I think Canon should check your camera. I have seen 10-15 posts from you similar to this one with other lenses also.
And I don't think your way of testing lenses (both this or your Sigma 35) for sharpness is good at all. Every test shot that I have seen is an extreme close up of a small child and often wide open like here! How do you know the child didn't move two cm when you where shooting? Take test shots at a more normal distance and with a still subject. Maybe use a tripod also. Because you will move a little bit when shooting.
I would do it in another way than you. But if every lens you test isn't sharp enough for you. Maybe you need Canon to adjust your 5D3
Agree with Lars... if you really want to test a lens (not that I'd encourage endless testing) then set up a more controlled situation with a static camera and a static subject. I've not done any testing with my new 24-70 II but I've not noticed any particular issues.
Also, have you made any lens microadjustments on your 5D3? It might be that the 70mm end needs some adjustment dialed in to nail focus
Mine is also sharp at all FL settings; easily as good as my 70-200mm Mark II at the 70mm setting.
I believe Lars may be right; it is possible that either you or your target may have moved some. Depending on the lighting conditions, the resulting shutter speed may not be sufficient to minimize motion, either camera/lens combo motion or model motion. No offense and to be frank, the sample posted of the 50mm setting doesn't seem to be critically sharp either. Sharper alright but..... To check for sharpness, pure sharpness that is, I am afraid you have to use a static target and put your gear on a sturdy tripod.
PhilDrinkwater wrote:
Actually I'm on my phone and just zoomed into the photo. The second one looks like motion blur, not a sharpness issue.
Agreed with Phil.
IMHO the shot at 70mm is slightly motion-blurred, it doesn't seem to show any issue with your lens.
My 24-70 II is slightly softer indeed at 70mm but this is because up to 50mm it is unbelievably sharp while at the longest setting it is only wonderfully sharp
As Lars said try the lens on still subjects, preferably on tripod.
I should have been clear that I'm posting 100% crops. These aren't really extreme close-ups.
A shutter speed of 1/160 at 70mm should be enough to stop motion blur, should it not? I was under the impression that 1 / 2x focal length with a FF camera would be sufficient. And the first picture at 50mm is 1/640.
I have calibrated the 24-70 with FoCal, and it returned a zero adjustment at 24mm and a -1 at 70mm.
I agree that a tripod and static subject would be a more controlled test. But I'm also interested in seeing how a lens performs in my normal shooting conditions. My 50/1.4 is very sharp above f/1.8 in normal shooting using the same technique and settings I used for these pictures. The same is true for my 40/2.8. The lenses I've had issues with are the Sigma 35/1.4 (obviously front-focusing by a significant amount), Sigma 85/1.4 (erratic focus) and possibly the 24-70 II (*maybe* slightly soft wide open, especially at 70mm).
Today I will try a more controlled test and see what I can learn.
And yes, perhaps I should send the 5D3 into Canon — with the 24-70.
Dof might be the issue. In the 24mm shot, everything is within DOF. In the 70mm shot DOF at f2.8 is very shallow. And it appears like the focus was closer to the front edge of the cap.
Are these 100% crops? Let's see a photo of something like grass where MA doesn't matter because there's always something in focus. Sharpness and MA are two very different issues that should be immediately divided.
badlydrawnboy wrote:
Okay, I shot some pictures of a tree in my backyard to try to separate the sharpness and MA issue.
One problem with focusing on the front of something is that if the focus is even further in front, you wouldn't know. This is why I suggested grass, because there will always be something in the foreground and the background. Nature's test chart.
