Journal of the Society of Motion Picture and Television Engineers (1950-1954)

values of a involved, and it is advantageous to study the other two figures first. The trends in Figs. 13 and 14 follow, at least in a rough fashion, the continuation of the experimental and other data of Fig. 7 toward short echo delays, as summarized by the lines (a) and (d). These lines are drawn in the later figures with those designations. The departures of the curves, for positive echoes running generally above, and for negative echoes running generally below the diagonal line, correspond remarkably to the more or less scattered observations reported in the unpublished memorandum of Christopher listed as source (d), Fig. 7. The conclusion drawn from the above is that for the very short echo delays the tolerance can be set effectively upon the amplitude or phase-characteristic excursion, from zero frequency to upper cutoff. The value of K, or measure of this double excursion in amplitude response or radians, is chosen to make the curves approximately tangent to the diagonal line. From Fig. 13 and the trend indicated by the other two figures it would appear that to merge with line (d) should set a value of K at 0.25 to 0.20. The merger occurs at an echo delay near one picture element. Toward somewhat greater echo delays from this point the experimental and other tolerances plotted in Fig. 7 can no longer be specified in terms of a fixed excursion R or $ within the nominal passband. They follow instead the diagonal lines (a) and (d). These are specified in terms of an envelope-delay tolerance, as was developed in connection with that figure. There is some sensible region over which specification in one set of terms or the other leads to about the same result. Thus the exact point of demarcation selected for a setting of tolerances is not critical. The one picture element at which the transition comes corresponds to the inclusion of a half-cycle of the scallop, on the transmission characteristics within the nominal passband (see Figs. 9 and 10). Thus it is on a structure of this order, or coarser, that the tolerance on the phase drift of Fig. IB is to be applied. Correspondingly, it is on a structure of this order, or finer, that the tolerance on the envelope-delay excursions of Fig. 1 C would apply. Influence of Distortion in the Echo Up to this point the detailed discussion has considered only undistorted echoes. By these are meant echoes representing a sharp and clear displaced representation of the original picture. Distorted echoes, in which the displaced image is not sharp and clear, are, however, quite common. An illustration of a simple type of distortion is presented in Fig. 1 5. Here the distortionless echo is shown at the left for comparison. The echo is indicated as coming from an extra transmission path between transmitter and receiver. The extra path takes a longer transmission time than the main path. For a distortionless echo the amplitude response a over the extra path, is flat over the frequency range. The case of the distorted echo is shown at the right. Here the extra path has a response of a type which could be obtained from a small series capacitance. The amplitude response, as indicated at the bottom, is directly proportional to frequency. In a logarithmic frequency plot it would have a slope of 6 db per octave. Such a type of response leads to a differentiation of the echo. This is indicated schematically at the top right of the figure. It is of interest to compare the overall phase and amplitude characteristics of paths having distortionless and differentiated echoes. This is done in Fig. 16. The response over the echo or extra path is repeated at the top of the figure. This response curve forms an envelope for the scallops in both overall amplitude and phase. At the left are repeated the curves which have been shown earlier. At the right the scallops are seen to have Pierre Mertz: Television Transmission Echoes 589