Photogrammetria, XIX, No. 6 
l 
perception (fig. 7). It is provided with indications for the amplitudes or frequencies (or 
detail size). 
When using the sine*-curve instead of the succession of circles we no longer have all 
scanning directions evenly represented. 
This is to overcome by two sine*-curves, perpendicular to each other. Consequently, 
there is no essential difference for the purpose of analysis, between the approach based on 
the Fourier transformation and the sine* curvatures. 
Fig. 7. Pseudo-sine edge representing the input Fig. 8. The output pattern 
III. 4. Execution of the test. 
The mechanical components of the plotting system can be analysed more reliably 
if the non-dynamic disturbances in the optical components and human operator are artifi 
cially suppressed, and vice versa. 
The dynamic system should be investigated in the representative and critical positions 
of the scanning and plotting mechanisms and along the representative and critical direc 
tions. The arrangement of successive individual tests should enable the separation of the 
principal groups of disturbances (6, d, D). 
In the assumed case the groups <5 and d must be investigated separately for each of 
the two projectors, while the group D only once. A more detailed arrangement of the ex 
perimental work is described in [6]. 
Apart from the plotting, the scanning speed per sine* wave is recorded. Within the 
range of higher frequencies it is about constant. Higher frequencies require more care 
than the lower ones, where the path to be scanned is longer. The graduation for the 
amplitudes (or the frequency) in the input and the estimated limit for the visual perception 
(v.p.) should also be indicated on the plot. An example of the recorded output is shown in 
fig. 8. The sine* pattern under 45° is an useful addition as a check and for a more accurate 
definition.