CheshireCat wrote in post #18284883
Uhm, close second is the Zeiss 135 APO. Has a very similar rendering, although the FL is quite different.
The 135L is a great lens, but doesn't come even close to the best apochromatic glass.
Actually, although the Sony is very clearly has better overall IQ, IMHO comparing these two, coming "even close", is exactly what the Canon does. Obviously this is a matter of opinion in use, but looking at measurements and numbers, given that the Sony is over double the cost, the Canon does indeed come very close.
One of the factors I have often compared the 135mm L favorably with the 200mm f/1.8L (and now f/2L) is it's autofocus. I'd weight that at about 2/5ths the reason why it is the best bang for your buck competitor for the ultra pricey 200mm options. Given that what I use this lens for is people, usually of the moving persuasion, AF is very important to me. So 2/3rds aside, for ME it would be more like 9/10 
Obviously the SONY can't compete there, so back to IQ.
The two lens' distortion measurements are the same, the light fall off is nearly the same (slight edge to the Canon) The Sony edges the out the Canon on light transmission, CA, and has it's biggest lead in sharpness and MTF wide open. (of course, wide open is where we shoot them, so the fact that Canon ties up by f/2.8 is a dubious achievement)
Given that pretty much everyone posting in this thread has said that for a SIGMA sharpness does not mean a better lens, I am unsure why it would be weighted so heavily in favor of a Sony, given it's double price.
SIDENOTE:
I looked at numbers from 3 sources, one of them is DxO . DXO is often a concern IMHO as the numbers can be really contrary to our own experiences. Case in point, they are saying the 135mmL has more measurable CA than the 85mm f/1.8. I find that of all things to be very hard to believe.