Why Film Could Be Key to Centuries-Long Data Preservation

[00:00:01] So today, recorded at the IT Press Store in Valletta, Malta, we're going to explore the fascinating innovation that could redefine how we approach data preservation. And joining me today is Antoine Simkin. He's the co-founder of Digifilm Corporation, a company that is at the forefront of preserving data on film for the long term.

[00:00:24] But they're now thinking much bigger than that. And with a background in cinema and decades of experience in visual effects, Antoine has witnessed firsthand the vulnerabilities of conventional data storage methods. From costly migrations to obsolete formats, these challenges inspired Digifilm's groundbreaking approach to putting digital data onto film.

[00:00:47] A medium renowned for its longevity. Not to mention it being offline and more secure. So in my conversation today, Antoine is going to join me and explain how Digifilm is blending the reliability of traditional film with modern day encoding. And creating an almost time capsule for critical information. And I want to discuss why current storage solutions like tapes and microforms fall short when it comes to long term preservation of information.

[00:01:17] And will also shine a light on the industries that could stand to benefit most from this technology. And the profound implications for sectors ranging from cinema and nuclear power to defence and architecture. So how can we ensure the digital artefacts of today remain accessible for generations to come? Well, enough scene setting for me. Let's get my guests onto the podcast now. So a massive warm welcome to the show.

[00:01:46] Can I tell everyone listening a little about who you are and what you do? So my name is Antoine Simkin. I'm one of the co-founders of Digifilm Corporation, which offers a solution for the preservation of data on the very long term. My background is cinema, actually.

[00:02:06] I was one of the co-founders of digital visual effects company in the 80s, 90s, where we have been involved into scanning and recording film from the beginning. Because this was in the days where cinema was shot actually on film and was exploited on film. So the challenge was to be able to actually digitize the film, do whatever tricks we had to do on it, and then record it back to film.

[00:02:30] But very soon we were confronted to one issue, which was how do we keep this huge amount of data that we were creating at the time? These were days where, I don't know, our first hard disk of 600 megabytes costed 5,000 euros. And it was really a miracle that we could already buy this incredible disk. So we were submerged with a huge amount of data.

[00:02:55] And we turned to industry-standard solutions made by Sony at the time, which was called the DIR-1000, for example, which was capable of storing data on tape. But somehow the format changed. After a while, Sony decided that they would move on to something else, the DTF. And we had to move on with them to the DTF, etc.

[00:03:18] And we discovered very quickly that although the solutions are absolutely great, for the short term, that's fine, but not for the long term. Because what happened if we didn't move the DIR tapes to DTF? And what do we do with the DTFs? We know for a fact, for example, that the people who have recently restored some of the film we did at the time from DTF struggled enormously because some tapes were unreadable somehow.

[00:03:42] So little by little, we thought, wouldn't there be a solution that would mix the best of both worlds? We know that film, for example, is a medium that lasts somehow forever. It's been here since 120 years. The last generation of film can last for even longer because it is polyester. So it's really very indestructible somehow. But we put images on there.

[00:04:09] How can we transform those images into a relevant image that would represent data? Because for everybody, the computer world is made of zeros and ones. Yeah. But those zeros and ones, actually in computers, they don't exist. It's only in our head.

[00:04:30] On a hard disk, on a tape, the zeros and ones are represented by some very thin magnetic dust that is oriented positively or negatively. And we decide arbitrarily that if it's a positive magnetic dust, it has a value of one. And if it's negative dust, it's zero. And it's the same for any computer chip. Is there a current going through or not?

[00:05:00] So if there is, maybe it's one. If there is not, it's zero. If there is five volts, it's one. Or it's zero. And we say we talk about bytes and this and that. But it is only in our head. This is the way intellectually that we made the computer world work. But actually, it's an analog world. It's not a digital world at all. And we've met here at the IT Press Tour in Malta. I've just seen your presentation. One of the things that blew me away is we're all passionate about entertainment movies and TV series.

[00:05:30] At the BBC in the UK, for example, they lost a lot of archived footage. And you mentioned, I think, three huge movies that you were a part of as well. And it's a very real problem, isn't it? This backing up of the data that we love so much and the content that we love so much. And especially when we're talking 2K, 4K, 8K, it's a lot of storage, isn't it? It's a lot of storage. I mean, our first goal, because we come from the cinema world, is about cinema.

[00:05:57] And so the cinema world, knowing and not trusting a digital world so well, actually very often records out on film an analog version of the film to preserve it in the long term. Yeah. Okay. But for me, it's exactly as if you would say, all right, the photos from my iPhone, I'm going to create an album so I can save it for later. Fine, you have an album. But if you want to get your photo back, what do you do?

[00:06:26] You have to scan them. And it's never going to be your original photos. It's a copy of your photos. That's a bit what the cinema world is doing. The other path, which consists of keeping it on the digital world, on tapes, hard disk, clouds, whatever, will make you dependent on a lot of things. And we think, wouldn't we be able to help cinema to preserve the films for the very long term once they have been made?

[00:06:55] And it's very important because we can see that the films, they have a very long career. And there's a whole world of film restoration, for example, of films from the 50s, from the 60s, 70s that are being restored. Because they're on film and we need them digitally now. But what happened to those films that are already digital where we're going to lose the data? Yeah, yeah. That's going to be a problem at some point.

[00:07:22] The question I've got to ask, because obviously we're talking about cinema and film here and the great work that you're doing. What was it that inspired you to use this idea, to use not just film as a medium, but also for encoding digital data? And how did your expertise add to the digital conversion in so many different areas? Because you're really expanding now on this idea, aren't you? Well, it's difficult to say how does an idea pop up. Yes.

[00:07:52] Yeah. What I can say is for us, and when I say us, it's our team, Rip Hampton O'Neill and Pierre Olivier, my partners in this company. We find it very, very natural to go back and forth from digital to film, because that's what we've been doing for a long time.

[00:08:14] And so it's pretty obvious for us that because we can write just anything on film, and as I was saying before, the world of data, of computer data, is actually an intellectual construction. Why wouldn't we say that if there is one pixel somewhere on the film, it would have a value of one? And if there is no pixel, it has a value of zero.

[00:08:40] And from there on, if you say that, then you create a sort of grid with pixels on and off, and that's it. And for us, it's very natural. I cannot explain how it comes from. It's a question of culture. And I think it comes from the 20 years that we've spent going back and forth from the two mediums and mixing them together in a natural way. I think probably someone who has not touched film or is only in the digital world or has only been in the film world wouldn't do this transition.

[00:09:08] But we are kind of a hybrid culture, and I think that's where it comes from. So LTO tapes are widely regarded as reliable and cost-effective, but what are their key limitations, and how does Digifilm address these challenges? Well, I love LTO tapes. I mean, it's a relief after all the different standards that have been there before. It's kind of an achievement where I think IBM is the lead into this consortium,

[00:09:35] where a lot of industry makers have gathered together and decided, all right, let's not fight with each other. Let's just make one format and do that. So I think that's a real progress. Fine. It's good, and I use it. And it's very good for the short-term and medium-term storage. What I believe is that LTO tapes, they have a certain lifespan.

[00:10:00] They are given to be valid for 25 to 30 years. That's what they say. Under the condition that you check them every six months to make sure that you can still read what's on it. And if not, well, you have to use the copy of the same tape to be able to recover your data, because if you can't read it, you can't read it. It's dead. The second thing is, and that's fine, I mean, then it means you just have to make sure that it's being done this way.

[00:10:28] Second thing is, you need a reader. And this reader is a very high-performance machine at a relatively low cost. The public price of a LTO drive is between 2,000 and 4,000 euros. That's fine. It's something you can almost have on your own if you have a medium-sized company. The thing is, progress is progress, and the LTO is evolving constantly.

[00:10:55] We are at LTO 9, and a LTO 9 reader or writer can write LTO 9 tapes and read LTO 9 tapes. They can only read LTO 8 tapes. They cannot read LTO 7 tapes. So if you have LTO 7 tapes, you need to migrate them on LTO 9. So every two or three years, you need to do this. So fine, you have companies whose business is to do this, and I'm perfectly okay with this.

[00:11:24] But from the user point of view, you have to realize that there is not a place where your data is on your tape. There's a place where there are tapes and people's data. And the day something goes wrong with the company that is doing this, how are you going to recover your data? Because there's going to be a very, very short time to come and get back a huge amount of data, which is spread amongst a huge amount of tapes under the pressure of people, banks and such,

[00:11:53] wanting to recover all the equipment that is there. And I'm saying this because it happened in the cinema world. There's a film lab that went bankrupt, and people had a very short time to recover their tapes, and films have been lost somehow. So although I still consider LTO tapes as an extremely good medium and probably the most economical way of storing data on the medium or short or medium term, for the very long term, I doubt of this.

[00:12:21] And I think that if you believe that in 50 years you will recover your data from an LTO tape, you'd better hope that the company that was dealing with it is still there, that you've paid your fees, and that all is okay. And that they actually really checked every six months that the tape was readable, that they had a second copy, etc. Now make the calculation of how much does this cost over 50 years. Nobody cares because you pay year by year. But in reality, it is a problem.

[00:12:50] It is a problem where someday this can be lost somehow. So the question I've got to ask there, and with technological advancements, you've highlighted a great problem there that we have right now. Many storage solutions lose compatibility over time. So how does DigiFilm ensure that that data stored today will remain accessible, not just in 5, 10 years, but 50 years or 100 years or even more? What are you doing differently here?

[00:13:15] Well, what we're doing differently is that what is stored on the film is an image. If you can imagine an image, maybe if I say a QR code, that's the most easy way to relate to it. It's an image that consists of little dots, right? This image is written on the film and will stay written on the film. Doesn't need to be migrated. Doesn't need to have anything done to it.

[00:13:43] It just lies there and it will not change because film does not change. In the very long term, maybe it will not be as bright, but it will still be on. So if a pixel was there, maybe it will not be as bright, but it will still be visible and it will not change color or layer or anything. It will still be there.

[00:14:06] So there's absolutely no need of migration or maintenance or anything or looking if it's okay or anything. We just store and forget. And this is a place where you can go back when everything else has failed. I'm not saying don't do the rest. I say absolutely do the rest. Yeah. So keeping in mind that the most important part of your data, you'd better have another sort of copy on another sort of medium. Actually, that's the way the UNESCO consider that you need to keep your data.

[00:14:36] You need to keep them on different sort of medium. Because if you have, for example, two hard disks with your data, it's not as good as having one hard disk and one LTO tape. The best is to have one hard disk, one LTO tape and an Archiflex database that you can put somewhere in case the others fail. And they will because a hard disk lasts only 30,000 hours. Yeah, yeah. I was speaking to someone recently and they said there are two kinds of people in this world.

[00:15:03] Those that are making backups and those that will be making backups in the future. Yeah. Some might argue that traditional methods like microforms or magnetic storage are sufficient in their own right. But what would you say to those skeptics who maybe question the need to do something differently? Well, regarding microforms, I think that's a very interesting topic. Microforms actually are pictures of files. Let's say you have an Excel sheet. You take a microform.

[00:15:32] It's like printing it on film. Yeah. So what you get is your Excel sheet on a piece of paper. It's a piece of film, but it's the same. It's a picture. Yeah. If you need to recover it, you scan it and then you have an image of your Excel file. Where are the formulas? Where is everything? Where is the date of creation of the file? Where is the name of... Nothing of this exists anymore because this is a TIFF file. It's just a reproduction. Yeah.

[00:15:59] So although microforms seems to be a very simple way of saving a ton of documents, you just have a picture of your documents. It's exactly like what I was saying about my iPhone pictures. If I have an album, it's fine, but it's not the pictures. Yeah. It's a reproduction of it. And it loses everything from the original file.

[00:16:22] So we consider that against microforms, we have a huge advantage because that file that you use to create the microform, whether it's a PDF, a CAD file, an Excel sheet or any kind of file, whatever it is, because we're completely agnostic. We're just using data.

[00:16:41] When you will recover it from Archiflix, you will recover your original file exactly as it was without being possible to tamper with it because on the film, you cannot change it at all. And you have the level of security because it's offline and people who need to access it, they need to scan it. It's, you know, somehow the security is much increased with this solution.

[00:17:04] And we've focused heavily today on the cinematic industry, movie studios, but I'd love to get people thinking bigger about this as well and how far-reaching it could be. So in your vision, who stands to benefit the most from your technology? And maybe share a few examples or use cases in industries that you'd find digifilm particularly transformative or potentially transformative in the future. Well, we started with the cinema world because it's the one we know the best.

[00:17:34] And it's very natural for us to go there. Although somehow we believe that the cinema world size in terms of market, the amount of data that is generating, we believe it's probably five to 10% of the global market of data that needs to be really preserved. Some data doesn't need to be preserved, but some data needs to be preserved. And so as we were developing the technology, we started to talk with other people.

[00:18:03] And we realized that in the industry, I'll take an example, for example, of nuclear power plant builders. A nuclear power plant is going to be exploited between 50 to 70 years. It takes about 15 to 20 years to build, and it takes more than 30 years to dismantle. We are way above the level of 100 years. So how do you preserve the data currently on microforms?

[00:18:31] So imagine what it is to recover microforms instead of the original files for whatever you need. I think this is something that could be really interesting for them because although they will preserve it through the regular route, LTO migration and stuff, they know that over 70 years, something wrong will happen at some point or potentially will.

[00:18:57] And the problem is not, the problem is how much does it cost to recover from such a disaster? It's impossible to calculate somehow. But it is true also for, I don't know, a boat, you know, a super tanker. It's going to work for 60 years. So it's the same problem. A lot of bridges, a lot of tunnels, a lot of buildings, a lot of, you know, all that kind of data. It's the same thing.

[00:19:24] It's like buildings, industrial installation that lasts for a very, very long time. A stadium, whatever. All these things, they need architectural plans to blueprints to be kept for a very, very long time. And currently there are no other real solution to keep them potentially alive, which by this I mean recovering the original file and not a picture of Nuln Rint. This is the only viable solution somehow.

[00:19:54] It's also potentially something for the defense industry. In the defense world, there are some big secrets and some very, some files that needs to be preserved. And it's better to preserve them offline because anything that's online will be hacked somehow. So the army already knows about that, has some special unconnected computers in bunkers and whatever. But what happened in case of an EMP? Yeah.

[00:20:23] If all the magnetic support is erased by something like that, well, if it's on film, it will last. So maybe it can be burnt. Well, the magnetic tape will be burnt as well. So that doesn't change anything. But at least it protects you from the EMP. And also in the politics of having different kind of medium to keep your data, it wouldn't be bad to have an extra copy on Archifreex medium and put it somewhere in another bunker in case something happens. So that's another application.

[00:20:51] It could be good for the notaries also. So like if you want to keep the notaries, they have to keep for a very, very long time all the wills and all the transactions that have been made during the life of someone. Some money has been given to somebody else and we need to keep track of that because maybe 70 years later when they die, we have to know who is given what to whom. And well, how do they keep that for the moment? It's very fragile.

[00:21:17] So there are ways, but they're not completely safe to my understanding. Yeah, it's such a great point about that need to keep storage offline and avoiding hacks and things like AMP, etc. as well. And as Digifilm evolves, are there any other advancements that you could possibly foresee in the future? And how do you envision the role of Digifilm in shaping the future of that long-term data preservation? Because we never talk about long-term data preservation.

[00:21:47] Well, our impression from all the talks we have with different people from different industry that we are meeting as we go is that because there is no real solution on the market, people don't really look at the problem. Yeah. Because everybody that is a bit reasonable knows in the back of their heads that something can really go wrong with data and we can do whatever we want.

[00:22:13] We can develop any kind of, you know, double, triple security. As long as it's magnetic and there's migration, there is a risk because each time you migrate, it's a copy. And if you make a copy, it can go wrong and all that. And also you're concentrating it, etc. The future of what we are doing is we currently are at a stage where we believe that we can already replace the microform world.

[00:22:43] We are not in the terabyte world somehow. We can replace a huge amount of documents in a reasonable space, but we are very conscious that we need to increase the capacity of our storage potential. So this is what we're working on. We're working on improving the density of what we can write on the film.

[00:23:11] Nevertheless, we are very careful with one thing. We want to keep it in a human readable capacity. Because if we go too far, then we're going to fall in the trap of all the other solutions that are absolutely wonderful and promising, but who all need a machine to be able to decode it and to read it, that is not a machine that can be made by amateurs.

[00:23:39] That cannot be made by, you know, let's get together and let's build the proper scanner to make Archiflix because there are some interesting data to decode. No one will be able to make a machine on their own that can decode a silica system. No one will be able to, we don't even know how we are going to write on DNA, how we are going to read from DNA. I don't know.

[00:24:04] And it's great that we do this and we must do it because we have tons of very interesting data to store for medium whatever. But I don't think it's so important to save all the data. Yeah, some data, and it's not so much, some data needs to be preserved. Not all of it. And for anybody listening anywhere in the world, in any industry, maybe you've sparked a few light bulb moments today of, hey, this might be able to work for me.

[00:24:31] Any way you'd like to point everyone listening on where they can find out more information? Well, there's a website of Digifilm Corporation, digifilm-corp.com, where they will find white papers and different things. And also, what's important about us is we are not a B2C company. What we are providing, we are not selling equipment. We are providing solutions for companies that are already into such and such business.

[00:24:58] If companies are in the business of doing long-term preservation of data and they know they have proper contracts, but they are struggling to actually be able to make it. Yeah. Well, maybe they can talk to us and we can look together at how best can we help solve their problem. Because we cannot imagine all the problems of everybody. We have our own path and we grow up understanding more of what's going on elsewhere.

[00:25:25] But we are trying to break the silo and trying to understand who else has what kind of problem. Because I'm pretty sure there are problems I don't even imagine that maybe we can solve. And I think that for the moment, yeah, a lot of people don't foresee that problem because they don't see a solution off the shelf. So then because this solution is there, maybe the problem will appear. It's like, oh my, oh my God, yeah, it's true that in 50 years, how is this going to run? Maybe people don't care because they won't be there.

[00:25:57] But some stuff are important to keep. 100%. Well, I'll have links to everything so people can find you nice and easily today. And I would urge anybody listening, if they're concerned or interested or excited about long-term data preservation, to reach out and contact you. But just thank you for stopping by and saying hi today and sharing your story. Thank you. Thank you very much for allowing me to talk about my time capsule. Wow.

[00:26:20] I think Antoine's journey from the cinema world to co-founding Digital Film Corporation highlights the urgent need for more innovative solutions in long-term data preservation. We're so obsessed with the right here, right now and next year, but we seldom think about the impacts of any technology or even protecting data 150 years from now. But what I love about today's episode is combining the durability of film with cutting-edge digital encoding.

[00:26:47] That collision has enabled Digifilm to address a problem that many of us haven't considered. How to ensure that vital data survives for decades, if not centuries. And as Antoine shared the importance of keeping critical information offline and secure, I don't think that can be overstated, especially in industries where data longevity is essential. And Digifilm's approach not only safeguards data, but I think it's bigger than that.

[00:27:16] It also opens up new possibilities of how we think about preservation. Whether it's architectural blueprints, cinematic archives or sensitive defence records, the potential applications appear to be vast here. But what are your thoughts on long-term preservation? Have you thought about how you protect data for 100, 150, 200 years? Do you see this as a growing concern for your industry? How are you tackling it? As always, let me know. Join the conversation.

[00:27:44] Email me, techblogwriteroutlook.com, LinkedIn, X, Instagram, just at me or CQs. Love to hear your thoughts on this one. But that's it for today. So until next time, thank you for tuning in as always. Bye for now.