Given the exponentially lower amount of data handled by still cameras, it was much easier for them to eclipse 35mm film in terms of quality. Keep in mind that e

Given the exponentially lower amount of data handled by still cameras, it was much easier for them to eclipse 35mm film in terms of quality. Keep in mind that even 'burst mode' on a high-end DSLR only runs at 3-5 frames per second (not 30), and still requires a few seconds between bursts to process and write to disk. Film not only shoots at 24 frames per second (or more), but the final picture has to play on a screen that is literally two stories high. The scale is unforgiving. Things that would fly on television get suddenly ugly - especially if you're only starting with 1080 lines of resolution. And that's before your picture has gone through layers and layers of heavy post production - every one of which constitutes a round of transcoding, and thus, a source of degradation and artifacting. In other words, not only are you shooting far more frames in far less time (and needing hardware that can do sustained encoding for each one in real time) but, you're doing so at exceptionally high bit-rates, oversampling like hell so that you'll still have a clean picture even after running it through the wringer of post-production. Encoding for broadcast, DVD and the web (which is par for the course with any professional shoot) only adds an extra layer of abuse. When Michael Mann was shooting Collateral (which made heavy use of HD), there were a number of scenes where they had to shoot film, just because the HD rigs capable of supporting adequate bitrates came with umbilical cords that were simply too cumbersome for handheld setups where agility was essential. There's also been an optical issue, in that sensors for digital motion cameras have typically been much smaller than those used for digital still or motion film cameras. Even if the resolution is commensurate (which is hasn't been), the fact that the same number of scan lines are packed into a smaller area has consequences in terms of the optical effects that the camera can achieve. Specifically, a larger imagine surface allows for a shallower depth of field. Given that really good cinematographers can make masterful use of DOF, it's understandable that they don't like the constricted range that comes with smaller, digital sensors. For guys like Lucas, Fincher, and Jackson - who all do a huge amount of pre-visualization and heavy CG in post - this isn't such a big deal. They can plan around it, or deal with it after the fact. But in a more traditional setting - and for filmmakers who don't have the same hardcore ILM/WETA ethos and skill - the trade-off isn't so favorable. All that said, my sense is that this RED camera marks a tipping point, where a lot of the advantages that film has held will start to erode in earnest. The thing that's really going to drive this is the exceptionally high dynamic range (that is to say, the range of light levels the camera can handle before the picture either drops into black, or blows out in a haze of pure white). Higher performance in this area really was the trump card for film. But now that RED is not just meeting, but exceeding the long-established benchmarks, a lot of people who have stuck with film are going to ditch it in favor of a much higher ceiling. I mean, wow, 18 stops - that's hot.