Masking by combination tones and bands

Masking of a pure tone by a cubic difference tone or band generated by two pure tones or by a pure tone and a narrow band of noise as a function of the level of the masker was measured for three subjects at three ratios of f₂/f₁, 1.125, 1,2, and 1.3. The three maskers used were f₁ alone, f₁ + f₂, and f₁ + NBN (60-Hz band of noise centered at f₂). At ratio 1.125, the amounts of masking produced by f₁ alone and f₁ + NBN were not significantly different for all subjects. For ratios = 1.125 and 1.2, f₁ + f₂ produced the least masking at least in two of the subjects. f₁ + NBN was the best masker at ratios 1.2 and 1.3. These results show that the masking data on cubic difference tone at very small ratios of f₂/f₁ are not reliable due to the masking effect contributed by f₁. For higher ratios, the combination band has to be used instead of combination tone because of the interaction between the signal and the cubic difference tone. Also the divergence of the effectiveness in masking by the three maskers gives an estimation of the levels of primaries at which the cubic difference tone is generated.