These tests were shot during the Hannover Hands-on workshops in August 2012
I'd like to start by thanking the organisers and their support staff as well as my students for making this work....
These are in no way meant to be any kind of definitive test
Alexa Studio recording RAW to a Codex recorder
RED Epic recording RAW at 5:1 to internal SSD (HDR +2)
Sony F65 recording RAW to SR docked recorder
All cameras set to 5600K and no tint, exposure was set with a meter rated at ISO 800
We simply shot the same move at several exposures and picked the exposure that seemed to cope with all lighting conditions best, in all cases I have used ISO 800.
Any size differences are due to us scrambling to compensate for the fact that the Epic sensor is a non-standard size.
Each frame has no grading adjustments at all.
Epic files were rendered to DPX using Magic motion in RedCineX using RedColor3 and RedLogFilm
F65 files were rendered to DPX using Sony's F65 RAW viewer using Sgamut and SLog2
Arri files were rendered to DPX using SpeedGrade CS6 using the inherent logC
This work-flow was the best compromise to get the best out of all systems, whilst we had intended to use Resolve 9 for this test we found that the lack of something like Magic Motion meant blending 2 exposures together and this lead to motion artefacts. I later discovered that I could use MM in Resolve but needed to use RedcineX to establish my base value with it.
The DPX log files do not necessarily manage to hold all the shadow or highlight detail that the RAW file is capable of, this is particularly the case with the F65.
In a change from my normal process there are thumbnails and below those there are links links for DPX files at 1920 * 1080 then RAW files and finally data files for those formats that need them.
There are moving versions of one lights here and best grade here