Marton Marschall

Sound-field reconstruction performance of a mixed-order Ambisonics microphone array

Recently, there has been increasing interest in using spherical microphone arrays for spatial audio recordings. Accurate recordings are important for a range of applications, from virtual sound environments for hearing research through to the evaluation of communication devices, such as hearing instruments and mobile phones. Previously, a mixed-order Ambisonics (MOA) approach was proposed to improve the horizontal spatial resolution of spherical arrays. This was achieved by increasing the number of microphones near the horizontal plane while keeping the total number of transducers fixed. The approach is motivated by the fact that human spatial hearing is most acute in the horizontal plane. This study presents simulations of the performance of an MOA rigid-sphere microphone array, and its robustness to variations in microphone characteristics. Specifications of a commercially available microphone were used to simulate self-noise, sensitivity, and phase response variations between the microphones. To quanti...

Measuring speech intelligibility with speech and noise interferers in a loudspeaker-based virtual sound environment

Loudspeaker-based virtual sound environments (VSEs) are emerging as a versatile tool for studying human auditory perception. In order to investigate the reproducibility of simple sound scenes, speech reception thresholds (SRTs) were measured with two interferers and in two spatial conditions (co-located and ±30° separated) using the Danish matrix sentence test Dantale II (Wagener et al., 2003). SRTs were measured in a typical listening room and in a VSE consisting of a spherical 64-channel loudspeaker array using simulated room acoustics with mixed-order-Ambisonics (MOA) playback. The speech maskers were taken from the same material as the target (different talker, same sex). The noise maskers had the same long-term spectrum and broadband envelope as the speech interferer but had random phase (Best et al., 2013). The co-located conditions were reproduced comparably in the real room and in the VSE, with both speech and noise interferers. However, spatial separation led to a 3 dB higher benefit in the VSE t...

Evaluating the auralization of a small room in a virtual sound environment using objective room acoustic measures

To study human auditory perception in realistic environments, loudspeaker-based reproduction techniques have recently become state-of-the-art. To evaluate the accuracy of a simulation-based room auralization of a small room, objective measures, such as early-decay-time (EDT), reverberation time, clarity, interaural cross-correlation (IACC), and the speech transmission index were measured in an IEC listening room for 28 source-receiver combinations. The room was then modeled in the room acoustics software ODEON, and the same objective measures were also evaluated for the auralized version of the room. The auralizations were generated using the loudspeaker-based room auralization toolbox (LoRA; Favrot and Buchholz, 2010) and reproduced in a 64-channel loudspeaker array, set up in an anechoic chamber. Differences between the objective measures evaluated in the real and the virtual room were within about twice the just-noticeable differences for most measures, and were comparable to the median results of the ...

The effect of compression on tuning estimates in a simple nonlinear auditory filter model

Behavioral experiments using auditory masking have been used to characterize frequency selectivity, one of the basic properties of the auditory system. However, due to the nonlinear response of the basilar membrane, the interpretation of these experiments may not be straightforward. Specifically, there is evidence that human frequency-selectivity estimates depend on whether an iso-input or an iso-response measurement paradigm is used (Eustaquio-Martin et al., 2011). This study presents simulated tuning estimates using a simple compressive auditory filter model, the bandpass nonlinearity (BPNL), which consists of a compressor between two bandpass filters. The BPNL forms the basis of the dual-resonance nonlinear (DRNL) filter that has been used in a number of modeling studies. The location of the nonlinear element and its effect on estimated tuning in the two measurement paradigms was investigated. The results show that compression leads to (i) a narrower tuning estimate in the iso-response paradigm when a ...