Mark L. Pederson
REDuser Sponsor
Nope, you're on point still. I'm exploring the lowest REDCODE RAW Compression Ratios lately for various reasons. 6K FF 3:1 looks very nice, but 5:1 is still pretty much standard fair on big productions with Weapon. The 5:1-8:1 range is still the most common for 4K-6K.
I've done endless compression tests. Each one more boring than that last cause REDCODE is so good. But I will say this -
I'm a 5:1 guy because - it is actually overkill for MOST shots - but the amount of detail can vary dramatically shot to shot and what you don't want to do is have to think .... "is this a 8:1 shot or a 5:1 shot" - you want one setting - and that setting needs to be good enough to compress your highest detail composition.
When I test compression - I will often refer to the EDGE DETECT to visually represent to me how much detail is in the composition that those wavelets needs to crush. Edge detect is pre debayer as well - so it's dead to rights precision accurate. One of the most under-used power tools on RED cameras.
So often, what you THINK is a "average" or "non-high detail image" is actually a high detail image or visa-versa - because our freekin brains are so advanced.
I have always wondered if Matt & the firmware wizards at RED could use the existing edge detect algorithms to determine if the image is "higher" detail - and then LOWER your compression ratio from your base compression setting.
How cool would that be
Auto dynamic compression setting.
Last edited: