Kholi wrote:How about those freezes in the middle of a good take, reboot scares, and half of the footage coming up as nothing but a green block?
THAT was a scary camera to use, but we still used it because there was nothing else at that price that gave up that quality.
Haha.. absolutely.. oh gosh, I don't want to remember, really... sometimes everything was fine and we shared a sigh of relief but sometimes it was green and/or scrambled blockyness galore or other errors.. esp tough on those commercial shoots when you had clients breathing down your neck.
So compared to that my BM experience until now was a walk in the park, thats true
Kholi wrote:I still have problems with EPIC... when I actually have to use it. Entire shooting day saw the right half of the image shifted down TWO pixels, not something very easily noticed until you get into post and a models nose shifts downward as she moves right to left.
Thankfully I was spared the pleasure of this defect/bug yet... although some of the rental OneMX/Epic/Scarlet over here in Germany are quite beat up to say the least... always needed to plan for at least a few hours of testing before you "go live".
Regarding AJA and linear.. that's quite interesting. The raw sensor data itself without fiddling with any of CMOSIS' "DR enhancing techniques" (knee like curves) is quite linear. So maybe they are more or less keeping that original data without applying a lot of lin/log conversions, color science etc pp... of course they have to do some of that.
Maybe log just doesn't make much sense with the CMOSIS "out of the box"... but I'm still fascinated by the findings of the Apertus guys regarding the sensors DR enhancement possibilities.
I wish these companies all would be a little more transparent in regard to what they do to the image. Of course the secrecy is partly understandable, on the other hand they now all seem to share the same CMOSIS chip so it's not like they designed their own super secret spaceship.