The End of Celluloid As We Know It

[MichaelB]

Archives will need to be able to print film pretty much indefinitely, so there'll be a permanent core demand.

[flyonthewall2983]

I was very bummed to learn this Thanksgiving my home town theater is soon to have a date with the wrecking ball, because converting to digital would cost what the theater makes in 5 years total.

[mfunk9786]

It sounds as if the theater has other problems besides digital projection.

[Brianruns10]

Well, it looks like the end of film may soon be on us. Recently, Technicolor announced it will be supporting digital projection. It looks like Kodak will also move in that direction. It may be soon that those who want to shoot in film may not have that option and generations of kids will not ever see the magic of celluloid.

Now, I use the term cinema because in the last few years you really have not seen any classics. No Citizen Kane. No Godfather. We often hear that digital will save cinema. There will be a renaissance of small filmmakers making great films and the studios will not be able to prevent this. However, most of the films we have seen shot in digital are not up to par to the classics.

No film in the last 20 years has matched the Black and White contrasts of German Expressionism. Digital color is no where near Technicolor's dye process. Even the most advanced editing software has not produced a film that can compete with Battleship Potemkin. The technology now is changing too often. Cameras are now obsolete after a couple of years. Movieolas lasted us for half a century. If you don't have the latest edition of AVID, you're in the stone age.

As we potentially enter this brave new world of cinema, I want to know your opinions. Is it good? Is it bad? Will it happen? Will it die from overhype? Will film and digital coexist peacefully? You know my thoughts. I want to know yours.

I think you're sounding a bit like the folks who called sound vulgar, and color garish. The fact is this technology is here to stay, and I'm grateful for the variety. Cinema isn't about format. It isn't about HOW you make it. It's about craft and artistry and telling a damn good story. I mean, 65mm is the greatest motion picture format in my opinion, but in the hands of some schmuck who doesn't know how to wield it, it'll look like shit. I've seen plenty of awful stories shot on film, and many great ones shot on digital, and it is a fools errand to say the films being made now will never be as good as the ones made during the good ole years. The fact is, you never know. You have to let time and the tastes of the people sort things out. Things have a way of working out, and not how you'd expect it.

Why, when sound came onto the field, along with motorized cameras, there were those who thought the cinematographer would disappear, because now one would need only push a button and the camera rolls. But of course, that wasn't the case.

More and more the trend is toward modular cameras, and I think the Epic is the way forward in this regard. So we don't have to buy new cameras all the time, maybe in the future we'll just upgrade new parts. That's my hope at least. Beside, the smart filmmaker doesn't buy cameras. They rent as the budget demands. I'm an industry pro, and I have yet to even buy an HD camera.

Will film go by the wayside? As an exhibition format, yes, and I'm glad for it. I'm fed up with crappy projection by teenagers who don't know the first thing about footlamberts and contrast ratios. I'm thrilled with the quality of top shelf digital exhibition, and having seen Lawrence of Arabia in 4K, I was completely sold on the format.

As an acquisition format? It's hard to say. I mean, for film supposedly dying out, we've had a renaissance of large format, with two features shot entirely on 65mm, and several more with 65mm sequences. And more large format films are in the works. Film is certainly more archivally sound and future resilient, and if I were a studio head, I'd be shooting all my AAA big budget productions in 65mm. They've shown it is cheaper to make separation masters of the original, than it is to constantly migrate and store digital files. But eventually Moore's Law will ensure the prices go down, and hard drives are becoming far more reliable.

In the end, a lot can happen. And we must not get so worked up about how we make films. We should worry about making them at all. I'm about to embark on a new documentary next year, and I'm going to shoot it any way I can, with whatever I can, format be damned. Just make movies. Good movies. If it's a good movie, it won't matter whether it was shot on Hi8 or 35mm. If it's a bad one, even the most beautiful, crisp, satiny celluloid won't save it.

[YnEoS]

But the coming of sound and color didn't introduce some pretty legitimate concerns about the ability to store and preserve cinema, the way digital does. I'm in agreement with you that creative people will always make great works of art regardless of the format, and I'm sure one day digital will surpass film in every way, even as an archival medium. However part of the process of making that transition is precisely worrying about what it is we might lose when technology advances. Films being lost forever is a reality, there's plenty of examples of lost masterpieces from the early years of cinema. And even though people are more conscious about archiving these days, films still continue to be lost, especially among lower budget filmmakers. And I have no doubt that quite a few films will be lost in these early years of the digital takeover. So I think now is precisely the perfect time to be worrying and spreading awareness of some of the risks that come with making digital cinema.

[MichaelB]

As restoration expert James White pointed out in his recent AV Forums podcast (it starts 30 mins in, and is well worth a listen even if you have no interest in Zombie Flesh Eaters, the specific project under discussion), film is still the best medium yet devised when it comes to longevity. Videotape has a very limited shelf life by comparison, and digital files can rapidly become unplayable - and often the filmmakers are blissfully unaware of the problems until they find this out the hard way.

When I interviewed Austrian experimental filmmaker Peter Kubelka the other month, he quoted a recent Kodak marketing initiative with some glee - they now manufacture a film that's specifically designed to create a more permanent copy of digital media. Kubelka was particularly taken with the phrase "asset protection", and noted that the notion of "protection" also extends to Kodak's entire business. But Kodak's right: this really is the best way of ensuring the long-term survival of these media.

[Brianruns10]

Archival problems have ALWAYS been there. Digital poses nothing really new. Stuff shot on tape is problematic. I deal with quite a bit of archival tape footage, and stuff not even two decades old is already starting to show problems...I've encountered several tapes already that were beyond saving. And this stuff has been stored properly. It's just the stuff doesn't last. Even film goes bad if not stored well. There are so many variables beyond mere format, and in the end, you've gotta have multiple copies across different mediums, each of redundant quality.

While I completely agree film is the best medium for preservation, I've come to regard it as a highly theoretical argument with little to no basis in today's fiscal realities for the vast majority of filmmakers.

Sure, I'd gladly preserve all of my films to this new kodak stock...but who's going to pay for it?

The digital revolution happened because there were many wanting to make films but couldn't because film and processing and prints were so damn expensive. Speaking as a filmmaker, who has struggled just to get his work funded, it's impossible to preserve it in the manner prescribed. It simply isn't a realistic option for 99.9 percent of all filmmakers, who barely can afford to get their films made, let alone worry about an extremely expensive digital to film conversion, or the creation of black and white separation masters.

Besides, digital is a convenient scapegoat for a larger, institutionalized neglect going on in this country. Say in a perfect world, every indie had the money to preserve their masterpiece to film. Then what? Who's going to store that film? Film is as only archivally sound as the archive it's kept in. Do we really have the sufficient climate controlled space to store all this material? Most would probably just welter in some closet. And who is going to periodically check up on that film, to make sure the dyes aren't shifting, or the base is turning to vinegar? These poor archives are understaffed and underfunded, and their stuff largely sits in cans un-noticed. My god, Universal can't even be induced to properly preserve and restore Vertigo, and it got voted the greatest film ever by Sight and Sound. What hope do all the little films have?

Digital has not killed film. People have killed film. Neglectful studios, myopic federal and state and local leaders who have slashed budgets for archives, and charitable and corporate sponsors who turn their gaze elsewhere. Projectionists killed film with shitty, murky, shaky projection.

Frankly I think Kodak is on a fool's errand with this new low fade stock for preserving digital assets. Because who is going to take advantage of it? Hardly anyone aside from the majors, and they have so much money, they can afford to really do it right, by making separation masters. For the rest of us, this new film stock means nothing. Absolutely nothing.

The film guys need to figure out a way to make film really, truly, democratically affordable to filmmakers everywhere. Get people ORIGINATING their stuff on film, so the master exists. Because there simply isn't money for most to preserve their stuff to film after it's all said and done. They're just trying to get back out of the sea of red ink in which they have been swimming.

Kodak and Fuji need to develop truly low cost, competitive film stocks. Offer up a cheap, really cheap 16mm stock. Contract with the labs, so I can get affordable processing and a cheap, SD telecine with keycode. Enable filmmakers like myself to get our films shot, so we have something to fund raise with, and then we can go back and pay the oodles of money for a 2 or 4K scan and color correction and all of that. Establish a means for us to shoot films on archivally proven formats from the get go, instead of this silly bid to get people to invest in film on the back end.

That or we need to accept what is coming, what is already here, and devise a means for everyone to preserve their material. How about a massive, redundant server array, supported by a onetime upload fee, modeled after the LoC's system of accepting prints of copyrighted films for their archives. Devise a means so that indies everywhere can upload their copyrighted material so it can be preserved in the cloud. Devise a sound method to preserve digitally. Let's think of a digital means of creating the answer to separation masters. Let's find a solution that works for what we have, instead of pining on and on about a method that, while time tested and vouchsafe, is totally out of the price range of most filmmakers, and poses long term problems of storage and environment and human resources.

[jindianajonz]

Can somebody please explain to me the risks of storing digital media? I would have figured that since digital files are so easily and cheaply duplicated, it would be the best way to preserve film. I could understand different file types going out of style, but I wouldn't think it would be too hard to retain a program that will be able to read whatever format the file is in.

[Gregory]

Digital has not killed film. People have killed film. Neglectful studios, myopic federal and state and local leaders who have slashed budgets for archives, and charitable and corporate sponsors who turn their gaze elsewhere. Projectionists killed film with shitty, murky, shaky projection.

It seems safe to say that the push to digital distribution and projection had far more to do with it than what's happening with archives. The writing was on the wall for the switch to digital with the launch of Avatar and the enormous demand for 3-D. The trend had already been underway due to the higher profitability of digital distribution.
The murky and shaky projection of 35mm you refer to also has a lot to do with the switch to digital, because a lot of the old projectors that commercial theaters have been using are increasingly run-down and in need of repair or replacement, which often doesn't happen because the theater anticipates the ongoing move toward digital, so why put money into projection systems that are on the way out for major theaters anyway (especially if virtually no one in a given audience is bothered by the shakiness or poor focus enough to complain, which is largely what I've experienced)?

[MichaelB]

Can somebody please explain to me the risks of storing digital media? I would have figured that since digital files are so easily and cheaply duplicated, it would be the best way to preserve film. I could understand different file types going out of style, but I wouldn't think it would be too hard to retain a program that will be able to read whatever format the file is in.

With digital media you have a double problem: ensuring the longevity of the physical materials storing the data, and ensuring that the information itself can still be interpreted decades from now (even assuming it isn't encrypted in some form to prevent piracy) - no mean feat when standards seemingly change every few months. In both cases, digital media are at a severe disadvantage compared with film.

The comments on this article are well worth reading.

[jindianajonz]

Thank you for that link. It did help clarify some things.

I still don't see how the two things you mentioned are insurmountable issues, though. Addressing longevity of materials, I figured this would be a non-issue due to the greatest strength of digital media: the fact that it is so cheap and easy to make duplicates of things. As long as you had a couple of backups of your data to fall back on and periodically cycled out old devices for newer ones, it doesn't seem like it should be an issue. Yes, having to rebuy multiple hard drives may be expensive, but when you compare that to the costs listed in this thread for creating a film reel, they seem pretty reasonable. And while film seems to be getting more and more expensive, digital storage just gets cheaper as time goes on- those uncompressed video files that seem huge now will probably just be a drop in the bucket 10 or 20 years from now. I would assume that 2 hours of uncompressed digital video at a certain resolution would always be the same file size. It may be pretty pricey to buy storage for it now, but 20 years from now it will most likely be very cheap.

As for being able to play back the data, really all you do when you playback a file is use an algorithm to convert the bits into an image. As long as you have a player that can do that you should be ok. Yes, you may have to make some tweaks to make it work with more current operating systems, but the basic idea of converting 1s and 0s into a pixel of a certain color doesn't seem like its anything that we wouldn't be able to preserve in the future. The reason formats like VHS and cassette tape (and eventually DVD and Blu-ray) go obsolete is because we stop producing players that play them, and starting up a hardware line is prohibitively expensive. But because players, like the video files themselves, are digital, duplicating and storing them should be fairly cheap and easy.

Now, this all seems pretty intuitive to me, but normally when things seem clear in subjects I don't know much about, it means I'm missing something. Can anyone explain what I'm not getting?

[warren oates]

Thank you for that link. It did help clarify some things.

I still don't see how the two things you mentioned are insurmountable issues, though. Addressing longevity of materials, I figured this would be a non-issue due to the greatest strength of digital media: the fact that it is so cheap and easy to make duplicates of things. As long as you had a couple of backups of your data to fall back on and periodically cycled out old devices for newer ones, it doesn't seem like it should be an issue. Yes, having to rebuy multiple hard drives may be expensive, but when you compare that to the costs listed in this thread for creating a film reel, they seem pretty reasonable. And while film seems to be getting more and more expensive, digital storage just gets cheaper as time goes on- those uncompressed video files that seem huge now will probably just be a drop in the bucket 10 or 20 years from now. I would assume that 2 hours of uncompressed digital video at a certain resolution would always be the same file size. It may be pretty pricey to buy storage for it now, but 20 years from now it will most likely be very cheap.

As for being able to play back the data, really all you do when you playback a file is use an algorithm to convert the bits into an image. As long as you have a player that can do that you should be ok. Yes, you may have to make some tweaks to make it work with more current operating systems, but the basic idea of converting 1s and 0s into a pixel of a certain color doesn't seem like its anything that we wouldn't be able to preserve in the future. The reason formats like VHS and cassette tape (and eventually DVD and Blu-ray) go obsolete is because we stop producing players that play them, and starting up a hardware line is prohibitively expensive. But because players, like the video files themselves, are digital, duplicating and storing them should be fairly cheap and easy.

Now, this all seems pretty intuitive to me, but normally when things seem clear in subjects I don't know much about, it means I'm missing something. Can anyone explain what I'm not getting?

Idk, jindianajonz, but the very things that make digital so appealing also make it almost absurdly vulnerable if you're thinking truly longterm, at, say, a Library of Civilization level about preservation in light of unknown future developments and epoch making disasters. A wet book might be dried out and salvaged. A wet server is fried forever. And multiple backups in different physical locations are still vulnerable to hackers, malware and unanticipated Black Swan events that uniquely effect the digital realm, like EMP strikes or the electromagnetic effects of space weather. Perhaps I've been watching too much Doomsday Preppers. But our increasingly digital world strikes me as more fragile and less rebootable than the analogue one we've left behind. And that's just crucial civic infrastructure I'm talking about. Not optional luxuries like film/video/media preservation.

[MichaelB]

The bottom line is that you can "read" a century-old piece of film with nothing more sophisticated than a light source.

With digital media, you need some form of assistance - which may not be forthcoming in several decades' time.

When the BFI transferred all its 2" videotape holdings to modern formats, the exercise cost millions of pounds, and was carried out in the nick of time - as a former colleague of mine put it, it's hard enough getting spare parts for equipment that's four years old, never mind 40.

[zedz]

All of those digital procedures you describe (migrating files, ensuring playback) represent ongoing costs, and rely on the rights owner being vigilant in perpetuity with their digital assets - which, in real terms, is extremely unlikely. A 35mm print of a film can sit in a barrel for a hundred years and be resurrected in all its glory (see the glorious Mitchell and Kenyon precedent). Five years of neglect with digital media and a work might be lost forever.

[warren oates]

The bottom line is that you can "read" a century-old piece of film with nothing more sophisticated than a light source.

That's it right there, Michael.

Zedz is right too about the constant attention digital storage necessitates. Ever try to spin up an external hard drive you haven't run in a year or more? Even then it's dicey.

This has me thinking of how tenuous the history of film preservation has already been even in the celluloid realm. Of how forgotten prints in places all over the world, including in the closet of a mental institution, have contributed to fuller versions/visions of masterpieces like Metropolis and The Passion of Joan of Arc. Or how Henri Langlois simply buried a vast swath of his country's film heritage before the Nazis could track it down and burn it. Would such films have been safer distributed into a worldwide computing cloud? Perhaps. Perhaps not. Not if they'd been lost or forgotten for decades and allowed to erode or be erased on neglected servers. And not in case of a preemptive attack on a country's cultural heritage and digital infrastructure by hackers, which seems a possible prelude to future global conflicts.

Or perhaps we'll all be post-human by then anyway, if Ray Kurzweil's right. And we, as our new (half) robot overlords, will laugh at old humanity's prior vain efforts to preserve its cultural artifacts that mean about as much in the future present as an anthill's architecture means to the kindergartner who steps on it.

[MichaelB]

When the BFI transferred all its 2" videotape holdings to modern formats, the exercise cost millions of pounds, and was carried out in the nick of time - as a former colleague of mine put it, it's hard enough getting spare parts for equipment that's four years old, never mind 40.

Oh, and stating the obvious, the whole exercise will have to be done again once Digibeta falls out of fashion. Assuming anyone's prepared to pay for it.

[Zot!]

I think the vulnerability of digital storage is being a little bit exaggerated here. It's not as though celluloid is somehow immune to destruction, loss, and especially not damage. My concern is more that a 2k scan of a 35mm source is not archival material, and if not done correctly can be ruinous. It's also not a good business proposition, as digital files are easilly shared, and companies may find that their product is being passed around with little regard for ownership. At this stage, I can't imagine why anyone wouldn't want to do both to archive a film.

[Drucker]

I can't add to the detail Michael and others have mentioned, but the idea that once something is digitally stored it's secure is not always true. Hasn't a Word Document or an MP3 ever become corrupted and un-useable for you?

[cdnchris]

A proper RAID set up would work for storage, so that shouldn't be a big concern. Multiple drives across multiple servers in different locations would work effectively. Like Zot says the concern should be more over the fact a digital scan, as of now, simply isn't as good as film itself.

[jindianajonz]

I think it's a bit erroneous to compare a file becoming corrupted to long term problems storing digital media. If my entire job was to make sure a file did not become corrupted over time, I'm sure I could manage. I would just make a lot of backups on different media (hard discs, DVDs, maybe invest in some cloud storage) and assume that no matter what happens, at least one of those would make it through. If I had a team of people at my disposal, I could probably also create some sort of test program that checks the files for signs of corruption on a regular basis.

Another thing that hasn't been mentioned is that digital media in all likelihood get more and more reliable as time goes on. We already have solid state hard drives that are much less vulnerable than our old ones. And another important factor is that unlike VHS and cassette tapes, maintaining a digital file, whether its a movie or a list of customer accounts or what have you, is something that every industry in the world is improving. With so many people looking for improvements in data security, it's inevitable that better and better resources will become available. And assuming these things are stored on something like a USB hard drive as opposed to an old tape style that requires specialized hardware to read, transfering from older media to newer media should be much simpler as well.

[htshell]

Keep in mind that average shooting ratios for feature films are now 10:1 what they were in the film era. Now many films are shot on RED Cameras (see this graphic) and similar that can capture up to 450MB of data per second. Films aren't necessarily edited with the raw files, so there may be a compressed mezzanine file for each, and so on. So if each feature film is a couple terrabytes, the cost of storage will be very high.

And then there's bit rot. Bottom line is we just don't know yet what will happen to a file that is on a hard drive for 100 years, but we do know that with proper storage, some films can still be in great shape 100+ years on. (I got see the original 35mm nitrate camera negative for The Great Train Robbery at the Library of Congress, which was in wonderful shape!)

[jindianajonz]

Oh, I don't know if it would be all that expensive. 450MB/sec * 60 sec/min *120 min (i'll use this for an average film length) gives us a tad over 3TB. Newegg has some 3TB hard drives for $170. According to the article linked earlier in this thread, each 35mm film print is $1500. I figure $1500 would be enough to buy 3 seperate hard drives and replace them every 5 years or so (assuming storage gets cheaper and cheaper until 3TB becomes inconsequential; http://www.mkomo.com/cost-per-gigabyte" onclick="window.open(this.href);return false;)

Now if you have three separate copies of the same file, bit rot won't be as big of a deal. You could just periodically check the files against eachother, and whenever one file has a discrepancy compared to the other two, you can correct it. This may take a lot of time right now, but as processing power gets better and better, scanning 3TB of data 3 times over is just going to get faster and faster. Since bit rot is random, it would be unlikely to have the same failure in two independent files, and even if there was, the only consequence is a single pixel of light may have a different color. Compared to all the things that could go wrong with film (where a single failure, even if less likely to occur, will likely have a much bigger impact on the overall picture)

Now again, i'm no expert here, so if I have any faults in my logic please let me know

[htshell]

You're only accounting for time of the finished film and not all footage shot. So that would be 30TB per film. How many films are made a year? And if we're arguing (as some on this thread are) that digital is cheaper and more cost-effective than film, is it cost effective for the independent filmmaker to store 30TB of material per film in three places and periodically check the files against each other?

And much of your argument is based on the slippery slope that "things will get better in the future" which is conjecture. I understand the logic you're following and while it might address the needs of one particular film, it doesn't account for the volume of all films produced. Also, I highly doubt that many independent filmmakers would realistically add a $1500 line item per film for preservation of the style that you're talking about above.

[Brianruns10]

Don't forget, the cost of a 35mm print is from a conformed negative. Most films are finished digitally, so you're needing to output to film, which is a hell of a lot more.

For example, the lab I use for my film processing charges $325 per minute to go SD to 35mm. For HD or 2K it's $400 a minute, and 4K it's $500! So it's going to cost you $1500 just to put a 3 minute short onto 35mm. A 90 minute feature shot in 4K, to be recorded to this magical new Kodak stock will cost $45,000. A big jump from $1,500.

[jindianajonz]

You're only accounting for time of the finished film and not all footage shot. So that would be 30TB per film. How many films are made a year? And if we're arguing (as some on this thread are) that digital is cheaper and more cost-effective than film, is it cost effective for the independent filmmaker to store 30TB of material per film in three places and periodically check the files against each other?

And much of your argument is based on the slippery slope that "things will get better in the future" which is conjecture. I understand the logic you're following and while it might address the needs of one particular film, it doesn't account for the volume of all films produced. Also, I highly doubt that many independent filmmakers would realistically add a $1500 line item per film for preservation of the style that you're talking about above.

Is it common for all elements to be archived, as opposed to the final film? I figured the final cut is all that most people would worry about when considering long term preservation.

As far as adding line items per movie (i use this term to avoid confusing the product with the medium), I'm not following your logic. Would paying $1500 up front for film stock be easier than paying $1500 over the course of decades?