The simplest and quickest way is to use the lens wide open and shoot a few pictures of a brick wall, camera oriented normally and camera upside down. With some lenses (maybe mostly longer lenses or extending zoom lenses) barrel play may have some influence. You can then also try taking a picture in portrait-orientation, comparing camera turned to the left versus camera turned to the right.
It is unlikely that you will notice differences with the lens stopped down (e.g. to f/4) that you do not notice with the lens wide open. I always perform the test with the lens wide open (largest aperture).
--
Slowly learning to use the Olympus OM-D E-M5.
Public pictures at
http://debra.zenfolio.com/.
Hmm, so wide open huh? I thought it made more sense to test stopped down for peak sharpness and more DoF but I guess it I'm just comparing one shot vs the other it doesn't really matter... Most issues are exacerbated fully open?
Yes, you should test wide open because that's where any deviations from perfection will show up most clearly. When testing whether you got a good copy, you don't want the inreased DoF and reduced aberrations that come with stopping down a bit to mask any problem. Of course, you may want to make a couple of shots slightly stopped down as well, just to make sure how the lens reacts to that. But you are unlikely to learn more about any asymmetry problem by doing that. Most likely, you'll just see things sharpen up a bit, primarily at the corners/edges.
You don't necessarily need a brick wall though. Pretty much any target will do as long as it has sufficient detail for you to see sharpness differences and is sufficiently far away that minor differences in distance or minor deviations from parallelity won't practically matter any more. You find a brief recipe for doing the upside-down test here:
http://www.dpreview.com/forums/post/51037446
There shouldn't be any problem using that test with the 12/2. I have successfully used it with that lens myself. But, as suggested by Paul, there may be a few cases where it is not the best bet. In that case, I'd turn to the four-corner test suggested by tt321 in this thread
here. See here for an example of this test
http://www.dpreview.com/forums/post/53894316
along with some associated discussion here
http://www.dpreview.com/forums/post/53894551
The advantages of both these tests (upside-down, four corners) over the alternative of shooting some test chart, real or improvised, are a) that they are very expedient, b) not very difficult to do right, and c) test the lens at a distance where sharpness across the frame is most practically relevant (far distance).
The test-chart method has its advantages too when you have a studio permanently set up to do that (and only that, as do the test sites) and you have already checked, double-checked, and triple-checked that the setup is perfect. But when that's not the case, it is in my experience quite error-prone. Such tests is what I used to do with new lenses until a few years ago, and I don't know how many times I have had to go back and redo it, sometimes several times, because I eventually discovered that the problem was with some minor detail of my test rather than with the lens I was testing.
