07-06-2012, 09:29 AM
[quote name='Bobcity' timestamp='1341531753' post='19331']
when you say it might be a bad sample, what exactly does that mean? How do I know? Does a bad sample mean it should be sent to Sigma? Is there a way to test it myself?[/quote]
Well, if you start with the manufactoring process, you will see, that albeit all
lenses of a type are produced from the same elements, but even those elements are
not exactly identical. They are just "very similar" and follow their specification within
reasonable limits. ... Now putting together such elements that are nearly the same,
would already lead to lenses that are only nearly the same, but within the assembly
process, there are tolerances as well. So in the end, of a batch of 1000 lenses, you
might have a few where all the tolerances add up in one direction such that the result
is outside the desired limits.
There is a variety of different defects a lens can have, some of them easy to detect like
unusual amounts of dust and dirt in a lens, uneven coating, scratches ands the like. The
most prominent one is "decentering" ... means the lenselements within a lens are not
in the positions they should be ... one or more are somehow out of their theoretically
optimal centered position ... shifted, rotated, tilted into a different position.
Klaus has, somewhere here in the forum, a link to a relatively easy decentering test.
But you will also find testdescriptions if you google for "how to test decentering".
My method ist as follows:
- Use a tripod.
- Test with focus set to infinity.
- Test against an object, that has good contrast and shows some fine detail.
- The object needs to be in a distance that allows infinity focus.
- Either set focus to infinity manually, or focus on the object, then switch to manual focus.
- Set your aperture and exposuretime manually, or at least use a fixed aperture setting.
- Take 5 shots ... one shows the testobject in the center, the other 4 show the test object in the 4 edges.
the switch to MF was done to have all 5 shots done with the same focus-setting.
- the image showing the object in the center will very likely show the highest sharpness.
- the corners will only come near this sharpness when you stop down. The important thing
is, that the corners show an equal loss of sharpness (and eventuelly contrast). If one corner
is much better or worse than the others, this is an indication for a centering defect.
With that, you should be able to see if there is something severly wrong with your lens.
Rainer
when you say it might be a bad sample, what exactly does that mean? How do I know? Does a bad sample mean it should be sent to Sigma? Is there a way to test it myself?[/quote]
Well, if you start with the manufactoring process, you will see, that albeit all
lenses of a type are produced from the same elements, but even those elements are
not exactly identical. They are just "very similar" and follow their specification within
reasonable limits. ... Now putting together such elements that are nearly the same,
would already lead to lenses that are only nearly the same, but within the assembly
process, there are tolerances as well. So in the end, of a batch of 1000 lenses, you
might have a few where all the tolerances add up in one direction such that the result
is outside the desired limits.
There is a variety of different defects a lens can have, some of them easy to detect like
unusual amounts of dust and dirt in a lens, uneven coating, scratches ands the like. The
most prominent one is "decentering" ... means the lenselements within a lens are not
in the positions they should be ... one or more are somehow out of their theoretically
optimal centered position ... shifted, rotated, tilted into a different position.
Klaus has, somewhere here in the forum, a link to a relatively easy decentering test.
But you will also find testdescriptions if you google for "how to test decentering".
My method ist as follows:
- Use a tripod.
- Test with focus set to infinity.
- Test against an object, that has good contrast and shows some fine detail.
- The object needs to be in a distance that allows infinity focus.
- Either set focus to infinity manually, or focus on the object, then switch to manual focus.
- Set your aperture and exposuretime manually, or at least use a fixed aperture setting.
- Take 5 shots ... one shows the testobject in the center, the other 4 show the test object in the 4 edges.
the switch to MF was done to have all 5 shots done with the same focus-setting.
- the image showing the object in the center will very likely show the highest sharpness.
- the corners will only come near this sharpness when you stop down. The important thing
is, that the corners show an equal loss of sharpness (and eventuelly contrast). If one corner
is much better or worse than the others, this is an indication for a centering defect.
With that, you should be able to see if there is something severly wrong with your lens.
Rainer