You are here: Home Blog Creating a Digital Archive

Creating a Digital Archive

by Dawn Chatty May, 2011

Thirty years ago I had the unique opportunity to live in the Sultanate of Oman and work for the United Nations researching and then putting into practice a set of projects to improve the lives and livelihoods of nomadic pastoral camel and goat herders in the central desert of the country. After an exploratory trip to the Jiddat il-Harasiis in 1978, I returned to Damascus, where I was living at the time, and sought professional advice as to what kind of film stock would most ‘truthfully’ record the colours of that desert landscape. I was advised by an American photographer to use ‘inter-negative’ Eastman Kodak film; this would give me deep colour even in the blindingly strong sunlight. For the next five years I captured an ethnographic record on this Eastman film. Near the end of this period, I began to see that some of my early images were losing colour density. I immediately shifted to a different film stock; mainly Fuji chrome and Kodachrome.

From time to time over the past few years I had reviewed my slide archive and noted with a deepening despair that many of my slides were shifting colours or losing density. I began to explore the possibility of scanning some of these images so as to preserve them as an ethnographic record. An opportunity to digitize a selection of 200 of my most significant images arose when the Bodleian Library wanted to test out some new equipment. This gave me a taste of what might be possible, but the accumulated dust and colour imperfections in this small scanning project left me dissatisfied with the outcome.

Late in 2009, I decided to look for funds to scan 2000 of my best slides and post them on a website so that they could be available widely for others to view and use. Without any prior knowledge of what was involved and no one to turn to for advice – I was not aware of any similar initiative - I estimated a figure of £10,000 as sufficient to accomplish what I wanted to do – purchase a slide scanner for £7-800 and train a graduate student to scan and colour correct the images. I calculated 10 minutes per slide or about 340 hours at £12 an hour would suffice. That gave me a sum of about £5000, which I then doubled to include the costs of a web designer and for purchasing the necessary software. I thought I knew what I was doing, but in hindsight I had no idea of what I was getting myself into and how poorly informed I was for the task at hand.

Then I did something right. In January 2010, I enrolled in two courses at the Oxford University Computer Services (OUCS): one an introduction to Photoshop; and the other on correcting and improving images presented by Carl Wenczek. That was when I realized that image correction was much more complicated than I had ever imagined. I asked Carl’s advice and was directed to the ‘Computer 8’ drop-in sessions at OUCS and met with David Baker, a specialist in teaching communication anddesign. We sat together and talked through my ideas for a website. That discussion further brought home to me how little I knew and how much I needed good advice. David agreed to help me get on my feet in this new world. And each step of the way he guided me through both jargon and important concepts to reach the clean and polished outcome that is www.nomadsinoman.com.

From the very outset David felt that my budget of £10,000 would be much better used if I had my slides scanned professionally rather than purchase an inferior machine - I couldn’t afford a high quality professional scanner on my small budget - and train a student to undertake the labour intensive scanning process. I asked for his help in identifying some possible professional outfits to do the job and he gave me a short list from which I selected three.

I wrote to each of them and asked what they would charge to scan 2,000 35mm sides to the highest dpi available (2400 to 4000) and in TIFF format. I also asked whether they would do a ‘pre-test’ on a selection of about six slides. And finally I asked what they would charge and how long they would take.

Each company replied with slightly different estimates ranging from 30p to 87p and more per slide and different scanning resolutions. But each company raised different issues. I was unclear as to what resolution I should be aiming for – anything between 3-600 dpi is all that I had ever used in the past. David suggested that the highest dpi [4000 +] and colour depth [48 bit instead of 36 or 24] would see the project in good stead, as it would remain a viable format for longer as technology raced on. Having little sense of what all these numbers actually meant, I likened it selecting the fastest broadband possible today so as to more easily upgrade in a few years time thus giving the project the longest archival value possible.

One company suggested that if the slides were ‘mountless’ or in strips then the estimate would be lower. Another suggested that if I required ICE it would cost more and take longer. Clearly I didn’t’ really understand what scanning entailed; it was obviously more than just laying a slide on a flatbed and pressing a tab! A quick email to David revealed that ICE was image correction and enhancement software that removed dust and scratches [an obvious necessity].

From the email exchanges, I was inclined to accept the ‘middle’ range offer of one particular company. However once the test-slides had been returned from each of the three companies, I sat down with David to review their work. All of the images ‘looked’ good. One company, however, had scanned and colour corrected the images to what they thought would be appropriate but jarred with my own recollection of the desert colours. David, however, had another test in mind and looked for more than colour authenticity. He was interested in seeing what image bit depth had been used. An examination of all the images scanned by the three companies led us to select the clearly superior one for the purposes of this project.

While awaiting the return of the test slides, David encouraged me to think through how the slides were to be identified – what metadata should be attached to each slide and what descriptive terms- key words - would also be attached to each image. This also meant that I needed to give some thought to how I wanted the website to work – was it to be only an archive of slides searchable with a core set of key words; would there be any predetermined ‘slide ‘ shows’; any written material about the subjects? At this point I contacted John Pilbeam who was the manager of FMO, Forced Migration Online, to help me locate a web designer and manage the construction of the website for me. David also suggested I get hold of an important reference book on managing visual images, The DAM Book: Digital Asset Management for Photographers by Peter Krogh and start thinking about workflow, hardware and software needed to see the project to completion. For the hardware (computer, screen archival storage boxes and cleaning equipment) I turned to Debbie Usher, the Archivist at the Middle East Centre’s Archive in St Antony’s College where a scanning effort had just been completed on a collection of images from Oman taken in the 1960s and 1970s by a British seconded army officer. From Debbie, I determined the specifications for a computer to hold this collection of images, which would reach nearly 1 TB once backup copies of the collection were included[1]. I ordered my archival slide sheets and storage boxes as well as cleaning equipment from the same company she had used – a decision I never regretted.

Before I could send the slides for scanning, however, I had to remove all the glass mounts – I had used glass mounts for nearly ten years, believing I was preserving my images. However I soon realized that those mounted behind glass were the ones that had the greatest colour deterioration. I also had to sort and throw out duplicate images and insert consistent metadata in the remaining set. I decided to identity each slide with the date (year), place, photographer’s initials and then a unique number. My original collection consisted of 4,000 slides. I had often taken duplicate images with the idea of using one set for slide shows and keeping the second set in safely stored in metal boxes to slow down any possible colour deterioration. Over the following month long period, I manually removed dust, and remounted slides in plastic casing while assigning each with its unique metadata.

Finally in March 2010, I was ready to personally deliver this collection to my selected company – requiring a trip around the M25. I had agreed that the images would be scanned to the following specifications: 4000dpi
Adobe RGB (1998) colour
Uncompressed TIFF format
48bit colour depth

Post scanning process would include:
Cleaning at 100% removing extraneous marks and dirt
Cropping and rotation when necessary

Naming convention was to be: Date_place_photographersinitials_unique number.tif
Images were to be saved on multiple DVDs
Images were to be held on the company servers until they had been successfully copied onto to my own computer and on to the University Archive.
The scanning would be completed in 8 weeks and would be charged at the rate of £1.37 per image.

With the slides now being scanned, I needed to turn my attention to the next stage in this project: what software to use to colour correct the digitized images and how to locate and train a graduate student in the work to come. After discussion with both David and Carl, it was agreed that a combination of Lightroom and Adobe Photoshop would be more than adequate for the job. I ordered the software as well as ‘i1Display’ equipment and software that colour calibrates the computer monitor. I also negotiated a one-day training session for two graduate students (at this point I considered one as a backup for the other] and began to envisage the workflow and timetable. I had hoped to launch the website in early November so set a ‘deadline’ for early October. John Pilbeam had by now located a web designer and we began to discuss what the website should look like and how it should function using Plone 4 technology.

I advertised for two graduate students and also spread the word in the visual anthropology department. I selected two students who have very different skill set and began to see the work flow in a different manner. One student seemed not to mind the possibility of a ‘repetitive set of tasks’ over the working day during the summer inserting key words. The other, already a skilled photographer with a remarkable fluency in using Adobe Photoshop, preferred to work evenings and on weekends. That combination of students meant that one could insert key words during the day (I had already determined the key words during the process of hand cleaning and remounting the collection) and the other would colour correct in the evening and over weekends.

In the first week of June, I collected the scanned images on 57 DVDs from the Max Communications Company. The following week Carl and David conducted an all-day workshop for three students (one back up should either of the first two drop out). And by late June, they began processing the sides, inserting key words, colour coding and identifying 4 star [****] and 5 star[****] images. This was an independent exercise to identify the visually most striking images to use for the two website slide shows. By mid August this task was done. Together we identified a slide show from the ***** images depicting the changes in the life of the community over the past 30 years. In addition, and as part of the recognition that what happens behind the scenes also has value, we created a smaller slide show of largely ***** images which included the anthropologists, doctors, nurses and their families who had worked with this community over the intervening years.

The website domain name had been purchased at the very beginning of this process and a server had been negotiated along side the development of the website. Finally in early October 2011, the prepared digital images with their meta data and key words were uploaded to the website along with the pdfs of written text regarding this way of life. There were a few hitches to iron out and on November 18st the website was officially launched commemorating the 40th National Day Anniversary of the modern Sultanate of Oman. If I had to do it again, I would have sought a grant of £20,000 to fund the project; but I would have followed the same path which I was guided along so expertly. My thanks go out to David Baker, Carl Wenczek, John Pilbeam, David Cordery at Max Communication, Jaanika Vider, Shama el-Salahi, Miranda Mylne, David Little, and Chris West and Arthur Konstanczak at Fry-IT.


1. The exact specifications were: Dell Precision T3400 Core 2 Duo 375W E6750 2.66 GHz on 1.33 GHz bus, 4GB RAM, 22" widescreen LCD Display, 1 TB Hard disk, 1 TB Secondary RAID 1 Hard disk, 16x DVD +/- RW.