Spatial Encoding of Finite Difference Time Domain Acoustic Models for Auralization

Alexander Paul Southern, Damian Thomas Murphy, Lauri Savioja

Research output: Contribution to journalArticlepeer-review

Abstract

A single room impulse response can reveal infor- mation about the acoustics of a given space in both objective, and, when used for auralization, subjective terms. However, for additional spatial information, or more accurate and perceptually convincing auralization, multiple impulse responses are needed. Higher order Ambisonics is a robust means of capturing the spatial qualities of an acoustic space over multiple channels for decoding and rendering over many possible speaker layouts. A method for obtaining Nth-order Ambisonic impulse responses from a room acoustic model, based on lower orders using differ- ential microphone techniques is presented. This is tested using a third-order encoding of a 2-D finite difference time domain room acoustic simulation based on multiple circular arrays of receivers. Accurate channel directional profiles are obtained and results are verified in a series of listening tests comparing the localization of a sound source placed within the given simulation to the same source encoded directly. This generic encoding scheme can be applied to any room acoustic simulation technique where it is possible to obtain impulse responses across multiple receiver positions. Although the proposed method encompasses horizontal encoding only, it can also be applied directly in 3-D simulations where height information is not required in the final auralization.
Original languageEnglish
Article number6214587
Pages (from-to)2420-2432
Number of pages12
JournalIEEE Transactions On Audio Speech And Language Processing
Volume20
Issue number9
DOIs
Publication statusPublished - 2012

Keywords

  • FDTD
  • spatial encoding
  • Auralization
  • Acoustic

Cite this