31 
Out of focuo curves - there is an average spread 
of 20% for all spatial frequencies above 3 cycles/mm. 
Measurements from BBC and OTI tend to be higher 
while those from 3DR, RAE, ITEK, and NBS tend to 
be lower than the '‘average”. However, there is 
no clear consistent pattern. 
The respective R and T cases are similar, and 
tend to agree to better than 10# in most cases. 
(b) Off-axis (flat field): 
The variations from lab to lab are much more marked, 
and up to a 30% spread in measurements is noted as 
low as 2 cycles/mm, and at least a 10% spread at the 
higher spatial frequencies, where the MTF itself is 
in the region of 10% response. Certain laboratories 
always read significantly lower than the others. 
Thus, SIRA, ITEK, NBS, and IOR in almost all cases 
give MTF curves much lower than the RAE and PTB labs 
by as much as 30% between 2 and 25 cycles/mm. These 
differences are much more marked in the case of the 
radial lines than for the tangential lines. 
Over a wide range of test conditions the PTB and the 
RAE curves both on and off-axis show good agreement. 
At least on-axis, IOR curves also follow closely 
the PTB curve. 
Wide Band Case 
The MTF, as expected, is significantly lower than for the 
narrow band case. In view of the poor correlation of the narrow 
band MTF curves, even less correlation can be expected in this 
case. The general pattern is again similar. The relative 
behavior of the curves when comparing the different labs for 
both the narrow and wide band spectral conditions is somewhat 
erratic on-axis. The curves tend to follow a consistent pattern 
in the off-axis case with certain laboratory measurements always 
being consistently higher (PTB and IOR) than the rest (though 
less so than for the narrow band case). 
Off-Axis (Best focal plane for off-axis position) 
(See Figure 8) 
The MTF curves in the off-axis case for measurements at 
the best actual focus (focused for maximum response at 40 cycles/ 
mm), are shown. For this case the agreement between the MTF 
curves for the two different spectral conditions for each lab 
oratory (not shown here) is much better than the independent 
measurements for the narrow band (or for the wide band) by each