Treating the Traité

From Mondothèque

Revision as of 14:56, 5 February 2015 by FS (talk | contribs)

Experiments with digital iterations of The Book on the Book.

Scan Tailor

Tomislav Medak spends two days with us to demonstrate a workflow for digitizing books. I use the opportunity to look at the Traité through the lens of Scan Tailor, "an interactive post-processing tool for scanned pages"[1].

First, I import the image files exported from the pdf into Scan Tailor and let it treat the Traité with all options set to 'automatic'. It produces exciting artefacts:

Printing the Traité

The Traité de documentation : le livre sur le livre, théorie et pratique is an almost hypertextual book on documentation, written in the 1930's by Paul Otlet. It has many cross-references, tables and illustrations; at times it is written in encyclopedic style, turns into a passionate manifesto, speculative fiction, and a practical manual for librarians. The pdf I have is badly OCR-ed and too heavy for reading comfortably on a digital device. So this morning I transformed the digital version into something that I can print at a copy shop.

I started with extracting the images from the pdf with the help of the imagemagick convert command:

$ mkdir spreads

$ convert Traite\ de\ documentation\ -\ Paul\ Otlet.pdf spreads/%03d.jpg

Next I removed front- and back-cover (they will be treated separately), and also 113.jpg (pages 118-119 are repeated), then cut each spread in half:

mkdir pages

convert spreads/*.jpg -crop 2x1@ pages/%03d.jpg

The properties of the original pdf mention a paper size of 200 × 260 mm (and also that the file was created with ABBYY FineReader on Monday December 3, 2007 16:25:51 CET (This file is already 6 years old ...). I am not sure if the measurements refer to the size of the spread or the single page, but from the detailed description in the catalog of the Universiteitsbibliotheek Gent [2] I gather that pages are 26cm high, and will fit comfortably on an A4: 431, [12], viii p. : illus. ; 26 cm.

I then simply put all images back into a new pdf:

convert pages/*jpg traite.pdf

Tomorrow I'll have the document printed and bound. Can't wait.

Transcribing the Traité

in progress on Wikisource

https://github.com/PaulOtlet/traite http://traite.czam.de/en/latest/otlet_traite_1934_FR.html#i-buts-de-la-documentation

Sources

Original scans http://lib.ugent.be/fulltxt/handle/1854/5612/Traite_de_documentation_ocr.pdf

OCR https://archive.org/details/OtletTraitDocumentationUgent
  1. http://scantailor.org/
  2. http://lib.ugent.be/catalog/rug01:000990276#reference-details