Page 32 - Loss of the VOC Retourschip Batavia, Western Australia, 1629

Figure 20. Recording the stem-post sheathing on transparent drawing film. note this is done at the same time as the photography.
Allhough Ilfo print paper was generally used for print- ing, it has been found that the new Ilfo speed paperis better for this type of work. The latter has better dimensional stability, and appears to suffer less from shrinlcage than the Ilfo print material
Underwater timber photomosaics
As each area of the ship's slructure was uncovered,
photomosaics were made to provide a record of the SIruC- ture. Mter all individual timbers had been labelled, a series of photographs were made of the area. Included in each photograph was a 1 m square or a 1 x 2 m rectangular grid frame, graduated in 0.1 m intervals. The grid frame was laid flat on the surface of the timber and a hand-held, semi- vertical photograph was talcen at a height of2.5 m, using the28mmlens,orataheightof1.4m,usingthe15mm lens. The grid frame was then turned over on one edge so that the next metre square of timber was covered with the grid frame. This was then photographed and the process repeated. By starting at One edge of the timber, a strip, 1 m wide, was photographed from one side to the other. The grid was then rotated laterally once and the process re- peated back across the slructure, to give a strip of the next adjacent metre. This process was continued until all the exposed area was photographed.
The surface layer of timber was then removed and the underlying layer was re-tagged where necessary and a second mosaic was made. This process of photography (removal of a layer, re-tagging and photography) was repeated until all the timber in that particular area was cleared. The excavation then progressed to a new area adjacent and the process was repeated. In this way, a series o f photomosaics o f each layer o f the structure was ob- tained, and Ihese could be fitted togelher to give an overall mosaic of each layer over the whole area.
The prints were rectified, as described in Green (1980). In this case, the rectification was quite suaighLforward, due to the use of a regular metre square. However, owing to the fact that the structure was curved in one plane, and almost suaight in the olher, the mosaic, although not suffering from serious matching errors, represented the surface of Ihe timber, flattened into two dimensions.
Later, when the 15 mm lens was available, it was found that, in good conditions, it was possible to photograph the timber in one single (or stereo pair) photograph. Thus photomosiacs were abandoned as unnecessary and verti- cal or semi-verticals of structure were talcen using a 1 x 2 m grid frame, in either stereo or mono, wilh Ihe 15 mm wide-angle lens.
20