|
The digitisation of the British Library’s
best copies of the first and second editions of Chaucer’s
Canterbury Tales was undertaken by a team of 14 Japanese
men and women, experts in a variety of fields, a professional photographer,
bibliographers and IT specialists, all from the HUMI project at
Keio University in Tokyo, directed by Professor Takamiya, a specialist in
15th-century English
literature, and managed by Masaaki Kashimura.

Before the books were photographed they were carefully
examined by staff from our conservation department to set the appropriate
standards for handling.
Keio University also digitised our two copies of the Gutenberg
Bible, the two projects amounting to a major benefaction to the
British Library, now commemorated on the Library’s tablet
of benefactors.

The camera which the team from Keio used for the
Caxton digitsation was a Kodak DCS Proback Plus + Mamiya RZ67 digital
imaging system (One shot, 1-CCD, 16M (4000x4000) pixels, 12-bit
RGB, IEEE1394 I/F).
All the equipment was flown in from Tokyo, from the single-shot
Kodak digital camera, which every three seconds can produce a 16-million
pixel image, the battery of computers processing the images, down
to stepladders and sticky tape.
As we were dealing with very precious items we of course imposed
special conditions of treatment.
The British Library’s Conservation Department specified the
degree to which the books could be opened. This meant that the books
could not be photographed flat on their backs, and this caused a
number of technical problems.
 |
 |
| First
edition opened at the maximum permitted by our conservation
department. |
Second
edition opened at the maximum permitted by our conservation
department. |

A cradle was specially designed in Japan to our
specifications, originally for the Gutenberg Bible project, but
modified to suit the exact dimensions of the rather smaller Canterbury
Tales.
Fundamentally it consists of three separately moveable parts, which
can be gently manipulated so that the book is never open more than
at 120 degrees and yet can present to the camera a nearly flat image
of each page.

As the British Library did not allow the camera
to be above the book, it was a challenge to ensure that the page
is parallel to the camera lens. The solution turned out to be mechanical:
a square mirror carefully placed on the page and suspended from
a central cord. If the square within the camera lens aligns with
it, the camera is in the right position.
The structure of the book moves gently as the leaves are turned,
so this procedure had to be repeated every five leaves, thereby
also allowing adjustments to be made for the slowly increasing distance
between the book and the camera.

The image once photographed went through a first
processing in the camera itself and then through a battery of three
laptops.
It was then inserted into the sequence
of the other images, given a permanent file name corresponding to
the leaf number and a final visual check for clarity and for true
colour. It was important that as much as possible of the processing
and checking of the image quality took place immediately, for once
the team and equipment had gone back to Tokyo it would be nearly
impossible to rectify any errors. Each page carries a number which
accurately records the printed leaves in the British Library's copy.
The process from shooting to last check took on average 23 seconds.
The whole digitisation of the two editions took 10 days. Because
the team was generously staffed, the speed was ultimately determined
by the physical process of turning the pages and adjusting the book
in front of the camera.
Tell me more: |