I've got the 1.2 and it's sharp, the DOF is just super thin and the contrast drops off enough to where it might *look* like it is not resolving very well. On celluloid it needs a contrasty film/development combo to look as sharp as the slower lenses.
The 1.8, 1.5 and 1.4 are definitely better value for your dollar, given that the 1.2 can go for $500 and up these days.
That test on Flickr that goamules posted is one of the things that made me seek out the 1.5. While the 1.8 is a great, and much more common, standard lens, the Sonnar look is pretty distinctive; if you prefer it over the somewhat more modern look of the double-gauss 1.8 or 1.4, as I do, the decision sort of makes itself.
I traded the lens even for a Contax II with a collapsible Sonnar (f:2, uncoated) and got pictures I liked much better under the same conditions with the same film.
They use the glass catalogue from the 50's the type IV summicron the glass catalogue from the 70's. The later lens will have better microcontrast on a MTF machine, if you tried, e.g. it is better then the type III summicron on a MTF machine.
The /1.8 will out perform the /1.2 and /1.5 detectable with slow film and solid tripod. but the /1.4 will be pretty good and very close to the /1.8 especially when both are at f/2... otherwise the difference is size and weight.
Comparing the Sonnar types to the double Gauss types really isn't fair. The Sonnars evolved as a large-aperture lens in the early 1930s, when lens coatings were not available. They gave a very good performance with a minimum of internal surfaces -4- compared to the dG, which has a minimum of 6 internal surfaces. The dG type had somewhat better performance at maximum aperture, but the price paid was flare. The Gauss lenses really had to wait for coatings to do their job well.
Canon LTM lenses - http://www.canon.com/camera-museum/c...s/s/50-85.html
But seriously . . . By my calculations the best I can observe on my monitor is ~90 dpi or ~3.5 lpmm. Viewing web images for sharpness at that resolution is not very useful.
Can someone provide actual "test data", such as lpmm resolution data (center and edge) for the Canon lenses in question? That would be helpful.
I have a J-8 as well, I've made only a few photos with it, but I'm very pleased with the results. Very similar to the Sonnar it's a clone of.:)