PhotoCounter presents an edited transcript of this week’s PMA podcast interview with Vint Cerf, a pioneer of the internet and currently Google’s ‘chief internet evangelist’ on the challenge of preserving images in the digital age:
Jennifer Kruger, PMA: …So you spoke about a coming ‘digital dark age’ and used the term ‘bit rot’. Can you briefly sum up your thoughts about the dangers that our digital information is facing right now?
Vinton Cerf: There is several, but the first simple one is that the digital media themselves don’t last forever, and the second – even if they are still workable, we may not have anything to read them. For example there’s some [computers and devices] that are made now that don’t have DVD or CD readers in them any more. So you have a problem with reader persistence,. You have a problem with digital bit persistence itself. But even when those two are still available you have a question of what the ‘bits’ mean. Often that means having software around that knows how to correctly interpret and render the ‘bits’ that you save.
By ‘rendering’, I use this term generally. It could be presentation of a document; it could be an interactive spreadsheet; it could be a video game; or it could be an image or movie. All of those things that require software imply that if you want to retain the meaning and render-ability of these digital objects you may need to hang on to the software that made them. If the software isn’t available any more; or doesn’t run on the latest operating system; or the company has gone out of business and hasn’t maintained the software, you may find the files that you’ve preserved regardless of what medium they’re on may no longer be interpretable. The software isn’t available to do that interpretation. That’s what I’m worried about and that’s what I mean by ‘bit rot’. It’s not just physical ‘bits’ that may longer be readable, but the software that’s needed to interpret them might not be available.
JK: …Can you tell us briefly about you’re ‘digital vellum’ concept to solve that ‘digital dark age’ problem?
VC: Well, it’s just a term I made up, so it’s not like I have a huge system available… ‘Digital vellum’ for me means an environment in which older software can still be executed, and therefore digital objects can be rendered. When it comes to imagery, printing can certainly help, especially if there’s high quality paper and so on… inkjet is sort of a mixed bag as far as I can tell, which is one reason why I am inclined, when I’m dealing with images that I think are really important to my family or business… is to have them professionally printed on long-lasting stock.
I do want to point out, that if you don’t print them in a time when the format of that image is still understandable, you may find that later on it’s not possible to print them because the format isn’t understood… The problem may still occur even if you intend to print something. If you wait too long then for some reason the format may no longer be known.
A lot of people claim ‘well y’know we’ll always be able to print JPEGS blah blah blah,’ but I have found that some other software that I use for rendering images doesn’t know how to correctly render some of the older image formats. I suppose that if I work hard I could find somebody who still has a piece of software that can do that, but it’s sort of a stunning – and shocking – realisation to find that a digital image that you could be able to display isn’t displayable any more, because the software doesn’t know what to do with it.
JK: A lot of the digital info we have is not human-readable. As you mention, games and videos all require machinery in order to be able to experience but images if printed aren’t like that. They don’t require any wires, you just put them on the wall.
VC: Well once they are printed that’s correct, but before they are printed you do need software to correctly render them.
JK: Exactly. Which is why printing, at least until ‘digital velum’ or whatever it may be comes into existence, is a great solution. Especially with archival inks and papers. Digital files are obviously important to humanity ,and there are those that are important to individual human beings, like family photos. We’ve got pictures that are precious to us that we would run into a burning building for, but they are at a much greater risk of ‘bit rot’ than they are of catching on fire. Why do you think that people really aren’t recognising this issue and taking more action to preserve their images?
VC: Well first of all, you point out that printing is not always the solution either. What if it’s a movie? What if it’s a digital movie? It’s not likely that somebody is going to be able to conveniently make prints of a movie. In fact it’s hard to find stock to make movie film any more because everyone is doing it in digital form, which is sort of terrifying in a way. The ultimate ‘bit rot’ problem is having a digital movie and not being able to print it, literately, because it isn’t printable.. It’s not like a 4×6 or an 8×10 picture. This is a very serious problem, and I don’t know why people don’t recognise that, besides that it hasn’t been thrust upon them. People tend not to do anything until suddenly it doesn’t work and it’s too late.
JK: There’s a wedding photographer that I know of that delivers to his brides a floppy disk to communicate to them this very idea – that formats change – and at least when it comes to still images, prints might be the best way to go. One of the biggest barriers to printing images for most people is the sheer number of digital images captured today. I have three kids and a puppy. I take thousands of pictures a year. Organising and curating that many images is just too daunting a task. Is this something you see us being able to move past, and if so how and when would you see that happening?
VC: There’s several ways to approach this. There are software programs like Picasso and others… that are intended to help you organise digital images. It’s getting easier to index images even by figuring what the images are. There’s almost artificial intelligence software that can recognise an image and help you catalogue it. I’ve also seen people who go off on adventure vacations taking these dramatic pictures of Antarctica for example, choosing some of them and then turning these into small picture books that they can keep and shelve – they have to pick their best photos for that purpose.
I think that the sheer volume is a big issue because storage space required for thousands of images can really be a daunting problem, and that’s why digital is so attractive and that’s why the ‘digital vellum’ notion is important. ‘Digital velum’, if we’re successful at designing systems like this, would be preserving the render-ability of these images over very long periods of time. I think you need to solve this problem two different ways. One is by creating this digital environment that preserves render-ability, but also choosing long life formats – printing formats on appropriate stock – that you can anticipate would last for a while, just like photographs from the 19th century are actually still usable.
JK: Why is this topic so important to you. Why did you choose to speak on this issue?
VC: Well I have encountered any number of instances were older files have no longer been interpretable. I’ve been in this game since the 1960s and I’ve accumulated a lot of digital content, and from time to time you run into exactly this problem. Because there’s such an enormous amount of digital content being created today, it seems timely to warn people that if you cared about this stuff being usable in hundreds of years from now, we need to start thinking about solutions. I was thinking more about the National Archives for example, where historical records are important. What about real estate transactions that need to be retained in perpetuity for reference purposes? All those digitised media have this combination of physical medium resilience and long life, and the interpretation of those files. I’m imagining that in one or two hundred years from now it may be very important to be able to get access to those records. When someone says ‘well, you could have printed it out’, the problem with that is that the searchability of physically printed stuff is weak. Even though you can see it, you have to go through it manually. So having digital versions of these things can be very powerful and it doesn’t strike me as odd [to have] both.
You might print things for retention and you might still keep them in digital form for manipulation, management, search and cataloguing and so on. But you have to have the tools available to do that and you have to be able to retain the correct rendering over long periods of time… What I’d like to see is serious thought about how to preserve digital records for a hundred or a thousand or two thousand or more years.
JK: What do you do with your own digital images? Do you ever lost any and do you make a habit of printing them?
VC: I have lost a number of digital images which has spurred me to print a number of them, and I have a bookshelf full of printed images. The difficulty of sorting through them is really daunting because you may want to sort them in different ways – I mean this is pictures of a person or a certain time, or pictures of places. And you want to catalogue them in different ways for purposes of reference. To do that with physical pictures it means you have to make multiple copies. Now that might make some companies happy, but it takes up more room. That’s why digital is so attractive. But at the same time, you run into this long term preservation problem. I’m trying to tackle the long term digital preservation problem because of the convenience of the digital form.
JK: What would be your advice to consumers who are wondering how they can best preserve not just photos, but all of their digital content that is very important to them?
VC: At the moment there aren’t a lot of places that are trying to solve this problem… I just had an interesting conversation with two young fellows in the UK who are preserving digital information and also preserving software that knows how to correctly interpret and possibly convert various file formats, some 1100 different file formats… I think we’re going to see services like this rising in the cloud, for example, which would be a natural place for it to happen. You upload the file, it gets converted, and you get it back. Or you leave it in the cloud for reference. My belief right now is that finding a place where conversions can be done, or where software can be run to correctly interpret those files is going to be important. But this is a nascent industry as far as I can tell. There are specialty places that will go to the trouble to read things on a disk drive for example… What people should be looking for is services that are reasonable cost that are designed to preserve these data and correctly interpret them over a long period of time.
JK: …What’s your advice to our listeners and to the photography industry as a whole when it comes to sharing this message?
VC: Alarmist probably doesn’t help, although I frankly think that my talk was taken as sort of an alarmist signal…My advice, frankly, is that the industry needs to hear from people who want to preserve their data, and in your case particularly images. They need to know what options there are, and I think you probably need to tell people real stories about real people who have discovered that they have lost, in particular images, that they wished they preserved somehow. We really have to hear about the problem as a reality, as opposed to a speculation, in order to get people to recognise that this is something that they should pay attention to.