30 
spectral band and (after refocusing for the best focus again) 
the wide spectral band curves, careful reduction of the data 
supplied was necessary where some other focal positions had 
been selected as the nominal "zero*’ focal position* 
The designation tangential and radial here means those 
orientations of the scanning slit (or spatial frequency patt 
erns) in the image plane of the lens under test* In the part 
icular scanning techniques, the scanning direction is, of 
course, perpendicular to the line orientation. Clearly, on a 
conventional optical bench with nodal slide, a vertical slit 
would be used for obtaining the MTF curves for the T (tangential) 
line orientation* Focus convention was such that a positive 
defocus inferred a focal shift along the optical axis away from 
the lens* 
The field angle convention adopted was such that a pos 
itive field angle corresponded to an anti-clockwise rotation 
from the lens principal axis to the principal ray as observed 
from above* 
The lens had very low response above 100 cycles/mm and 
probably for this reason very few laboratories submitted results 
above this spatial frequency, although up to 300 cycles/mm had 
been called for in the specifications. Although not specifically 
requested at this stage, phase transfer function (PTF) data was 
submitted by both SIRA and by RAE together with limited data 
from Itek. There was only relatively poor correlation here. 
The MTF data is shown in summarized graphical form in 
Figures 3 and 8* Examination of these indicates that while 
the same general form of the curves is obtained in all cases, 
there are relatively large differences in the actual numerical 
values, particularly at the lower spatial frequency points where 
one might have expected better consistency in the measurements. 
As the full graphical results are very voluminous, they 
are not presented here, but a brief summary of the main points 
is attempted* 
The examination of all the curves demonstrates that both 
systematic and non-systematic variations occur between the meas 
urements from the different labs* The magnitudes of the differences 
are relatively large in many instances, and this suggests a serious 
standardization problem* 
Monochromatic Case 
(a) On-axis: 
In focus curves - there is an average spread of 10% 
for both the R and T orientations, for spatial fre 
quencies ranging from 3 to 100 cycles/mm.