Archives of the future

Reading Time: 6 minutes

It is certainly a truism to say that we’re taking more images than ever before, the majority of which are created and stored digitally. RAID and cloud storage services are widely used by individual photographers, but for gallery and museum archivists these approaches don’t offer 100 percent security in the long term.


Digital data stored on RAID multiple disk drives must be migrated every few years as drives wear out and need to be updated and replaced, while cloud storage – no matter how safe it is purported to be – is no substitute for a hard copy of an image. There is also the issue of human error that may come into play, or the possibility that a future archivist may delete important digital files to make room for others without properly looking at what is being deleted. Ultimately, no matter how safe current digital archival approaches are claimed to be, there is no guarantee that the files stored on hard drives will survive 50, 100 or more years into the future.

Two industry professionals, Graham Diprose and Mike Seaborne, have come up with an archival method for storing digital images that they hope will become a legitimate alternative to existing digital storage options. In a paper due to be published later this year, they write that their approach involves “selecting and sending our most vital digital photographs and documents forward into the 23rd century as inkjet print ‘artefacts’ rather than as digital data”.

Selected digital images are reduced to half or a quarter of their original size, printed using pigment inkjet technology on A2 or A1 paper sheets and stored flat in plan chests with sheets of inert, acid-free paper placed between them. These images can then be reproduced, enlarged and printed using either a high- resolution digital camera or scanner technology in the future, with a database set up to catalogue them.

“Ten years ago, the life of digital inkjet prints was around a third of the life of a darkroom print,” they write. “Now, certain combinations of acid-free inkjet paper and pigment inks can claim a possible life of 250 years or more.”

The method works on the basis that not every digital image needs to be saved and stored as a high- quality digital file. What would be more efficient, they argue, is to select “the most important” images to archive as reduced-size hard copies so they can be referenced at a later date. Contemporary curators will have control over which digital images are sent down into the future and at what size.

“Curators will be able to send the most important images down in time at a larger size, while reference or contextual pictures can be smaller,” explains Diprose.

“At present, every archive is trying to save ‘exhibition-quality’ 60MB TIFF files of everything. But to send a huge mass of data down in history is asking for trouble. If everything is sent down as 60MB files, there will be no time for our great-great-grandchildren to wade through them all and decide what is significant or worth keeping. There is a risk that vast swathes of images might be wiped off because there isn’t any more room on the servers.”

Many, if not all, major museums, picture libraries and archives will already have their own systems in place to tackle the issue of digital storage, most likely using RAID drives. Diprose and Seaborne’s intention is not to undermine existing practices. Rather, they see their approach as an addition to existing digital storage strategies.

“We’re not suggesting using our method instead of RAID drives – it would be very wrong of us to say that,” explains Diprose. “We’re the insurance policy, if you like. RAID drives can go wrong and the same RAID drive motor won’t be spinning in 50 years’ time.

“You also have to factor in changes to technology. It may be that the same technology won’t be around in 100 years’ time, so you may have a situation where images stored on drives can’t be accessed because the technology is obsolete. Whether cloud storage is as safe as people assume also remains to be seen,” he adds.

Paper approach

The idea first came about in 1997, when Diprose, Seaborne and Charles Craig began work on a project to produce a contemporary photographic record of the banks of the River Thames for the Port of London Authority (PLA), based on a 1937 silver gelatine panorama. Shooting on 6×17cm Fujichrome colour film, the team wondered if their version would last as long as the original silver gelatine image.

When they returned 10 years later to shoot the same project digitally, they began a series of experiments to see if they could find the optimum paper and ink combination for archiving their work. “We had no idea how long the digital files would survive, even with expert curation,” says Diprose. “There was every possibility that the 1937 panorama would outlast our digital files, so we started to think about alternative approaches to archiving digital images.”

The team used an HP Z3100 digital printer and a range of papers, including Fujifilm Fine Art Fibre Baryte Matt 310gsm, Harman Crystaljet Luster RC paper and Canson Infinity Rag Photographique paper. They experimented with reproducing a selection of their own images and those from the PLA’s collection at various sizes on A2 paper – for example, printing both monochrome and colour images at four, eight or 16 to a sheet (approximately A4, A5 and A6 size respectively). The quality was then compared. As they write in their paper: “While the [rescanned or rephotographed] images may not necessarily be a perfect replica of the original, this project offers hope that a usable, reproducible image will have survived.”

“You probably wouldn’t want to put a [re-enlarged] ‘16-up’ image on a gallery wall but the quality is good enough if you view the image on a tablet,” says Diprose. “For a future historian who wants to know what a place looked like in 2013, it will exist as a visual record. In any case, the way we view most images in 50 years is likely to be screen-based, rather than in an exhibition space.”

To measure accurately which paper produced the best resolution, the team used a method of printing small sections of text, which were then analysed for errors. “We had to find a way to measure which paper produced the sharpest dot, so we printed out and compared sections of a book I was writing at the time for Thames & Hudson,” explains Diprose. “If you take an area of text and scan it using a flatbed scanner capable of optical character recognition, you can count how many mistakes there are in the text. If the dot isn’t formed properly then the scanner won’t be able to read it [and will produce an incorrectly spelled word]. The more mistakes there are, the worse the dot.

“We tried lots of different papers to see which would give us the best result and found that on the Canson Infinity paper the dot is about as perfect as you can get. We can prove this by showing that on this type of paper the text can be read by the scanner.”

The estimated cost of printing 2000 images at 16-up on A2 paper is £3000, or £1.50 per image as calculated by the authors. Costs could be reduced slightly, they say, by printing out double the number of images on A1 paper.

Gaining support

Reception to the proposal has so far been largely positive. The authors have already had interest from major archives, including the National Archives based in Kew, London, and they are currently running a pilot with the Cass East End Archive, an online digital resource documenting life in the East End of London.

“The willingness of the likes of the National Archives to invite us to present an alternative [archiving] idea says everything without saying anything,” says Diprose. “They started out being quite sceptical, but by the end of our meeting they were looking at possible applications of [the technology].”

The duo’s hope is that gallery and museum archives across the country and farther afield will gradually adopt their approach. Consequently, they are currently concentrating their efforts on approaching curators at larger institutions. Their next meeting is with English Heritage.

“At the moment, our focus is on speaking with galleries and museums,” says Diprose. “I’d like to speak to the Victoria & Albert Museum and the Science Museum, as well as government archives and so on. We want to get head curators of large archives [on side] first. What we’re really looking for are people’s ideas and opinions.

“If the National Archives had shot us down in flames, we would have retreated sheepishly into the sunset. But the fact that they are interested is [promising].”


Diprose and Seaborne also intend to discuss possible future collaborations with Canson and hope to work with suppliers and manufacturers to develop their idea further. “There may be other applications of the technology – we’re exploring options at the moment,” says Diprose. “For us, it’s about getting the word out. The idea is to help archivists get started so they can use the technology themselves.

“I have no doubt that institutions such as English Heritage and the National Archives are doing everything they can to ensure their digital data is safely preserved and can be viewed in 200 years’ time,” he adds. “But there will be lots of smaller archives, galleries and picture libraries that don’t have the knowledge or resources to archive digital images for the future. To do what we’re proposing – save and store images as hard copies as well as digital files – could be a very good way to protect against the disasters and mishaps that can happen.”

The paper, An Alternative Approach to Conserving Digital Images Into the 23rd Century, was first published for the Electronic Visualisation and the Arts (EVA) London conference in July 2011 and can be downloaded from the British Computer Society website. It will also be published in a new book, entitled Electronic Visualisation in Arts and Culture, by Springers International later this year.