rscheffler
Offline
Image Upload: Off
|
Dawei Ye wrote:
I just took the plunge and purchased the 14L II (got a good deal, price was 2250 AUD on special, around the cost of a 100-400L in Australia)
The lens is weird. It's almost as if it has field curvature or something. DOF is very think despite the wide focal length
It is sharp at f/2.8, but DOF is very thin that it is difficult to get corner to corner sharpness (unless I have a misalinged element)
At f/2.8 the centre is VERY sharp. As sharp if not sharper than the 35L and 85L II. However the 35L and 85LII are way better at equivalent apertures. The 14L II is a bit disappointing at f/8 compared to the 35L, it is not soft, but just not mindblowingly sharp. The 35L I have is ridiculously sharp at f/8 corner to corner.
Still, I'm quite happy with the results.
The lens when in focus is sharp wide open right to the corner except for about a centimeter or so as demonstrated above in the f/2.8 Crop (the left side of the Pelican logo. I focused on the centre of the Pelican logo)
I won't show centre because it is perfect.
Here's where it gets weird. I need to more testing on this though
This shot is at f/8. As seen above it is sharp corner to corner at f/8. But look at Crop D, it is horrible, but can be explained by DOF limitations, YET Crop B is even more outside DOF but way sharper. The lens is very weird, items that are close to the camera are extremely sharp, even when the camera is focused at infinity, yet items closer to infinity on the edges are OOF. It almost defies the laws of physics.
^PLEASE NOTE WITH CROP 4, IT IS MYSTERIOUS WHY THIS HAS HAPPENED BECAUSE AS YOU CAN SEE IN THE CAMERA STORE TEST IT IS TACK SHARP CORNER TO CORNER AT f/8
^That one is at f/8. The one below is at f/2.8 and Crop C is sharper  , which suggests a DOF issue rather than sharpness issue
Here's the shot at f/2.8 (almost the same)
^You'd think that B would be more OOF than C. Maybe user error...
It's possible you're not seeing the smearing in crops 4/C in 3/B because 3/B is along the horizontal edge of the sensor which is not nearly as close to the extreme edge of the imaging circle, where image quality tends to fall apart, as the vertical edge from which 4/C are taken. Also, if there is field curvature (and it appears to be likely), then the railing is considerably closer than the object in the 4/C crops and benefits from the curvature. I'm curious how the right edge of the frame looks where the people are standing in comparison to the 4/C crops.
As someone else suggested, you really need to shoot a flat subject to determine how significant the field curvature is with the lens.
I haven't tried the 14II, so can't say if field curvature is normally an issue, but I own the 16-35II and notice significant field curvature with it. I can't photograph a flat wall at f/2.8 (or f/8) and expect it to hold critical focus to the edges. In this situation the lens is very frustrating. On the flip side I can use it to my benefit by placing close subjects around the frame edge and distance subjects in the center and all will appear in focus at a wide aperture.
You wrote: "The lens is very weird, items that are close to the camera are extremely sharp, even when the camera is focused at infinity, yet items closer to infinity on the edges are OOF. It almost defies the laws of physics."
This sounds a lot like field curvature to me, assuming the closer objects are located around the edge of the image. It could be worth focusing on a distant subject in the 4/C area using live view, then comparing how the center of the frame looks. I tried this with the 16-35 II wide open for a brick wall test and I can get the edges sharp, but then the center goes soft.
If edge to edge sharpness of a flat plane wide open is critical then I think the only Canon super-wide to successfully tackle this challenge is the new TS-E 17. If the 17 isn't wide enough, remember that you can stitch exposures shifted to the opposite extremes to capture an even wider angle of view.
Ron
|
