Research

3D auditory scenes for Extended Reality (XR) environments are usually recorded using spherical microphone arrays and then binaurally rendered through head-wearable devices. Considerable research is ongoing to integrate spatial audio recording capabilities into these devices. Plausible microphone array designs mountable on head-devices will have arbitrary geometry and unknown scattering effects. In this case, we can not apply conventional eigenbeam processing methods of regular arrays. In this paper, we propose an alternative method using Wearable-Device-Related Transfer Functions (WDRTFs) to extract higher-order ambisonics from the recordings of the head-device. WDRTFs are directional responses incorporating the sound wave transformations caused by the device’s and wearer’s body. The inherent diversity of these functions is exploited to estimate the ambisonics of the recorded soundfield. This paper presents a preliminary evaluation of this method using two irregular microphone array designs worn on a spherical head and the results are promising in comparison to an equivalent SMA. Therefore, we can further adapt this method to achieve perceptually accurate binaural rendering using the microphone recordings of the head-device.