-200 f/2.8	—	1/250	1/250	1/125	—
canon 75-300 f/4-5.6 IS *	—	—	1/250	1/125	—
canon 35 f/2.0 *	1/1000	1/500	1/250	1/125	—

* except for canon 35mm (shot at 35mm from 4') and canon 75-300 (shot from 75mm, with tripod scooched back a bit)