Auditory Scene Synthesis for Distributed Audiences in E-Learning Applications

TY - THES

T1 - Auditory Scene Synthesis for Distributed Audiences in E-Learning Applications

AU - Kearney, Gavin

PY - 2006/8/7

Y1 - 2006/8/7

N2 - The enhancement of learning processes through electronic presentation has led to the development of e-Learning strategies which merge traditional classroom instruction with teleconference capabilities, allowing for the creation of a more effective learning environment. Currently, teleconference based e-Learning strategies underestimate the importance of the audio components and do not address the issue of correct localisation of both stationary and mobile sound sources with the video data on a teleconference screen for all audience members. In light of this, an investigation into the application of conventional sound systems as a solution to this localisation issue in classroom based e-Learning is presented. Firstly, sound reinforcement solutions based on stereophonic principles are discussed. These include two and three channel Stereo, Quadraphonics, Ambisonics, and Dolby Surround Sound. These systems are found to be inadequate for the requirements of a distributed audience in e-Learning environments, since they do not provide the necessary auditory cues for correct localisation of the sound source for all audience members.Secondly, existing sound reinforcement solutions which provide the correct wavefronts for source localisation for all audience members are investigated. Two such systems are Delta Stereophony and Wave Front Synthesis, both of which are found to provide sufficient enhancement of the direct sound for correct localisation for distributed audiences. However, this study also reveals that installation of such systems is uneconomical in the context of classroom based e-Learning. Thus, a solution to this localisation problem is presented as a 5 speaker frontal line-array, easily adaptable to existing commercial sound system technology. Its performance as a competent localisation and direct sound enhancement solution is investigated. For this, accurate computer simulation models are constructed, based on existing rooms, and the system performance is compared to reference sources located at the notional source positions. These tests verify that the correct auditory cues for accurate localisation of the source are presented to all listeners. For subjective evaluation, a physical implementation of the design is also presented. Here, the system is found to notably enhance the direct sound whilst providing the correct localisation cues for notional sources to all membersof the distributed audience.

AB - The enhancement of learning processes through electronic presentation has led to the development of e-Learning strategies which merge traditional classroom instruction with teleconference capabilities, allowing for the creation of a more effective learning environment. Currently, teleconference based e-Learning strategies underestimate the importance of the audio components and do not address the issue of correct localisation of both stationary and mobile sound sources with the video data on a teleconference screen for all audience members. In light of this, an investigation into the application of conventional sound systems as a solution to this localisation issue in classroom based e-Learning is presented. Firstly, sound reinforcement solutions based on stereophonic principles are discussed. These include two and three channel Stereo, Quadraphonics, Ambisonics, and Dolby Surround Sound. These systems are found to be inadequate for the requirements of a distributed audience in e-Learning environments, since they do not provide the necessary auditory cues for correct localisation of the sound source for all audience members.Secondly, existing sound reinforcement solutions which provide the correct wavefronts for source localisation for all audience members are investigated. Two such systems are Delta Stereophony and Wave Front Synthesis, both of which are found to provide sufficient enhancement of the direct sound for correct localisation for distributed audiences. However, this study also reveals that installation of such systems is uneconomical in the context of classroom based e-Learning. Thus, a solution to this localisation problem is presented as a 5 speaker frontal line-array, easily adaptable to existing commercial sound system technology. Its performance as a competent localisation and direct sound enhancement solution is investigated. For this, accurate computer simulation models are constructed, based on existing rooms, and the system performance is compared to reference sources located at the notional source positions. These tests verify that the correct auditory cues for accurate localisation of the source are presented to all listeners. For subjective evaluation, a physical implementation of the design is also presented. Here, the system is found to notably enhance the direct sound whilst providing the correct localisation cues for notional sources to all membersof the distributed audience.

KW - AUDITORY SCENE ANALYSIS

KW - E-Learning

KW - LOCALIZATION

M3 - Master's Thesis

PB - Trinity College

CY - Dublin, Ireland

ER -