Damian Luchow wrote:
1) Is rescaling to lower resolutions more labor intensive on the hardware than recording in raw at 2.5k? and if so, why do cameras offer the option of shooting with higher framerates on a lower resolution?
Rescaling requires computation. It would be an extra layer of calculations in the pipeline.
2) Is shooting in prores more work on the hardware than shooting in raw?
It's a lossy perceptual codec. That means math, which will eat computational power. The tradeoff is do more computing up front, or consume more bandwidth up front.
3) Please correct me if im totally not right here, but i though a lighter program or codec generally means its coded more efficiently and requires less work, thus lightening the workload of the hardware, with the codec not necessarily increasing the compression?
"Coded more efficiently" is basically stating that the current code is crap. Odds are that's not the case; this isn't the IT industry, which is dominated by garbage code written by the cheapest dumbass or the best brown-noser. (This is based on far too much experience in the industry. Forgive the rant, it's the reason that I'm changing careers to being a cinematographer, and let's leave it at that. Should you prefer more, just be informed that Dilber isn't fictional, it's sugar-coated.)
When it comes to compression, there's no free lunch. Most of the best video codecs are perceptual and asymmetric. That basically means that they analyze the data stream, look for the details that the viewers won't be able to see, and don't save them. Increasing the compression ratio means lowering the threshold of what details to discard from the data stream. It takes considerably more computing power to encode h.264 than to decode it, for example.
A "lighter" codec, one that requires less computing power, is most likely also a codec that doesn't do as much compression, or that like most lossless codecs, doesn't do any analysis to determine what the viewer may or may not be able to see. The down side is that it will as a result require a higher data rate to save the video.
Its clear as day that Blackmagic's main priority was to make an awesometacular camera and get it out into the public's hands based on its current features, and that the engineers have so many other priorities and things to work on/out before they even consider higher frame rates.
That's the most logical thing for BMD to do. They're a newcomer in an industry where their primary competition is film, Arri Alexa, Red Epic, and Sony CineAlta. This isn't a market that tolerates crap; if you start selling a camera that flakes out or overheats regularly, you're done. Black Magic targeted a price point and made the compromises that they needed to in order to hit it without sacrificing image quality, and their belief was that "image quality" was driven primarily by having accurate color and a huge dynamic range.
The stuff from Andrew Julian alone shows that BMD got it right, even though they still have some bugs to work out with the audio side of things, but at least they got the hardware right, and they can fix the software.
If you can't afford a Red and you need a high frame rate, look into a Hero 3 Black.
This is a good time to be a cinematographer.
I'm going to see if I can arrange a trip to climb Mount Adams, and if that works out I'll be dragging along the BMCC. My only frustration will be having to choose between the BMCC and the Arca-Swiss for weight considerations, unless I can find enough sherpas... I mean, crew... to join in.
Sat Feb 16, 2013 3:03 am
