A tentative method of predicting audibility of modulation noise

In the previous paper, (Neelen, 1970), experiments were discussed on the audibility of amplitude modulation of narrow band signals and wide band signals, using a narrow noise band as modulating signal. For a narrow band carrier like a sinusoid it turned out that the frequency analysing faculty of the ear is important for modulation detection. It was possible to calculate the masking curve of the carrier from the measured modulation thresholds. For a wide band carrier like a periodic pulse the masking curves of the individual harmonics in the frequency spectrum overlap. Amplitude variations of the successive pulses smaller than about 1 dB (corresponding to 5% modulation depth) were inaudible. This paper deals with a tentative method of predicting whether modulation noise with given parameters will be heard in a given signal as a carrier. As a reasonable approximation of modulation noise in practice we shall use low-pass filtered noise with a cut-off frequency f1. As before, a narrow band signal and a wide band one will serve as carrier. In two well-defined situations we will compare the predicted modulation threshold M1' with the measured one.