Digitizing the ANC Archive Mountain One Manuscript at a Time

For over a year Africa Media Online has had a team resident in Alice at the University of Fort Hare to continue the process of digitising the archive of the African National Congress. The project is a continuation of a project that was initiated in 2011 by the ANC ahead of their 1ooth Anniversary. Funded by Multichoice, the digital archive was launched by President Zuma in 2013. That project, however, only saw a small portion of the archive digitised. The dream was always to have the entire collection digitised, a dream that began to come to fruition in 2015 when Multichoice gave the go ahead again for funding a further phase of the digitisation of the collection.
The 2011-2012 project had only been able to accomplish so much. The team of service providers managed to digitise most of the video, most of the audio tapes and we managed to capture about 24,000 images, a similar number of manuscript pages, and a couple of hundred museum objects. That, however, was only a scratching of the surface when it came to the manuscript collection. At the time of putting the proposal together, based on the finding aids we anticipated there were in the region of 800,000 manuscript pages in the collection. We had captured just 3% of those pages. As it has turned out, however, we had probably only captured just over 1% of the archive.
When we relocated the team to Alice in November 2015 the first challenge we encountered was the fact that the manuscript collection had only been organised down to the folder level. The archive is well organised into collections, series, boxes and within each box are folders. All of those levels of organisation had numbers on them, but when you opened a folder, the items inside did not have an item number. Digitisation could not begin until we had organised down to that level. And as we started to do that, we discovered that there were very many more manuscripts in the collection that was calculated at first. Over a year later we are now more than half way through creating a detailed inventory and dividing the contents of each folder into three workflows – plain paper, fragile manuscripts and bound manuscripts – and by our calculations there are close to 2 million pages in the archive.
When we first took on the digitisation of the ANC Archive manuscript collection in 2011 and 2012 we were using flatbed scanners. The quality was fantastic, but it was really slow work. It may have worked for the 24,000 pages captured in that phase of the project, it was not, however, going to work for this phase. The vast majority of the manuscripts in the archive are plain paper that can be captured on a sheet-fed scanner. For that we sourced the German-manufactured Inotec Scamax machine whose belt driven mechanism is very gentle even on archival pages. For really fragile documents, however, we are employing our Alpa camera with Phase One back on a copy stand. And for bound manuscripts we have brought in a v-cradle capture device that utilizes full-frame Canon cameras.
We have owned an Atiz machine for some years. For this project, however, we felt we needed to make some improvements to it. We installed additional lights to improve the even illumination of the pages. We invested in high quality Zeiss Milevus 1.4/50 mm lenses for the cameras. And we replaced the perspex v-cradle “glass” that holds the bound manuscript flat with hardened glass. We also worked out how to set the cameras to capture at exactly 300 dpi in line with Metamorfoze standards. So far so good. We are aware, however, that it is no small task and we expect to have in the region of 420,000 bound manuscript pages to capture!