[ariahspam]

While it is good to keep track of the serial number, the essence of that thread is to see whether it is viable to buy a 16-50 or not.

So...

A slightly better approach:

Data to be filled:

-Lens serial#
-Date and source of purchase
-Bought "supposedly new" or bought used?
-Tested on which camera
-Nature of the problem (De-centered or Front/Back focus)
-Brief description of testing technique, recommended to post sample image (i.e. used tripod? did you scientifically validated that the camera was perfectly perpendicular to the plane of the image?)

Naturally, some of the data can't be filled anymore (for example, people that don't have the lens anymore, or people that can't run the test with extreme precision).

But, when we have a high enough number, we can then do some simple statistics, like:

-% of De-centering problems
-% of Front/Back focus issues
-% of Good Samples

(The posting of image samples is to avoid the subjectivity of "good").

The result of this analysis may lead to some scenarios:

Scenario A:
-Low % of De-centering problems
-High % of Front/Back focus issues

This would mean that a 16-50 is a good choice for people with K20D. For people with other cameras it would be a high risk option.

Scenario B:
-% of De-centering problems approximately the same % of Front/Back focus issues

This would mean that for people with K20D it is still a risk, although lower than for other cameras.



Then compare to the average % of de-centered lens and see if the 16-50 is on de-centering standards or not. If it is within standards, then buying a 16-50 would pose the same risk as buying another lens, so it wouldn't be an issue for people with K20D.

Yes, I'm aware that this kind of test only benefits people with K20D. But, in the long run it is going to be useful (some people will eventually upgrade to a camera that allows FF/BF correction).

As for myself, I won't buy this lens unless 2nd hand (properly tested) because I can't afford even 10% de-centering chance as I don't have the possibility of returning a lens, but for other people this additional data can be very informative when taking a decision.

I just thought that the database thread required a new direction as in being a bit more useful.