Performance of time-position discriminator of composite signal

Optimal time-position discrimination involving multiplication of the signal+noise mixture by the time derivative of its video copies with maximum slope in the range of small deviations, is compared with time-position discrimination involving multiplication of the signal+noise mixture by the difference of its video copies shifted in time symmetrically forward and backward from the instant of readout. In terms of the output signal-to-noise ratio, the advantage multiplying by the derivative of video copies at time Tau = 0 is equal to the advantage of multiplying by the difference of video copies at times Tau = plus or minus Tau sub kappa. As the time shift decreases toward zero, both methods of discrimination become equally efficient but the magnitude of the transmission coefficient decreases sharply. This is demonstrated on a signal with a uniform finite spectrum.