28 
readily. Availability, however, is not everywhere secured, copies and en- 
largements are more costly. These disadvantages also apply to colour photo- 
graphs. 
Colour images, like CIR, provide the capability of differentiating by 
colour hue; their “natural” colours can be an advantage in specific inter- 
pretation tasks. Lo (1971) states that “colour aerial photography appears to 
have the great advantage of recording clearly the dynamic aspects of the urban 
environment, ie growth, redevelopment and movements”. Research should 
try to establish the limites of validity of this statement. 
A neglected field of study is the oblique aerial photograph. Its low cost 
and timelessness (Moore and Wellar 1969) suggest that urban planners should 
pay more attention to obliques as a data source. Moreover, obliques contain 
some information that is more difficult or impossible to extract from vertical 
photographs (number of floors, views of facades). For monitoring development 
of construction (or demolition) activities, some cities already use obliques 
systematically (figure 7). 
Traditionally, flight specifications are more geared to photogrammetric 
than to interpretation requirements. In our experience, shadowless or soft 
shadow (black-and-white) images provide a definite interpretation advantage 
in densely built-upt areas (figure 6). Also, a larger overlap and sidelap result 
in enhanced utility as a data source, in particular when combined with a wide 
angle lens. Adequate attention should therefore be given to the formulation 
of specifications really adapted to the urban user’s needs. 
Whereas professional interpreters will generally prefer wide angle or 
super wide angle photographs with an 80% overlap, it can be expected that 
urban agencies will give priority to topographic information in the near future ; 
they will therefore request photographs of very large scale (1 : 2000, 1 : 1000 
and 1 : 500) with a minimum of terrain and roof surface hidden by relief 
displacement. These images may be standard-produced by enlargement of 
high quality negatives from 3 to 10 times, the negatives to be produced by 
lenses of long focal distance (300 - 600 mm). Rectification or conversion into 
orthophotography should remain optional. For statistical and census appli- 
cations, a start with photographs at a scale of 1 : 5000 or 1 : 10,000 may be 
appropriate, while updating of the information sets may be done on the 
basis of much smaller scales.