Regarding the storage of information, we've never had it so good. Hard disks with a minimum capacity of a Terabyte are now available on the High Street, and we can back up all of our personal information to the cloud. Gone are our concerns in terms of physical volume; we can now take a seemingly infinite number of photos, and store them on our netbook's hard disk, or send them to an unseen RAID array somewhere on the other side of the world, for later retrieval by us or anyone else.
When there is less physicality, there is less of a need to be choosy. Storing photos on a CD-ROM required some consideration as to the number and parameters of photos. When personal storage has increased at a factor exponential to the amount of content that we produced, we're less bothered. 100 holiday photos – no problem. You could get every holiday for a lifetime onto one cheap hard drive.
Storing data in this way makes it easy to forget that there are still mechanical elements involved. In a standard hard disk, accessing and adding data requires a read/write head to move across a platter which could be spinning at 10,000 RPM. Faster devices, such as solid-state storage, are still made up of manufactured parts: SDRAM chips, PCBs and the like.
While all of this brings spontaneity to the user, it is actually creating a problem. A problem that is currently overlooked by many, but is likely to grow year on year, generation to generation. That problem is one of storage and archival.
Data storage onto mass RAID arrays, or even onto smaller media such as CD-ROM, is prone to quality control. We know that through hard disk corruptions and CD misreads on our own computers. However, if that hard disk or that CD is kept for over a century, it is likely to come out as damaged. This makes the original content, even if it was carefully curated for the purposes of storing, worthless.
Graham Diprose is a former lecturer in photography at the London College of Communication and recent co-author of a book on Henry Taunt, an Oxford-based photographer from the Victorian era, whose tireless work along the full stretch of the River Thames inspired many authors and fellow photographers to produce work which also celebrated river life. Taunt's silver gelatin photographs, covering the source of the Thames at Lechlade all the way to London, have been stored through the years with a level of care and consideration sufficient to allow their reproduction in Diprose's new book. As the book required Diprose and co-author Jeff Robins to retrace Taunt's steps from one end of the Thames to the other, their challenge was to store their own, new, photographs in a way that encouraged long-term archival without degradation of quality.
The challenge to Diprose's thinking on digital storage arose when English Heritage asked him for his new photographs. It led him to wonder what the future for his CD of images would be, given the lifetime of 150 years (so far) that Taunt's gelatine prints have enjoyed. It may even be the case that Taunt's photographs could go on for another 150 years, something that the CD is unlikely to achieve. Diprose and colleague Mike Seabourne decided to find out the best way to store data in order to maximise its potential to still be accessed in the next century.
They started with the notion that papers of 150 years ago were in good shape now, suggesting that paper remained a good medium for storage. They set about printing a large number of pages at high resolution onto single pieces of paper, meaning that 64 or even 128 pages of content could be stored onto an A2-size sheet. Retrieval would then be a case of scanning them back in. This reduces the impact on storage facilities, as there is less material to store.
Once that was settled, Diprose and Seabourne started to think about how the media treated data. Even with the best magnification loupe, it was hard to see how different papers treated the photograph. "You really couldn't tell from looking at the images, about whether the dot was bleeding, whether it was dried up and not making a proper-shaped dot, when you were printing it. That's where we got the idea."
Diprose was writing a text book at the time. He struck upon the idea of taking 64 or 128 pages from the textbook, turning them into a PDF, printing them out into rows, and using a measurable way of comparing one method of printing to another.
Their method to measure this quality was to the printed text back into an OCR device, and to see how many mistakes it made. An indirect benefit was that they also learned about the preferences of OCR devices in terms of font: "We could tell that Times New Roman with serifs was bad news, Arial was good news, and so on; but Century Gothic, which you would think it would like due to its round letters, it didn't like as much. So we learned some things about typefaces. We ended up settling on Arial, because it was font that most computers had."
Once the font was settled, they experimented with different papers and printers, donated by suppliers for the project. Their test became structured: "We had this lovely test where could take 64 or 128-up on a page, stick them under the OCR, and count the number of mistakes in it. That gave us a result which was Canson photo rag paper, which proved sharper than any other. On 64 A4s up on a sheet of A2, it got everything right."
There is a reason. This particular paper has a layer of potato starch between the baryta coating, and the top surface that takes the ink. In many papers, the ink sinks into the baryta, and bleeds. Cansen's paper did not, and gave the best printed dot with their chosen printer, the HP Z3100. They could continue their experiments, knowing that the paper and ink combination was the best that they could find.
These three components – the printer, the paper, and the way in which the ink dot was printed, were the most important facets of the experiment, with the successful combination giving a printout that could last for 300 years.
It was the quality of the dot became critical to the results. "My background is an advertising photographer, not a printer. I was on some new ground, but it made a difference. Putting the same paper through the same machine and you could get such different results really quite surprised us. Some matt papers bled; When you were looking at images you didn't really notice it, but when you had tiny type of 64 pages of A4, you could not read it, even with a magnifying glass. Sometimes we would look and say it's very pretty, but not as sharp as it could be."
"That's when we went back to putting 16 up on A2, 8up, 4 up, and seeing how much was lost when recopied using a very high resolution digital camera or a scanner. With the Cansen paper, you can get 16 up on A2 because many printers can't print on A1; that's 32 up on A1. That's a reasonable amount for a museum to pay for, and they could set up a file to drop the images in automatically, print the material out, store in large A1 flat cases, and that would do the trick."
It caught the attention of the Museum of London, who invited them to run a whole collection of theirs through this technique, which is where their attention now lies.
On the night of the launch of Diprose's exhibition ... in the footsteps of Henry Taunt in Henley-on-Thames, sponsored by English Heritage, Diprose had a surprising encounter with his sponsor.
"There I was, standing up and saying that they only way you're going to get your digital pictures down 100 years or more is to actually make prints of everything. Someone from English Heritage came up to me afterwards and said, 'We had planned not to build any more cold stores. You do realise the implications of what you're saying!' ... because they thought that they were going to stick everything on RAID disks and it would all be fine. I don't know if they will build any more cold stores, but they did take our files as A3 digital inkjet prints which, according to the Wilhelm Institute, have a life of 250-300 years."
With Victorian photography, many prints have not faded. Ironically, this was less to do with preservation, than the commercial environment at the time. Photographers would compete with each other for family portraits, on the promise that their prints did not fade, but the prints from competitors with less of a quality assurance, did. Henry Taunt's impetus was just that: he was driven by selling commercial guidebooks of the River Thames. What was important to Taunt in terms of storage was not the retrieval of a perfectly-stored photograph after 100 years, but ensuring that his customers did not return their prints to him, complaining that they had faded.
The issue of English Heritage not considering the development of more cold stores for storage, is symptomatic of a much bigger problem: of how data is preserved, and how legacy data is treated by future generations.
"We have a question that nobody is addressing or thinking about. Everyone assumes that it will be all right. You're trusting your grandchildren to choose your files and your things to migrate to the next generation, when there will be a lot of other things which have come along by then. If we want to be in control of what we send down, then we will have to send it down in a format that they will treasure. An artifact in their hand. If you have a picture of Henry Thompson in your hand, then you're not going to tear it up or stick it in the bin, whereas it's all too easy to chuck a digital file in the trash or delete it, because you want to make a bit more room on the server. If you have a picture that was taken 100 years ago you look after it; you're far less likely to bin it."
Diprose sees a number of applications, both artistic and commercial, for storing information on high-quality paper. They are currently considering working on documents such as 99-year leases. For obvious reasons, these need to be preserved in order to provide a legal and social continuity. The amount of documents that require this treatment is not inconsiderable, which then raises a new problem: how should we choose which documents are stored, if we – once again – have physical restrictions on storage? "You can't go into a museum and print out everything. With a museum containing 500,000 pictures, they won't do it with all of them. There is an opportunity to curate what we, as a society, as museum curators, as individuals, think is important: our work, our company's work, a particular photographer's work. We are taking more pictures than we have before; but less of them will be around in 10 years' time."
The actual method of storing is identical to how museums are currently storing printed matter: a cool, dry room, with the images laid flat. This latter nuance is as important as the quality of the printer, paper, and ink dot.
Instead of individual flat A4 sheets, the project also looked paper rolls, which could cater for 32 or 64 pages per sheet. Although this may be more convenient, it actually creates new, specific, problems.
Unrolling old scrolls and paper rolls is difficult, as being rolled up for so long may lead to the paper cracking. For inkjet printing, which uses glycol-solution ink, research is currently taking place regarding what happens to this ink solution after it is sprayed onto paper. Diprose cites the project, saying that the ink "...continues through the back of the media, and attacks the surface of the next print. It's touching it.
"So, having papers rolled up would be a very bad idea, because you could damage the surface of the next print. By laying them flat, you would need to have a sheet of something impenetrable, such as crystal parchment: something inexpensive, but that the inks would not soak in to. We did start with the 'roll' idea and rapidly went off it, once we realised what the problems would be."
Diprose believes that there are some organisations that are starting to explore these alternative methods of storage, but most are still buying giant RAID arrays and not considering the implications. When museums are cutting back on funding, this method of storage is more expensive initially. However, it is the current mechanism that may prove to be more costly – in many different ways – over time. The equipment will require preservation; the data formats will need to be understood; and curators and archivists will carefully need to consider the implication and value of what is lost, if some files are deleted now in order to free up space for "more important" content to archive.
The interest of a well-established museum such as the Museum of London opens up new possibilities for this work. It is feasible that commercialisation is possible, such as with the possibility of archiving legal documents in this manner. What is important for museums, as much as commercial organisations, is that what is archived is the right material. The forthcoming Olympics is a good example of this; there will be a huge number of excellent photographs taken on behalf of publishers, news organisations, galleries and museums. If these photographs have a future, then they need to be chosen in terms of their value to future generations, rather than their inconvenience of occupying server space which could be used for something else.
Reframing digital content in terms of its long-term value may lead to some very different choices in terms of its personal and social curation, as well as change the way that we store. There is as much value in how the material is stored, as the material itself.
"Digital photography has made things so transient, that everyone thinks that their pictures will still be around if they store them very carefully on a writable CD which they have written to at best 52x - which won't be around in 3 years, particularly if it's a cheap one.
I don't think that [long-term archival is] something that occurs to people. They think that they have saved all their pictures, then their computer hard drive doesn't work one morning. People find out the hard way."