106 



RADAR SCOPES 



noise levels as reference points. This leads 

 to discrepant statements as to how far below 

 average noise level a signal can be seen. 

 Since visibility measures will have to be on 

 an arbitrary scale for a long time, it is a 

 pressing problem to reach agreement on the 

 reference intensities.* 



Psychophysical Methods 



Memory screens pose a new problem as to 

 the actual procedure of taking a threshold. 

 Their decay characteristic precludes imme- 

 diate re-excitation of a screen spot which is 

 still glowing. It usually is necessary to 

 adopt a sequence of dimmer-to-brighter for 

 the experimental runs on a given day al- 

 though it is possible to deactivate the 

 phosphor by red light between sessions. 

 Traditional methods of psychophysics are 

 applicable with some modifications. 



The Method of Average Error, wherein the 

 subject adjusts a signal until it appears "just 

 visible," works out very well, provided ob- 

 servers are trained enough to be capable of 

 maintaining a constant criterion of judg- 

 ment. The interval of uncertainty, the 

 range over which adjustment is made prior to 

 settlement, is usually no more than a few db, 

 which is sufficiently small to minimize decay 

 and build-up effects. The method has the 

 great advantage that it does not require ex- 



* The decibel is one-tenth of a bel, which in turn 

 is a power ratio of 10 to 1. Although customarily 

 applied to sound intensities the decibel can be 

 applied to any kind of energy and is commonly 

 used in electronics. To make sense, a reference 

 intensity must be defined "absolutely," that is, in 

 some other unit. In sound, an arbitrary ref- 

 erence pressure has been agreed on by scientific 

 workers. No such agreement exists among radar 

 investigators and each laboratory chooses its own. 

 No matter what the reference signal voltage might 

 be, however, use of the decibel unit means a 

 logarithmic scale. Ten db is | log unit in voltage; 

 one db is a voltage ratio of approximately 1.13 to 1 ; 

 six db is approximately 2 to 1. For example, to 

 double a signal voltage, we increase it by six db, 

 which is accomplished by reducing the attenuation 

 of the reference signal by six db. 



tended observations over long periods during 

 which electrode voltages are likely to drift. 



In a modified Method of Limits, the sub- 

 ject simply watches a scope initially devoid 

 of signals while the experimenter gradually 

 introduces a pip. The subject may or may 

 not know the location of the pip before hand, 

 depending on the nature of the experiment. 

 One obtains an appearance threshold. The 

 method is similar to the sighting of a target 

 in operational practice, but apparently was 

 not used in radar research until Garner. A 

 description of its use with multiple targets, 

 staggered in strength and therefore in order 

 of appearance, can be found in Gamer and 

 Hamburger (15). 



A modified version of the Method of Con- 

 stant Stimuli is usable though tedious. 

 However, because of phosphor decay and 

 build-up, it is impossible to employ a truly 

 random order of signal intensities unless the 

 signal is shifted in scope position from trial 

 to trial. This tends to confound the thresh- 

 old and obviate much of the method's tradi- 

 tional precision. Using an unconventional 

 version of the method, the M. I. T. re- 

 cearchers (1) discovered that the operator's 

 guesses of marginal stimuli improved in ac- 

 suracy as stimulus strength increased — a fact 

 known for nearly a century by psychophysi- 

 cists — and that one had to be arbitrary in the 

 selection of a probability criterion. Not be- 

 ing acquainted with psychophysics, they 

 decided to use the 90 percent point on a fitted 

 hnear function instead of the traditional 50 

 percent point on a fitted ogive, a difference 

 which happily is negligible for most engineer- 

 ing purposes. The method may also require 

 a circuit stabihty not easily achievable with 

 electronic equipment. 



Kinds of Factors Influencing 

 Visibility 



Physical Characteristics of the Glass Tube 



The light generated by a spot on a phos- 

 phor screen travels many paths; only about 

 25-35 percent of it is transmitted through 



