A THEORY OF SCANNING 503 



An Estimate of the Idle Frequency Regions 



As mentioned at the beginning of this paper, the frequency regions 

 between the strong bands appear to be empty when examined with a 

 frequency analyzer of limited level range, or when a narrow band 

 elimination filter is used in connection with visual observations of the 

 reproduced image. These regions are not really completely empty, 

 but do contain weak signal components as shown by the preceding 

 theory, which are not, however, particularly useful inasmuch as, in 

 the final result, they give rise about equally to components simulating 

 the original picture and to masking extraneous components. The 

 regions may, therefore, be considered as idle. 



The factors determining the extent of these idle regions are too com- 

 plicated to permit an exact theoretical evaluation of their width, but 

 an estimate may be attempted from an inspection of Fig. 12 and of the 

 curves given in Appendix II. 



From Fig. 12 and the experience that along the direction of scanning 

 the minimum recognizably transmitted elementary signal length is the 

 length of a rectangular aperture it can be deduced that in the absence 

 of extraneous components the useful band of an aperture extends up to 

 the point where its relative admittance, for a single aperture, is in the 

 neighborhood of 0.65. For two apertures in tandem the corresponding 

 relative admittance is 0.65^ = 0.42. 



Now in Fig. 27 of Appendix II the extraneous components are very 

 small and may be considered negligible. According to the above cri- 

 terion, therefore, the useful frequency band constitutes approximately 

 54 per cent of the total space. The idle frequency regions would, 

 therefore, occupy the remainder, or 46 per cent of the total space. 



Experimental examination of a television signal with a narrow band 

 elimination filter gave the width of the idle regions as 50 to 60 per cent 

 of the total space. This was for a field scanned with a circular aperture 

 giving a one-quarter overlap of scanning strips. The discrepancy for a 

 quantity so vaguely defined is not large but is probably due to incom- 

 plete utilization of even the theoretically active region by the television 

 set because of inherent imperfections in parts of the complete system 

 outside the scanning mechanism proper. 



The width of the individual idle bands is then about half the fre- 

 quency of repetition of scanning lines. For most systems of telephotog- 

 raphy this runs in the order of magnitude of one cycle per second, mak- 

 ing the waste regions very narrow and close together. For systems of 

 television the waste bands come in much more significant "slices," al- 

 though the same fraction of the frequency space is wasted. For ex- 



