Barry_Green
Moderator
I think people are reading way more into it than what's warranted with the info that's been released. For example -- Canon currently has h.264 at 50mbps in their DSLRs, but it's nowhere near as good as 21mbps of h.264 in an AVCCAM product, and it's not even as good as 35mbps of MPEG-2. There are a lot of optional components when designing an h.264 recording system. Not all implementations use all the same toolsets, and therefore you can get wildly different results even at the same bitrates. Will 360mbps of Canon I-frame "blow away" 100mbps of AVC-Intra? Or even be comparable? That remains to be seen. You can't judge a codec or its capabilities on bandwidth alone, or you'd think that 50mbps of Canon h.264 would be better than 24mbps of Sony h.264 NXCAM, and that's just absolutely not the case, NXCAM is dramatically better than the DSLR codec.*IF* the I-Frame is 3x the IBP codec *AND* the IBP codec has a comparable bit-rate to the "bog-standard" Canon H.264 implementation (As intimated) then we get 120-150 mbps, comparable to AVC-Intra, so 10 bit 4:2:2 is *JUST ABOUT* possible.
Without some actual testing and some clear technical details released, any speculation on the Canon codec implementation is just that -- speculation.