2.3. Focus Setting 
Cameras shall normally be supplied for calibration and for use focussed to give best average 
photographic resolving power. A different focussing criterion may be used only if it is clearly 
justified by a user requirement and is stated in the report. Calibration shall be carried out 
in the actual focal plane of the camera, that is, in the position the emulsion surface, if 
it were a true plane, would assume in service. 1f in any type of camera the intended shape of 
the emulsion surface is not a plane, the calibration shall be carried out in the intended 
surface, or by a method which produces equivalent results. 
No comments. 
2.4. Target 
A high contrast target shall be used. 
Discussion 
Two types of targets need to be considered here, the targets for the determination of the lens | 
distortion and the targets for the determination of focus setting and, at the same time, a check on the 
resolution of the lens. A test target is designated as high contrast when the ratio of the 
transmittances or reflectances between the light and dark areas is greater than 100 (a density 
difference greater than 2.0), and medium contrast with a smaller ratio, typically about 10. If the 
target for the determination of the lens distortion is one of medium rather than high contrast it is 
easier to guarantee that a calibration plate can be developed such that all collimator target images 
are located in the straight-line section of the D-LogE curve when all collimator targets are 
illuminated identically. 
The requirement of high contrast was originally intended for the resolving power targets used to 
check the focus and the stability of the resolving power of a lens with time. If the determination of 
lens distortion and the check on image quality are not combined in one operation the definition of the 
target should be based on the requirements for the lens distortion determination. 
Nothing is said about the shape of the target. Collimators are usually equipped with crosses as 
targets, and the photographic images of these crosses are measured with pairs of crosshairs. This is 
contrary to the photogrammetric practice where circular targets are predominantly used when high 
accuracy is to be achieved; and the images of these targets are pointed at with a circular measuring 
mark. It is possible that differences in point locations may result from the two basically different 
targets when the image is produced by a non-symmetrical wave front. Investigations are needed to 
clarify this concern before the target shape is included into the "Recommended Procedures...". 
2.5. Collimators 
Each target shall be placed at the focus of a collimator. A single collimator which is set 
at various angles relative to the cameras may be used, or an array of collimators providing all 
the necessary field angles. Collimator lenses shall be of sufficient quality and focal length 
shall be focused accurately enough that displacements of the camera lens which permit of it | 
being entirely filled do not introduce significant errors in measurement. bu 
NOTE: Spherical aberration of the collimator objectives or small focussing errors could 
introduce significant inaccuracies. Such effects should be checked carefully. a Hh 
2.5.2. The collimator apertures and the positions of the collimators relative to the camera shall be 
such that the entrance pupil of the camera lens is filled with light from every part of the 
illuminated collimator field. 
2.5.3. Luminance shall be uniform over the collimator field and of the spectral quality specified in 
section 1.4. 
No comments. 
2.6. Number of Field Angles 
Targets shall be photographed at the principal point of autocollimation and along at 
least the two diagonals of the image format. They shall be photographed at a sufficient 
number of off-axis angles (at least 6 on a semi-diagonal) in each half field to permit 
accurate plotting of the distortion to the limit of the field. 
Discussion 
The number and the location of the collimators may influence the result of a calibration and 
therefore contribute to differences between calibrations of the same lens at different 
institutions. The introduction of standard lens distortion reference curves for all lens types and 
the use of these curves by all institutionswould reduce the task of lens distortion determination 
to the determination of the deviations from these curves. Hence, the number and locations of the 
collimators would no longer influence the result to the same extent; also, a calibration would be 
92