The digital Classicist seminar on Friday 3 July: Roger Boyle & Kia Ng (Leeds)
‘Extracting the Hidden: Paper Watermark Location and Identification’.
This seminar was given by two scholars from the University of Leeds presenting their project to locate and identify watermarks in paper. Their technique involves 'back-lighting' and scanning to obtain digital images which are then processed to reveal the hidden watermarks.
They have two approaches: a 'bottom up' method with a pixel-first approach followed by image manipulation. Alternatively they create a computational model of what the back-lighting actually does using a 'top-down' approach which proved more successful in identifying details of marks that had not been previously seen with traditional methods (using the hand and eye).
A full and lengthy discussion followed around the specifics of their project and the possible application of these techniques in other related (and non-related) areas.
The audio recording and side show are available on the Digital Classicist seminar webpage:
http://www.digitalclassicist.org/wip/wip2009.html
Re: Roger Boyle and Kia Ng (Leeds) ‘Extracting the ...
I've now uploaded the audio recording of this seminar (apologies for the audio quality of parts of this recording, which was due to a poor match between microphone location and the seat of the chairperson), along with the updated presentation slides to which Roger kindly added the publication information for the research presented here.
One of the most interesting aspects of this presentation, for me, was that--although the primary focus of research was early modern (especially musical) manuscripts, and features such as watermarks and chain lines that Roger and Kia's research don't often occur in Classical manuscripts (certainly not in ancient documents)--the general approach to studying unique traces of paper manufacture is applicable to a whole range of documents within manuscript studies.
Of course parchment is probably denser and thicker than manufactured paper, and so less amenable to digitization via backlighting (and imperfections in the surface are likely to owe more to the uniqueness of animal skin than the paper-makers technique or apparatus). And papyrus, being hand-made, would give unique sheets in any case, so digital comparison of backlit scans of individual sheets would be unlikely to allow the study of distinct manufacturers' techniques.
But combined with other techniques for studying the physical nature of manuscripts, such as multispectral imaging, laser scanning, and CT/MRI volumetric scanning (as our colleagues in Kentucky described for us in two paper [Seales; Baumann] in this series last year), this work enhances our arsenal for the detailed physical analysis of the textual and non-textual features of manuscripts.
==
GB