edwardkaraa wrote:
Honestly I haven\'t noticed any, but this doesn\'t mean it doesn\'t have it. It\'s an optical phenomenon that happens to all wide aperture lenses.
You are incorrect. A fast aperture does not cause focus shift in of itself, and therefore it is not something that happens to all fast aperture lenses.
Focus shift has its roots in spherical aberration. A retrofocus lens that has a wide angle of view tends to have more issues with spherical aberration than telephoto lenses. However, wide angle lenses also tend to have more DOF, thus hiding any focus shift. So there is a sort of \"sour spot\" (to put a turn on a familiar phrase) in the wide-normal to medium telephoto range for which lens designs are not only feasibly fast, but also spherical aberration might not be corrected well.
Some lenses are intentionally designed with some degree of focus shift because it creates more flattering background blur. Lenses designed for portraiture, such as the ZE 85/1.4, are such examples. When a lens is designed with undercorrected spherical aberration, objects behind the plane of sharpest focus will render blur circles whose intensity decreases slightly from center to edge, whereas objects in front of the plane of focus will render blur circles that have a less attractive, ring-like appearance. When overcorrected, the opposite holds true.
Hence, in practice, because focus shift is generally a very small effect, lens designers will permit it to some extent, as a compromise so that the lens \"renders\" well in other ways that are deemed more important. If stopping down 1-2 stops at close focusing distances is desired, then the photographer should choose a different design.
The ZA and EF are AF lenses, so I would suspect there is some automatic compensation happening in the background.
Again, incorrect. If focus shift were noticeable in these lenses, you would still be able to observe it simply by manually focusing in 10x Live View at the fastest aperture, then stop down to shoot, and compare. Furthermore, no AF system at present incorporates any such compensation.
PS: Just realized these two also have floating elements, so the shift is probably insignificant.
A well-corrected lens effectively has no focus shift to speak of.
Mar 11, 2011 at 03:42 AM
Previous versions of wickerprints's message #9393835 « Risk of focus shift issues in new 1.4/35? »
