EOS M6 Mark II
- Feb 9, 2012
I seem to remember seeing DR tests showing the R5 sensor is only capable of 11 stops in video so well within CLog 1 even when shooting RAW. Most of the DR of CLog 2 when shooting RAW on the R5 is noise so totally pointless. What the R5(C) needs is a better sensor for video.
It's difficult to loosely compare DR numbers across various forums. People mean and quote different things depending on brand loyalty and agenda. 11 measured stops for the R5 seems low to me, but would fit better in a 'usable stops' context. But such statements are very subjective.
And again, Canon Log 2 doesn't 'have' or 'increase' DR of a camera. But due to the way it maps light information, it makes darker and brighter regions more easily available for grading by having more data values given to them at the time of encoding (non raw). But giving more values to dark shadows and bright highlights mean you are taking those values from somewhere else: namely the actually important part of the image that might benefit more from having better tonality.
For lighter codes such as 8-bit, long GOP compressed, it's better to encode the light information in a way that more closely resembles the end result you're aiming for (even if it's at a cost of less flexibility in post).
With today's 10-bit, 422 codecs, it starts to make sense to use flatter log curves like Canon Log 2, but more often than not, something like Canon Log 3 is a good Goldilocks acquisition format.
Once you commit to recording in high bit depth (12+ bits), high data rate raw codecs, the flattest choice—Canon Log 2—is a great choice for maximum flexibility. But since you record "raw" you can apply/change the gamma curve in post.
The DR of the R5C is surely a non issue. It will be sufficient for any project—that's for sure.