This is precisely why it's a good idea to test out workflows.

Okay, you stupid 5D MkII w/ Magic Lantern, listen up! When I say "use 200% normal bitrate", I expect 200% bitrate like in that test we did. You do not silently fail back to 100% during the actual shoot!

ARGH! Grr.

Note to self: Do a thorough test of enhanced bitrates and find out under what conditions the camera silently falls back to 100%. Then curse Canon for not building this functionality directly into their firmware, necessitating the need for third-party reverse-engineered kinda-buggy firmware add-ons.
 
Oh, and how did I notice this? By checking the file's bitrate and then complaining?

Nope. Visible artifacts. I'm sitting in front of the newly-loaded footage thinking "Why am I seeing compression artifacts in the out-of-focus areas?"

Then I checked the file's bitrate.

FFFFFFFFFUUUUUUUUUUUUUUU....
 
Yeah, that's the problem with bitrate hacks. Canon more-or-less set it to the highest safe amount that the equipment could handle. Not every camera is created equal, just like other computer processors an what not... Some you can over click successfully and others you can't.

Still, when it works it's amazing!
 
It's worked before with this exact same setup (even used the same card), so now I just have to figure out what I set differently. Maybe I had more of the visual doodads disabled when I did the successful test. That would reduce strain on the processor...
 
Nearly had a heart attack the first day of shooting for my film - we were recording to an Atomos Ninja and when I checked the footage, compared to the camera's native AVCHD, the Ninja ProRes looked noisy. It was only after running a number of tests I realised the Ninja was bypassing the camera's internal noise reduction and actually giving me what I wanted (footage straight from the sensor).

A few test runs through Neat Image later and I was happy to carry on filming :D
 
Back
Top