Externalization of virtual sounds using low computational cost algorithms for hearables
In binaural sound reproduction, it has been shown that externalization improves the listening comfort. Using individualized binaural room impulse responses, it is possible to simulate sound sources in a given room for a listener wearing headphones. However, in some real-time binaural sound applications such as miniaturized hearables, it is not always possible to use such optimal filtering. This potentially results in the perception of virtual sources inside the head rather than externalized. Being able to be aware of surroundings in space, referred as spatial awareness, is another crucial feature in this type of application. This study assessed three sound spatialization algorithms that aim to optimize externalization while preserving spatial awareness. Those algorithms were designed to be implementable on wearable devices, using low computational power and little memory. These algorithms were evaluated in terms of externalization as well as spatial awareness. The results show that a convincing externalization can be achieved with those low computational cost algorithms while preserving spatial awareness.
Publication ➔
In binaural sound reproduction, it has been shown that externalization improves the listening comfort. Using individualized binaural room impulse responses, it is possible to simulate sound sources in a given room for a listener wearing headphones. However, in some real-time binaural sound applications such as miniaturized hearables, it is not always possible to use such optimal filtering. This potentially results in the perception of virtual sources inside the head rather than externalized. Being able to be aware of surroundings in space, referred as spatial awareness, is another crucial feature in this type of application. This study assessed three sound spatialization algorithms that aim to optimize externalization while preserving spatial awareness. Those algorithms were designed to be implementable on wearable devices, using low computational power and little memory. These algorithms were evaluated in terms of externalization as well as spatial awareness. The results show that a convincing externalization can be achieved with those low computational cost algorithms while preserving spatial awareness.
Publication ➔
