12 
5.2 
5.4 
  
NOTE: This may be made clearer by an example: a hypothetical f/5 lens 
with 100% transmission would have an axial image illumination relative to 
the source of 1.00. À real f/5 lens with an axial transmission of 86% would 
have an axial image illumination of 0.86 and a corresponding T-stop of 
T : 5.4. If the image illumination at 45° off axis was one-quarter of that at 
the axis, that is 0.215, the T-stop at 45° would be 7 : 10.8. If the average 
illumination over the whole area of the format were found to be 0.4 the 
average T-stop would be T : 7.9. 
  
  
  
DEVIATION OF FILTERS AND CAMERA PORT GLASSES 
The angular deviation which a filter or a camera port glass produces in a 
beam of collimated light incident normally upon it shall be determined for all 
positions on the filter or glass which may be used. Maximum values shall be found 
for the deviation and for the change of deviation occurring with variation of 
position of the incident beam at the filter or glass. 
Since the deviation produced by a filter or a camera port glass is not sig- 
nificantly dependent on the method of measurement it is not necessary to use a 
photographic procedure to simulate practice nor is it necessary to standardize on 
any procedure in detail. Generally the deviation will be measured by placing the 
filter or glass between a collimator or a real distant object and a telescope which 
is fitted with some angle measuring device. For ease in reading the angles, teles- 
cope apertures up to 3 inches may normally be used but for filters or glasses 
having rapid variations in deviation with position on the filter or glass it may 
be necessary to mask the telescope to an aperture equivalent to that of the camera 
involved in order to obtain the true effective value of the maximum variation in 
deviation. 
The deviation produced by the filter or glass should be regarded as a two- 
component vector perpendicular to the collimated beam of light. The variation in 
deviation is the difference of two such vectors and is itself a vector in the same 
plane. Values to be reported are the maximum magnitude of the deviation vector 
and the maximum magnitude of the variation in the deviation vector. Of course, 
to obtain the difference between two deviation vectors, which generally differ from 
one another in both magnitude and direction, it is necessary that a reference 
direction be maintained during the measuring procedure. This is most easily ac- 
complished by having a constant reference direction, that is by not rotating the 
filter or glass about the axis of the collimated beam during measurement. Alter- 
natively rotation of the filter or glass about the axis of the collimated beam may 
be allowed for. 
  
NOTE: Although filters are most generally used at angles of incidence up 
to about 45°, measurement at normal incidence has been specified for sim- 
plieity. The effect of other angles of incidence can be readily calculated and 
should be taken into account when tolerances are being established. For 
example, the maximum deviation produced at 45? incidence is 1.74 times 
that at normal incidence. This ratio is for an index of refraction of 1.5 but 
it does not change seriously for any other likely index. 
  
  
  
Presentation of Data 
The filter's maximum deviation, and its change of deviation shall be stated. 
Orientation of the filter on the camera shall be stated if it is not clearly established 
by marks or indexing devices on the filter and camera. 
——