I took a total of 10 pictures: the widest apertures of F/1.8, F/2.0 and F/2.2; then from F/2.8 to F/22 in full EV increments. Exposure metering (matrix, -.3 EV) resulted in shutter speeds ranging from 1/13s to 13 seconds.

At left you can see the full frame at F/2.0, reduced and with sample location marked; clicking on this will bring up the full-size original.

Below you can see 1:1 samples from the sharpest area in the frame, the AF target location. The AF was well-behaved, except for the first frame, where it seems to have focused not on the teeth, but rather on the red bandana on the far side of the face. (I forgot to use my normal trick of switching to MF and autofocusing on demand, using the [AEL/AFL] button.)

Central sharpness considerably improves at F/2.2 and reaches a low-sloped plateau at F/2.8. It still keeps going up, but slowly.

The difference between F/5.6 and F/8 is close to none, and going to F/11 we see first signs of degradation due to diffraction effects.

This makes sense: diffraction limit at F/16 is, as I remember, 100 line pairs per millimeter, and the dependency is linear — so at F/8 it will be 200 LPPM. With the image being 13 mm tall this is 2600 LPPH (pairs per picture height) which is a very good value, probably a tad lower than this camera/lens combination has.

It is only at F/16 when the adverse effects of stepping down are clearly visible to the naked eye, and at F/22 things look quite bad: even the contrast suffers.

Oh, but you wanted to see by yourself. Sorry, go ahead.

Well, this was to be expected; this is one of two, maybe three, best μFT lenses. I really did this project to get a pretext for buying the props.

I also used the same images to show how the depth of field depends on aperture. Here is a slightly altered version of that demonstration, a collage of vertical strips from ten frames:

This is just a thumbnail; you may want to see the full-size image (45MP) or an economy version (4 MP).