RESEARCH ON VISIBILITY: INTENSITY-MODULATED SCOPES 



121 



treated plexiglass (also untreated vs. treated 

 halves of the same pane). With negligible 

 external illumination, there was no advan- 

 tage of the treated filter. There was a defi- 

 nite though slight advantage of the treated 

 panes when there was a significant amount 

 of room illumination; signals required 0.5 db 

 to 1 db less strength to be seen behind 

 treated than untreated panes. This is easily 

 attributable to reduction in specular reflec- 

 tion. Both types of pane caused some loss 

 in visibihty. For example, in the presence 

 of 20 ft.-c of illumination, the untreated 

 plexiglass was 1.5 db worse than no filter at 

 all but the treated plexiglass was only 1 

 db worse. Visibility tests of coatings on 

 the bulb itself would seem desirable. 



Uniformity of Background 



Excitation of a PPI scope is a type of 

 discrete stimulation: pulses are separated in 

 time. Successive pulsations do not ordi- 

 narily fall on quite the same spot on the 

 phosphor screen because the time sweep 

 (radial or range deflection) keeps "painting" 

 them on at gradually increasing ranges and 

 the rotation of the antenna continually 

 moves the sweep around the circle of the 

 scope. Obviously, if the pulses are few and 

 far between and if the rotation speed is 

 great, the rotating sweep will actually miss 

 many spots on the phosphor. "Silent" gaps 

 of no excitation appear and the observer 

 sees only an intermittent flickering and 

 "spoking" of lights. The intermittence de- 

 creases as the repetition frequency of the 

 pulsing increases and as the rotation becomes 

 slower. To a lesser extent it is also depend- 

 ent on pulse length; and its visual appear- 

 ance, of course, depends on brightness as 

 well as on construction features of the phos- 

 phor screen, etc. The critical construction 

 feature is the "build-up" ratio of the screen 

 which is a function of its decay law; lumi- 

 nous build-up occurs when successive stimu- 

 lations overlap, as they do under the usual 

 firing and deflection rates. With a given 

 screen, and a constant pulse length and bias, 



though, the background will cease flickering 

 and appear uniform above some critical ratio 

 of PRF and rotation rate, jor a particular 

 observer. (Here is an application of flicker 

 data to radar.) A unique variable in the 

 equation is noise. Noise is by definition 

 a random population of pulses which, of 

 course, is no longer entirely random as to 

 duration after passing through frequency 

 filters in the receiver (r-f, i-f and video 

 bandpasses). Because of its random char- 

 acter, noise can be mathematically evaluated 

 only on statistical assumptions. It is here 

 that the elaborate statistical theories be- 

 become important (44, 46, 50, 59). 



From a perceptual standpoint uniformity 

 of background may be a kind of intervening 

 variable, i.e., a requisite condition which is 

 independent of its manner of production. 

 Such an hypothesis needs two kinds of data 

 for confirmation: direct, subjective judg- 

 ments of the "quahty" of uniformity — a 

 spatial analogy of critical fusion frequency; 

 and pip visibility thresholds. Both should 

 be dependent on the electrical parameters 

 in roughly the same way. 



Subjective Judgments 



Attempts in the writer's laboratory to 

 judge the threshold of afterglow uniformity 

 have been rather disappointing because of 

 the diflEiculty of maintaining a constant cri- 

 terion of judgment in the face of so hetero- 

 geneous a stimulus display. Nevertheless, 

 the judgments do correlate roughly with 

 PRF and rotation speed in the expected 

 manner. A PRF of 50 or 100 pulses per 

 second requires a very slow rotation speed 

 (3 rpm) to produce uniformity, whereas a 

 PRF of 300-400 produces the effect with as 

 much as 30 rpm. Judgments of the sweep 

 line are not quite so difficult. Even here, 

 though, frequencies low enough to appear 

 intermittent are disturbing. The bias re- 

 quired to produce the minimum sweep is 

 lower for low PRF's than for high and the 

 PRF for low antenna speeds is smaller than 



