Digital Cinema Dreams
Film, Goings-on, Good stuff, Star Wars, Techie Stuff No Comments »[Update Jun 1st: there were several inaccuracies in this article now corrected, and I’ve added some too - all in bold.]
I just ordered a Canon VIXIA HD30 camcorder - this is a higher end HD consumer beast. I’m excited about it. There are many things I’ll do with it.
I’ve been asking around and doing a lot of research about HD and particularly how it may transfer to film and/or project on a big screen, and I want to say I notice a fairly sharp divide between people who insist video should never imitate or copy to film vs. people who say go for it. What’s odd to me is that folks against it seem to usually describe that as the more realistic or practical approach, or that copying video to film is only “dreaming.”
Well, gee, imagine any film maker dreaming.
I’m in no mood after writing my thesis
to cite the sources of facts I present and form my conclusions on. Suffice it to say I believe you could verify these facts.
My take is that in truth it is more practical to go digital if you can. We are in a digital cinema revolution, and physical film stock may always have its place, but the reality is that the blockades to shooting digital film which audiences don’t perceive as different after transfer to film - never mind the options to just distribute digitally increasing every year! - blockades to that quality break down steadily every year.
[Why am I speculating? With a ruler I drew a grid on a post-it note at the resolution of HD - 3 pixels an inch assuming a 30 foot screen - and filled it with alternating black-and-white squares, and looked at it from 40 feet back. There would really need to do be some image processing and projection magic with the way pixels transition into each other to make it look good. Fairly obvious “I am pixels” look at that resolution. But I need to know. I’m looking for sources that give a lot more detail on this, and I’ve also simply got to do real application visual tests on all this theory myself, somehow.]
About digital film projection, I’m going to speculate now. I don’t know how this actually plays out, this is theory, and I’d love to know of the real-world tests that certainly are playing out on these questions. But my speculation is that depending, digitally projected high-definition video could look not only anywhere from sufficiently as good as film to just as good, but better. Consider resolution available from the Red One. Here’s a picture of a 2006 model with some kind of super-exo-death-armature-skeleton-frame thing around it.
[This section had inaccuracies about the resolution of the Red One when I first wrote it - it’s fixed now.]
It shoots 2k (just over 1080 vertical pixels). That’s a bit more resolution than George Lucas thought was good enough (snobs have turned against him after the Star Wars prequels - yes, I will make that abusive statement, anyone who derides Lucas over his Star Wars prequels is a snob - I have qualms with the stories and writing on Star Wars Episodes I, II, sorta III and totally VI, but IV and V still rock the world, and I give Lucas full faith as a technological pioneer: arguably, he has single-handedly initiated the special effects revolution, and then the digital cinema revolution. Whether he simply vanishes like a good Jedi or makes it to heaven or not, before and if you pass the pearly gates, you’ll at least have to give him a hearty “thank you”.) Never mind that the Red One looks like a Star Wars Tie Fighter or something, and has a name reminiscent of Luke Skywalker flying the Death Star trenches - they have their market down - but it can record 1152 vertical pixels (or rows) at 120 frames a second, so that if projected at the same rate, it’s showing images exactly five times the frequency of standard film. I’ve read of tests going back to the 1970s demonstrating that people see a difference between 24 frames a second vs. 60 frames a second, and 120 is twice the upper range of those tests. I’d think that would probably look brilliant. Or you can do about a five hundred more rows of pixels at 60 frames a second, or again about five hundred more than that (or 2048 rows) at 30 frames a second - still a better frame rate than film. And digital projectors that do this are steadily spreading to theaters worldwide - my dear local Wynnsong has some now
Interesting math: that highest resolution mentioned (4x) has 4,096 vertical pixels, and if you divide that into 30 feet (for the typical height of a theater screen, and that link passes those paramaters into google calculator), and express that in inches, it’s about 8 pixels an inch. The math for x1080 resolution gives 3.333… pixels an inch. How does that look when you’re sitting 30 or 40 feet back (or further) from the screen? Losing detail and size for distance, they’d likely appear a lot finer and closer together I’d think. How does the density of pixels multiply across the visual range with distance? [When I first posted this I wondered if ten inches would shrink to 1, visually, so that what used to look like 3 pixels in an inch would be 30, and whether that would be enough - but no, 300 pixels an inch (or 100 times as many as 3 per inch) would correlate with the usual baseline for digital images.] That would seem like a reasonable baseline they’d go for in apparent density for “digital film”. If the visual density multiplies by about a hundred - would it? - I haven’t done that math or looked it up - but if that were the case then x2 resolution might be effectively 600 dots per inch, and x4 resolution maybe 1200 dots per inch?]
I’ve read of cinema house worries over the fickle and perhaps difficult to manage aspects of digital media, and there may be a lot of kinks and things to figure out with digital cinema along the way - but what do we expect? - it’s a brand new medium. Besides, those kinks will probably be worked out fairly fast. It took a good 60 or 70 years or so to figure out how best to technically work film, but vast improvements with digital film are advancing over stages of years, not decades! Ten years ago nobody would have thought you could buy a camera that shot at 1080 vertical pixels for under a thousand dollars. Three years ago the same camera would have cost several thousand dollars. If the trend continues the same quality camera will be available in a few years for half the price, and a camera twice as good will be available at the same price. Expanding that trend to decades it’s easy that around, say, 2020, teenagers from middle to low income families could be armed with camcorders that shoot at a resolution you can blow up to an IMAX screen - and by then there may be some bid-to-rent digital distribution network in place so that they can show their independent film at a local theatre for costs low enough that independent filmmakers of today might gasp. You can distribute for what cost? That low?
That all sounds like a dream, and it could be, but again, given the way these specific technologies have advanced in the past decade it’s easy they may advance to that stage in another decade. In my book digital cinema has to be the way motion pictures go. (I think high definition and beyond will also radically transform home entertainment.) We’ll still use film a lot, I think, especially for long-term storage because digital storage is notoriously destructible and fickle.


