07-14-2015, 01:21 PM
Judging from the 70-200, I don't think a "good copy" of the 11-24 exists. They must all be wrong at one end or the other.
Canon's MTF shows a strong astigmatism, so by playing with the focus you should be able to fix the bad corners at 11mm. In real life images anyway - that'd be fudging it for a resolution test.
Regarding MTF; many lenses improve considerably when you add the coverglass of the sensor to the path. If a lens has undercorrected spherical aberration there is a solid chance it will improve. (And almost all do).
The actual sensor itself tends to have shit MTF towards the edges. This is a big issue with digital and requries a telecentric lens design to solve, which the 11-24 is decidedly not. True cine lenses do better in that respect.
Here's the mesaured MTF of the 135L. Agrees quite closely with canon's published chart.
https://www.dropbox.com/s/4kgroq1gz3b0lq..._.png?dl=0
Most lens models are assembled too imperfectly on average to resemble the mfg published charts, but it is possible some copies do resemble them.
Canon's MTF shows a strong astigmatism, so by playing with the focus you should be able to fix the bad corners at 11mm. In real life images anyway - that'd be fudging it for a resolution test.
Regarding MTF; many lenses improve considerably when you add the coverglass of the sensor to the path. If a lens has undercorrected spherical aberration there is a solid chance it will improve. (And almost all do).
The actual sensor itself tends to have shit MTF towards the edges. This is a big issue with digital and requries a telecentric lens design to solve, which the 11-24 is decidedly not. True cine lenses do better in that respect.
Here's the mesaured MTF of the 135L. Agrees quite closely with canon's published chart.
https://www.dropbox.com/s/4kgroq1gz3b0lq..._.png?dl=0
Most lens models are assembled too imperfectly on average to resemble the mfg published charts, but it is possible some copies do resemble them.