Time-Shift Modeling-Based Hear-Through System for In-Ear Headphones

Abstract

The hear-through (HT) technique was developed actively to compensate for passive isolation by enhancing the perception of ambient sound by wearers of headphones. The passive material in headphones reduces the high-frequency components of sound above 500 Hz. The HT algorithm compensates for the loss of ambient sound by generating an artificial sound, derived using a relative transfer function (RTF) between a microphone and the user’s ears. Conventionally, HT performance depends on the direction of arrival (DOA) of the ambient sound. This work develops a time-shift scheme for identifying the RTF and an active equalization algorithm to calculate accurately the compensating sound. A shaping filter is also developed to prevent acoustical interference by the artificial sound at low frequency. Finally, the proposed approach is integrated into in-ear headphones and its HT performance is compared with that of a commercial product to verify the effectiveness of the proposed algorithm.