387 
?heric pressure is 
'ease, so that the 
sre is no attempt 
ry largely by the 
>al distance and 
re, although the 
ion. 
ng through the 
nomenon, dealt 
ngus-Leppan cl9J 
erature gradient 
imputing it from 
correction to be 
He has deduced 
of both effects 
r curvature and 
5 small, perhaps 
grapher, so that 
e accuracy. Of 
hy. 
airstream close 
window. As air- 
ases, turbulence 
ver the window, 
appear first as a 
¡ting account of 
'rom supersonic 
uture, but if the 
ey, then we will 
from refraction 
' reduce misting 
> “shimmering” 
oscope and can 
For this reason 
uring the actual 
:e they are part 
d as the camera 
ss and even if 
a occur. More- 
aterferometry is 
o interpret and 
deviations on a 
era and is quite 
/er the windows 
ble temperature 
gradients can occur across its thickness. The air outside the aircraft will be at - 5°C. 
at 10,000 feet or -25°C. at 20,000 feet, while the camera itself needs to be kept 
at + 10°C. This differential has the effect of bending the window into a quasi- 
spherical shape and this introduces radial displacements to the image. The order of 
magnitude of these effects can be readily calculated by simple theory. W. P. Smith 121 
states that experience with and without a camera window shows that better answers 
from aerial triangulation are obtained without a window. He gives no quantitative 
information and there is no evidence whether this is due to window defects or tem 
perature gradients. It is very possible that he has experienced refraction errors due 
to hot air passing in the narrow space between the camera window and the camera 
filter. 
15. Attitude 
The camera is usually vertical in the aircraft. It is not always so during 
calibration, and examination of the literature certainly suggests that the camera is 
more often held horizontal. Gravity is then acting in a different direction on the 
various optical components and this may affect the performance. For example, 
lenses may contain components which are “knife-edge” thin at their edge. To 
prevent breaking under strain, these must not be held too tightly by their metal 
mounts, and there may be some slight movement of these when the camera is turned 
on its side or they may sag under their own weight. The register glass may take up a 
slightly different position and may deflect under its own weight or even more so under 
the pressure from the backing plate when the magazine is used in the air. For 
cameras with a suction backing plate, either the backing plate or the film may take 
up a slightly different position under their own weight. There may also be reduced 
suction, due to lower ambient pressure, holding the film against the plate. 
16. Vibration and Image Movement 
The camera in an aircraft will be constantly vibrated, even if it is given an anti 
vibration mounting, and the whole aircraft will also be moving over the ground. 
The image is therefore moving over the focal plane during the exposure period. If 
this movement is sufficient to cause loss of image quality, there may also be geometric 
errors. First, loss of image quality leads to a lower precision of setting in the 
plotting equipment. Secondly, differences arise if settings are not made on similar 
objects. The judgement of the centre of a thin line image may not be affected by 
vibration or movement, since both the image and the movement are symmetrical 
about the centre. The apparent position of an edge may well be altered since this is 
not symmetrical. As vibration gives a movement at the edge of the format of approxi 
mately twice that at the centre, the errors introduced are likely to increase with the 
radius. There appears to be no quantitative information on this effect. In practice, 
the use of the fastest possible shutter speed, 1 /500th or 1/1000th second, not only 
gives a marked improvement in image quality, but also improves residual errors 
associated with these image movements. 
17. Field Calibration 
The obvious method of making a correction for all the possible errors associated 
with the installation of the camera in an aircraft would be to calibrate the whole 
system from negatives taken in the air. In the simplest form this would imply flying 
over a known target area, containing accurately measured ground markers. The 
negative would then be measured and a set of corrections deduced to be used in later