Interfering Effect of Tones and Noise Upon Speech Reception

Abstract

The interfering effect upon speech of tones and noise was measured by determining the percentage of vowel, consonant and word articulations correctly heard by an observer who listened in the presence of a tone or noise to meaningless monosyllables spoken at a distance of 2 meters. The interfering tones varied in pitch by octaves from $C_2(128\mathrm{}\mathrm{d}.\mathrm{v}.\mathrm{})$ to $C_7(4096\mathrm{}\mathrm{d}.\mathrm{v}.\mathrm{})$ and in loudness from slightly above minimal audibility to near the painful limit. The interfering effect increases with increasing loudness of tone or noise at an accelerated rate; for tones with a loudness less than the loudness of the speech, the effect is almost independent of the pitch, but as the interfering tone becomes louder, the tones of lower pitch produce a relatively greater and greater interference. These results are consistent with the auditory masking data of Fletcher. Wegel and Lane. A noise produces twice the interfering effect of a tone of the same loudness. These results indicate that for good hearing in an auditorium the speech energy should be from 1000 to 10,000 times the energy of any interfering noises, and that sound-absorbent materials for the reduction of noises in auditoriums should absorb equally tones of all pitches below $C_7$. Some experiments with rather deaf persons will be reported elsewhere.