An octogenarian male with gently sloping audiograms with average hearing threshold levels (0.5, 1.0, 2.0 kHz) of 56 dB was tested extensively with a two-channel master hearing aid. Each channel includes a headworn microphone and a button receiver in a fitted earmold. The electronics consist of laboratory-quality amplifiers and 13-octave-band equalizers. Speech-reception tests were conducted in an acoustically absorbent room with a closed set of 100 monosyllables presented live by a female talker in quiet and noise. In a single mini-experiment, three or four frequency responses were compared in two to four sessions of 1 hour with a counterbalanced and randomized design that allowed frequent rest periods. When the field-to-eardrum transfer function is matched in the hearing-aid over a bandwidth of 6.5 kHz, reception of monosyllables is superior to competing curves by 6%–15%. Competing curves have included: those of a good quality commercial hearing aid; high-frequency pre-emphasis of +8 dB/octave defined at the eardrum ; curves with less than +18 dB relative gain in the 2–4 kHz region; curves with bandwidth less than 6.5 kHz; and so on. Finally, it is suggested that the role of frequency distortion in the design of hearing aids may be understood by relating aided and unaided audibility curves for sound fields to measures of speech reception.
Read full abstract