Masked threshold changes associated with angular separation of noise and signal sources.

Threshold changes associated with separating a signal source and a masking white noise source from 0 degree to 90 degrees were determined for 0.5, 1 and 8 kHz pure tones and click trains. No changes occurred for the 0.5 and 1 kHz pure tones. Masked thresholds of 8 kHz pure tones and click trains decreased linearly by 9 and 13 dB respectively as angular separation was moved from 0 degree to 90 degrees. Changes in click train stimuli masked thresholds did not change significantly when the ear directed toward the masking source was occluded (11 dB drop at 90 degrees). The absence of changes at low frequencies and the similarity in magnitude of the changes in signals containing high frequency components with the responses to the monaural click trains, suggests that the threshold changes can be attributed to a head shadow effect. The casting of a sound shadow effectively lowers the noise level on the shielded side. These findings question the importance of cross-correlation techniques when detecting signals in noise.