The DCP is just a digital print (what movie theaters play for digital releases instead of the reels of film that go with analog releases). It's not really about resolution per se, it's just a term for the whole package that the theater gets.
This all works due to proprietary hardware - the modifications to the 4k cinema projectors that enable decryption almost at the lens/imaging chip can cost as much as the projector itself, doubling the price of the projector (last I heard from around 15k to 30k USD).
With a more sane key distribution (no master keys in equipment, like with dvds) - I don't believe DCP is likely to be cracked in the same sense CSS was: probably the more likely scenario is that some insider leak some of the distributors private keys which would allow certain releases to be cracked - but would likely also trigger key roll over.
Also worth nothing that the DCP format is pretty nice, has optional encryption - and you can make your own. With the right contacts you can view your own footage on on a cinema screen :-)
See eg:
a typical cinema release is a
pretty big file - frame-by-frame
encoded as jpg2000 in 4k. Often
shipped on hard drives, or
downloaded via dedicated fiber
lines.
This all works due to proprietary
hardware - the modifications to the
4k cinema projectors that enable
decryption almost at the lens/imaging
chip can cost as much as the projector
itself, doubling the price of the
projector (last I heard from around
15k to 30k USD).
There'd be no real need to tamper with proprietary hardware or intrude on the chasis of the projector. With reasonable equipment, it's likely that a person could reproduce a re-recording of a movie that is very-nearly indistinguishable from the factory source, perhaps except for negligible artifacts that might irritate only the holiest of true believers.
In a way it's shocking that huge cinema screens ONLY project 4k. When we had film instead of digital projectors, we actually had a much higher resolution on screen. It's a shame we replaced it with an inferior technology.
Similarly, a highly priced vinyl player might have slightly better sound quality than a cheap CD player.
But for most people, digital Cinema and CDs represent a vastly superior experience.
It's not because a technology is newer that it is necessarily better. CRT still holds some advantages over LCDs, Paper still holds advantages over e-ink, and so on. Of course Digital photography wins by a large amount for cost and convenience, but that's not the whole picture.
Anyways, most of old stock (print, negative, ip) have grain such that over cca. 2.5k won't yield anything better. However, kodak and fuji stock from mid to late eighties definitwly is worth capturing at higher detail. Highest-detailed stock from nineties we've measured to have spatial rwsolution of around 8.5k and that's about as high as you will get (rare) out of 35mm.
tl;dr; depends on the stock
edit: Here's previous mini discussion regarding telecines, if anyone's interested https://news.ycombinator.com/item?id=11377678
There are enough good 4k releases of classic movies to see this is not the case. You can compare the 4k and HD release (which is just under the resolution you provided).
I regularly scan Vision 3 film (movie stock) and Velvia 50 and Velvia 100 (E6). I scan it professionally at 10 Mpix, but when I have a good picture, I pay big bucks to have it drum scanned at 100 Mpix. Yes, it's not 100 Mpix, there's about 15-25 Mpix worth of detail there, but it's much, much more detail than the 10 Mpix scan.
I've seen both, digital is better.
Film has grain, which is a type of non-chroma signal-dependent noise that is not unpleasant like chromatic signal-independent sensor noise or shot noise you get digitally in photography (not in movies, they have good lighting there).
In fact, grain is actually useful, as it's aesthetically pleasing in small amounts, can increase apparent sharpness. In fact it's so useful digital creations fake back in post-production a small amount of grain.
> The resolution is not as high as you are imagining.
The resolution of 4k projection is not as high as you are imagining either. It's just 4096x2160 lines (often less, depending on aspect ratio) or 8.8 megapixels.
In the best circumstances you can get 25Mpix of useful data from a 35mm frame shot on low grain stocks like 5203. In the worst circumstances you get 10Mpix. This is 35mm, The Hateful Eight was shot on 65mm stock, which has a camera frame surface are of 3.44 times the camera surface of 35mm film.
Either way, the resolution of 4k digital cinema is fine, I miss more the old contrast and gamut.
And no, it's not aesthetically pleasing in any amount. Directors that add it back artificially should be criticized. It looks like there is a swarm of gnats flying across the screen, and unless there are a ton of bugs in the scene it doesn't belong.
> This is 35mm, The Hateful Eight was shot on 65mm stock, which has a camera frame surface are of 3.44 times the camera surface of 35mm film.
It won't stay that way by the time you watch it, it will have been copied multiple times, and it's on 35mm again. The effective resolution as seen in the actual theater does not beat 4k, and you have additional noise which lowers the resolution even more.
The main thing though is that noise is worse than lack of resolution. Depending on how far you are from the screen more resolution is meaningless - but noise (like film grain) is always a problem.
> I miss more the old contrast and gamut.
I suspect you are missing what never was. Digital is better on both.
No, it's not on 35mm again. I've watched the 70mm projection of The Hateful Eight. At no intermediary point the film was on 35mm.
The intermediate film stocks and the print stock are very special low-sensitivity stocks (Kodak 2383 has ISO 1.6, Kodak 5302 has ISO 6) that have no discernible grain whatsoever. You can't see the Kodak 2383 grain with a grain focuser. All the grain comes from the original negative. You do lose some spatial resolution with multiple copies, but you don't add grain. I'll note that modern print film like Kodak 2383 can resolve 550 lines per mm, and contact printing has a minimal loss of quality due to imperfect optics. Granted, a showprint will always be better than a release print, but the copying process is much better than you give it credit for.
> noise (like film grain) is always a problem
Not really, dithering is essential for all video processing. Your very clean image is actually full of noise. Without dithering, videos (and movies) would look bad.
> I suspect you are missing what never was. Digital is better on both.
Absolutely nonsense. Kodak 2383 has a D-Max or 4, giving a contrast of 10000:1. Kodak Vision Premier Color has a D-Max of 5.5 giving a contrast of about 310000:1. Yes, not a typo, one-third of a million to one. There is no other technology that can achieve these deep blacks. Digital projection doesn't come any close, the best projectors today, (the ones that you won't find in any run-of-the-mill theatre) can barely do 2000:1.
DCI-P3, the color space used in digital cinema was modeled after print stock gamut, but because of technological limitations, it's not identical. Cyans and yellow has a higher chroma on film. Green on film has a higher chroma at the expense of luminosity. Other colors are pretty much the same.
The only way for digital projection to have a higher gamut is to switch to laser projection AND to retire DCI-P3. You need a green laser to produce more saturated greens.
Digital cameras are capable of recording a larger gamut than film, but unfortunately there is no technology to display these colors outside laser projection, and you need to switch the whole movie industry to a new color space.
If you go down to the microscopic level film goes way beyond the resolution you can ever hope to achieve with the best digital cameras out there. It's just not something that is expressed in pixels.
> If you go down to the microscopic level film goes way beyond the resolution you can ever hope to achieve with the best digital cameras out there.
That simply isn't true. At the microscopic level film pixels look like "clouds" of colored ink. The pixels are not square, but they are there, and they are not the size of atoms, but much larger - you can see them in a typical microscope (I know, because I tried it).
Digital is cleaner not because of filters, but because the actual capturing technology produces a cleaner signal.
A couple of years ago there was still an argument, but these days the argument is over. Digital clearly won.
No, only color reversal film (slides) looks like clouds. Negative film, the one used in movies, looks like crystals (because they are crystals).
> actual capturing technology produces a cleaner signal
This is true, albeit for a film like Kodak 5203 or Kodak 5213, you'd have a hard time proving that this is true. These stocks are so smooth with good exposure and proper development (as it done when making movies) that you need a grain focuser to see the grain. Many blockbusters and big productions that you probably saw in a digital cinema were shot on this film, and you never knew.
> Digital clearly won
Digital is much cheaper, it's very good, and allows much easier intermediate processing (editing, vfx, color timing, etc). Of clourse it "won", although film is still widely used, and although archival copies are still done on film.
The fact that digital won doesn't mean that projection has better resolution (it's roughly the same for 35mm, but worse for larger formats) or better contrast (it's far, far worse for digital) or better colors (film has more saturated cyan and yellow, although this doesn't matter much). It sure has less flicker though.
There are still plenty of reasons why people shoot film, use a digital intermediate workflow, and print film today.
You are using survivor bias to prove your point. My point is that Digital is inferior in SOME aspects to Film - and it still stands. The fact that Digital won has nothing to do with the facts to be considered when doing a comparison.
DCP can be a pain in the ass for indie filmmakers, though. These days it's easy to generate a DCP from footage, but basically impossible to test it without actually putting it through a DCP projector in a cinema. This is crazy.
But it takes a lot of experience to imagine what a frame that looks great on a ~24 inch monitor will look like when projected in a theater.
I would strongly recommend that anyone dreaming to screen something in cinemas make a deal with their local cinema in order to do test screenings - to check picture and sound. Just like with anything else, there's no substitute for testing under "real" conditions.
Note that most(?) cinema setups will happily take some form of digital audio/video signal and run it through dolby surround and project it - so you don't have to start testing with a DCP file. But you certainly should test that in a cinema before distribution.
Also note that the "player" part of a fixed installation might have quirks/bugs that software that's more easily updated might work around (think mplayer vs a cheap old DVD player).
It could also be the case that someone leaked the plaintext symmetric key(s) for this specific movie's DCP. If someone gained access to private/secret keys on a compromised DCP player somewhere, it'd be smarter to leak symmetric keys for individual movies to avoid detection.
What does that workflow look like?
Where can I buy one?
Last I heard, smallest 4k projectors start at $400k.
If you have the right friends, and hang out with the right group of people, you can get sometimes unencrypted material, just like in the old days you could get film reels. But a 35mm film projector was relatively affordable, this new digital stuff is still crazy expensive.
I'm curious what I should be keeping an eye out for, what sorts of people would be the most interesting to chase, etc, to maximize my chances of causing this opportunity to present itself at some point in the future. (I'm not particularly social right now.)
I'm not defending DRM, but it's important to know one's enemy.
It's frustrating wanting to give money for something, and not being able to.
This way he made a legal purchase and exchanged money, but still gets a DRM-free high quality copy to move around to his various devices.
If the focus is on quantity and diversity, I can understand just getting a decent-bitrate MKV or MP4 of just the film, but it might be worth it to see if there are Blu-Ray rips of the media in question "around" before just grabbing what's out there; at the very least, complete (and perhaps even decrypted, since that's finicky) rips have reasonable value on private trackers.
I'm guessing this sort of thing has probably already been taken into account.
For each movie, copies are released over time and a correlation can be looked at between quality of rip, distribution by file sharing, and box office receipts.
The "problem" is that it would obviously show that the impact of torrenting is much less than one lost sale per download. And reflect that even the downloads that are lost sales are offset to some degree by the people who download to sample and then buy when they otherwise wouldn't have, and the free promotion from the people who downloaded it talking about it with other people etc.
The result is that the likely "negative impact" would be small or could even be positive, which isn't the narrative Hollywood wants to spin. Because they actually want DRM for entirely different reasons than that, specifically because it allows them to control content delivery and thus prevent independent content from being displayed on equal footing with big studio content.
Yes and now, You cannot say that people who pirate games are necessarily going to buy the game if it were not "free". That's the argument used by the industry to justify DRM and it makes no sense whatsoever since the market would be like 100 times bigger if you calculate things this way. It would not.
This can't be true as written, as long as people expect it to show up there eventually. Every day it stays off piratebay is an extra day of revenue intermediate between full and "already out on piratebay" levels.
The DRM works because the value of these things appears to be greatly time-sensitive -- most of the cost is already getting-it-right-away, which DRM can protect, rather than the fundamental timeless value of the product, which DRM can't protect.
It depends on what the desired outcome of the DRM is. A prime example of effective DRM is Denuvo [0]. Loosely speaking, there are two main groups of people who pirate games - those who would buy it otherwise, and those who wouldn't. There's not much you can do about the latter group, so they're not really worth considering. The former group however are the potential lost sales. The sales for a AAA video game usually peak on release or very shortly after, and fall off a cliff very quickly., If your DRM system manages to keep a crack for your game from being leaked for the first few weeks of its life, you will capture many of the people who fall into the "might buy your game" category.
Also, I would be surprised their isn't a watermark in there that will point to the cinema who leaked the version.
And exactly the time and motivation needed is why DRM can work in the end. Modern DRM schemes like AACS employ methods to change keys for new works. So when crackers find a key that allows them to circumvent the DRM scheme they can't publish it or it would soon get useless and they would have to do the work to extract a key again. This is probably what we're seeing here.
This means that for consumers the DRM stays effective. They can't do copies of things they legally bought.
Modern DRM schemes like AACS
employ methods to change keys
for new works. So when crackers
find a key that allows them to
circumvent the DRM scheme they
can't publish it or it would
soon get useless and they would
have to do the work to extract
a key again.
Who would EVER care about a cipher key, when YOU STILL NEED TO EXPOSE THE RAW PIXEL RASTER, AND AUDIO CHANNELS TO THE END USER, IN ORDER FOR ANYTHING TO BE VALUABLE AT ALL?
Users will always eventually get the whole thing in the clear, in straight-up plain text somehow, eventually anyway. And all anyone needs is a buffer big enough to capture it, and it's trivial to assemble one.
You need to be able to watch a movie with the naked eye, and hear the sound with your ears. That's how movies work.
It's trivial to capture the raw data, and people have been living with NTSC quality picture and sound for decades.
This is not about perfectly matching the SHA256 hashes of the original MPEG artifacts. People just want a copy, and it's easy to skim one, somehow, one way (cracking) or another (brute force direct copies of the image frames and pulse code samples, at the signal source).
In theory, if the picture is revealed on a display, upon which the full pixel raster can be distinguished for every frame (in real time or slower), an optical capture may be performed, which preserves the original fidelity of the image, and one need not recompress as lossy for peers who have the capacity to receive the full duplicate.
All images are eventually revealed optically somehow. Such images may be captured directly, and corrected and restored according to original quality, if you have good video equipment.
Only one good copy needs to make it into the open, and then the cat's out of the bag.
HDCP is encryption during transmission (like SSL), e.g. encryption while transferring the movie data through the HDMI cable. It is not used for encrypting data on a storage medium (like CSS, and apparently also DCP are).
In the end, copyright is based on laws created by governements as trade agreements, rarely involving the voting public when it's not a dictator or monarch making the rules. In effect copyright is a matter of law, not moral, and sometimed the law is not moral. In this case, I stay my judgement as it can only cloud the real issues.
Personally I think everybody who is not an adult, without a job or a home should by default be allowed free access to anything we call culture. Being part of society is - which everyone should be - is down to having a shared context. The so called entertainment is a major part of this, and without access to it you become an 'other', the outsider.
Moral or not.
So many movies are only available commercially in their objectively worst version (pan and scan, poor subtitles, no subtitles just dubbing, missing scenes, censored scenes, poor audio or poor picture), people who actually care for the film more than the copyright holders go out of their way to combine several awful versions into a version better than commercially available because they actually care about what they're distributing because it's for passion not profit..
And this is the framework the US is pushing the rest of the world to adopt.
> That ideas should freely spread from one to another over the globe, for the moral and mutual instruction of man, and improvement of his condition, seems to have been peculiarly and benevolently designed by nature, when she made them, like fire, expansible over all space, without lessening their density in any point, and like the air in which we breathe, move, and have our physical being, incapable of confinement or exclusive appropriation. Inventions then cannot, in nature, be a subject of property.
--Thomas Jefferson
The use of DCP seems to be limited to very high resolution releases. HDCP is still what current era consumer hardware uses, and it's totally wide open.
As always with DRM, it's defective by design. https://en.wikipedia.org/wiki/Defective_by_Design
If you can view it, you can copy it... it's just a matter of time, motivation, and desired quality.