44 BKI.L SYSTEM TRCIINICAL JOi'RXAL 



far from satisfactory. On the whole, the general opinion expressed in 

 an earlier paragraph is borne out, that an enormously greater number 

 of elements is required for a television image for general news or 

 entertainment purposes. This, how'ever, was anticipated, and the 

 real question is whether the results of this experiment indicate that 

 the finer grain image is best attained by resort to multi-channel means. 



This leads to a discussion of what has proved to be a serious practical 

 difficulty with the multi-channel apparatus. This is the problem of 

 keeping the several channels properly related to each other in signal 

 strength. In the e.xperimental apparatus, the direct current com- 

 ponents (introduced at the receiving end) and the alternating current 

 signals, are separately controlled, manually, by potentiometers. 

 These have fine enough steps so that with care, with a non-changing 

 image, a uniform picture may be obtained. If, however, for any 

 reason the signals on one of the channels becomes too strong or too 

 weak, the picture exhibits at once a strongly lined appearance. The 

 eye is quite sensitive to irregularity of this sort, and the transition 

 from a smooth grainless image to one showing a periodicity of 1/3 the 

 number of constituent lines largely offsets the higher resolving power 

 afiforded by the actual number of scanning lines used. A characteristic 

 practical defect of the system as set up is that any marked change in the 

 general character of the signal, such as is produced by a shift from 

 close-up to a wide angle view may throw out the existing signal 

 balance sufficiently to show objectionable grain in the picture. 



DifTerences of this sort in the three signals are of course caused in 

 general by differences in the characteristics of the three circuits. Such 

 differences can arise from overloading of amplifier tubes, whereby one 

 or more may be working on a non-linear portion ; by rectifying action 

 of different amounts in the tubes immediately associated with the neon 

 lamps, or in the neon lamp electrodes themselves. A remedy is the 

 careful design and test of all parts of the system to insure the greatest 

 possible uniformity of performance. When this is carefully done, the 

 behavior of the three signals is reasonably satisfactory. 



Conclusion 



We are, as a consequence of this work, in a position to make a 

 general comparison of the two chief theoretical means for achieving a 

 television image of extreme fineness of grain, which are (1) extension 

 of the frequency band, and (2) the use of several relatively narrow 

 frequency bands. Both, because of the diminished amount of light 

 which finer image structure entails, demand enhanced sensitiveness of 

 the photo-sensitive elements at the sending end, and increased efficiency 



