
Originally Posted by
Daniel Browning
I'm glad you're getting good results. There are only three possible explanations for what you see: spectral QE changes in the CFA, updated raw converter software, and/or placebo effect. According to DxO tests, the primary change that Canon made to the CFA from the 5D2->5D3 is that the blue pixels are much less blue -- now they are 25% green and 75% blue. This improves noise performance, but noticably reduces color accuracy for subjects that have blue spectra. But perhaps the blue change is more pleasing, even though it's less accurate.
I think a more likely explanation is just newer raw converter software changes. If so, it would be possible to get the exact same zero-corrections printing from the 5D2.