I am curious if anyone has noticed problems with peripheral AF points on the 5D3 and this lens like those mentioned at http://www.dpreview.com/forums/post/42545234?
Sorry if this has been asked previously.
Seeing no one jumping up and down I decided to test the lens. Sadly, my results are consistent with the above link.
I tested the lens on a brick wall and on some more challenging subjects as well, using aperture priority and three focal lengths: 24mm, 50mm, and 70mm. All shots were made on a tripod using a two second delay. I only tested at f/2.8. Since I was comparing only f/2.8 shots I believe it is a fair comparison. If I want to use an f/4 lens I have the 24-105. I took multiple images (at least three) of each test shot, except the center point shots where I generally only took two.
The results were consistent at the three focal lengths I tested. Using the center AF point of the 5D3 the shots are nice and sharp. They are sharp over the peripheral AF spots as well. Using a peripheral AF point the image is soft throughout. These are sample screenshots of the RAW image at 70mm in DPP at 100% showing the selected AF point.
I would send it into Tamron. Their customer service is excellent and they will fix this for you.