I did my careful (but admittedly not scientific) lens comparison this morning. I shot several lenses off from a tripod at a wall covered with newspapers. Shot on a 5D Mark II, Controlled lighting, with all shots at 1s exposure @f/8, at a distance of 14'. All focusing was done at 100% via the EOS utility on the a 26" monitor. I have attached a view of three 100% comparisons. I tested my Canon manual focus exposure against an Auto-Focus version , and it was a bit sharper.
Sorry to the low-light shallowDOF folks, but since I always shoot at around f/8 I didn't do a comparison set wide open.
I'll list off a couple of observations:
Zeiss 50/1.4 vs. Canon 50/1.4
1: The Zeiss is much more enjoyable and usable to focus in a critical setting. The play on the rubber Canon focus ring is frustrating. However, if you shoot AF, then voila problem solved. If you are doing landscape or studio work the precision of the Zeiss play is very nice. However, if you are focusing through the 35mm viewfinder, you would never be able to discern tack-sharp focus anyway.
2: Sharpness on these two is pretty darn near identical. If anything the Zeiss is a teeny tiny bit more soft, but the Canon is a teeny tiny bit more soft at the edges, while the Zeiss is consistent throughout. By "teeny tiny" I mean really difficult to tell whether this is due to focus shift or a minute irregularity with my manual focusing. Given the amount of time and effort I spent manually focusing however, in real world terms you would get near identical results with these two.
vs. Canon 45 TS-E
When I originally switched to digital I was a 'two-exposure stitcher' with a 45 TS-E: to get that nice 4x5 aspect ratio, keep parallels straight in-camera, and I thought I'd get a bit more resolution per-landscape. The reason why I now own two prime 50s is because it dawned on me that by using a less-sharp tilt-shift lens I was losing detail and throwing away the benefits of having a $8K 21MP camera. Duh. So I decided tho throw on my 45 TS-E and compare it to the Canon 50/1.4. The image was cropped to the 50mm coverage and the Canon 50/1.4 was rezed-down to match the TS-E image size. Results:
1. That extra 5mm of coverage makes a very large difference. it feels like a 35mm lens in comparison to the 50mm
2. As expected, the 45 TS-E is a touch softer at the center, and sharpness does noticeably fall-off to the edges, but not tragically. I know from experience that this does get much worse with shifting. You get nice and blurry detail at the edges of your frame as a reward for all your effort.
vs the Canon 24-105/f Zoom lens @50mm:
This is my second 24-105, after I returned the first for seeming very soft relative to the prime 50mm. This one is better, but of course still softer than the prime. Considering what a zoom it is, it still is pretty sharp, even edge-to-edge. You would notice the different pixel-peeping and printing over 30x40, but otherwise it is close.