96 
Bryan, Andrews 
4. DRAWINGS 
The initial requirement for the Peterborough project was to provide a set of traditional photogrammetric line drawings. These were to 
be printed on A1+ sheets at a scale of 1:20, and were to cover only the visible architectural detail within the first 7m of ceiling at the 
very eastern end of the nave - which was to be the physical extent of the first phase of conservation works. In addition a set of 
smaller A3 format drawings at the larger scale of 1:10 ( Fig. 5) would also be needed during the actual conservation works on site. 
These would allow the pinpointing and referencing of all visible structural features such as nail heads, screw threads and tie bolts, as 
well as the painted detail itself, that can only be seen when viewing the ceiling from close proximity. 
In discussion with both the Cathedral conservation consultant and Architect, the precise level of detail for the survey was agreed. 
Photogrammetric plotting was commenced during December 1996 by the Photogrammetric Unit using a LEICA SD2000 analytical 
plotting machine. All of the detail was recorded commensurate with the standard specification for such English Heritage surveys, 
although the CAD layering convention was reduced down to include only three basic layers - Boards, Fixings and Painting. During 
this initial plotting work a number of interesting problems arose. 
PANi b Ni> 3 6/11 
PETERBOROUGH OATH 
NAVE CEILING 
a.& 1 m 
8- Serico 
sufNw TBcm k pnijiBBfuwiWMiiO urn 
Gtmjj 84/^45 6 Bx^Vd 
3 nû'-iCsx -vgacT“k.b-3 
O 
Mccoaytf- «Su*!*«s. “tra, 
'■"'ffttiSriiii” 301 
Orf* | DRrfWfi Orffll 
s?*ic no,| oavyw no. 
Figure 5: Drawing extract taken from base photogrammetric survey 
The first was how best to represent each of the oak boards that made up the actual ceiling. As each board actually overlapped its 
neighbour it was impossible to accurately record all four sides, as one side was often hidden from view. Therefore it was decided that 
where part of a board was hidden by the overlap of another, the visible edges would be plotted as an unclosed 3D polyline. 
Secondly as some of the painted detail was difficult to interpret successfully as a line, particularly when viewed at high 
magnification, it was decided not to plot any of this detail other than a basic outline, and infill the remainder with a scaled 
photographic image. 
Finally some of the fixings within the ceiling itself, such as the original nails and the later inserted screws, proved very difficult to 
differentiate between during plotting. Therefore simple circle symbols were used to denote all similar features. Although each fixing 
would be recorded at the correct diameter, they would require verification by the conservators themselves, from the scaffold, when 
the actual works commenced on site. 
Using this initial plotting work as a benchmark, it was possible to provide an accurate estimate to the architect on the anticipated 
time, and hence cost, of photogrammetrically processing the remaining 55m of ceiling to the same level of detail using the existing 
stereo-photographs and survey control data produced by Photogrammetric Unit. This information has already been used in the