Presentation Title

Acoustic Impulse Response Capture and Reverb Convolution Modelling

Format of Presentation

Poster to be presented the Friday of the conference

Abstract

For our study, the goal is to simulate, capture and analyze the acoustic properties (in particular reverb) of real space locales to the best of our abilities with the equipment we currently possess. An impulse response is a sort of unique audio signature which describes how a given impulse of sound will reverberate and echo at a particular location. If one can capture that response, it’s possible to create a model which can transform input sounds to the soundprint of the measured location, thus giving the illusion that those sounds were played at the original location. We will construct said model, and present it at the conference along with a poster describing the research we did to come to it. Interested viewers will be able to listen to the different reverbs through a pair of headphones that will ”transport” them to different locations around the school. This technology has a lot of applications in sound design for film, video games, virtual reality, music production, audio forensics, among other related areas. The process of capturing the impulse response is essentially to measure the input and output signals generated by a location and deconvolve them. This will be done using a set of microphones and speakers, with the deconvolution done digitally in the Csound audio programming language. Our reverb model will convolute the impulse response with some

input signal (a users voice for example) and play back a transformed version of that signal to the user, repeating the process in reverse. Our model will be contained within a VST file suitable to be loaded into a multitude of Digital Audio Workstations (DAWs). By examining several acoustically interesting areas around TRU, we hope to successfully capture there impulse responses (soundprints) with minimal pre-ringing, abrupt pulsive noises, skewing, cancellation, and timesmearing. To accomplish this, we have conducted research on related protocols and propose to implement the preliminary tests, equipment alignments, signal capturing, and analysis methods presented by Angelo Farina and suggested by our co-supervisor Richard Taylor. These methods include Equipment Equalization, Exponential Sine Sweep (ESS) and Artifact Reduction. Analysis of the results will be conducted through the use of various software applications and hardware devices. To simulate the captured models we are developing a software application called a VSTi (Virtual Studio Technology Instrument) that will contain the captured models to be used for presentation. The VSTi will also contribute to the audio industry by being made available to the public for free.

Department

Physics

Faculty Advisor

Mark Paetkau & Richard Taylor

This document is currently not available here.

Share

COinS
 

Acoustic Impulse Response Capture and Reverb Convolution Modelling

For our study, the goal is to simulate, capture and analyze the acoustic properties (in particular reverb) of real space locales to the best of our abilities with the equipment we currently possess. An impulse response is a sort of unique audio signature which describes how a given impulse of sound will reverberate and echo at a particular location. If one can capture that response, it’s possible to create a model which can transform input sounds to the soundprint of the measured location, thus giving the illusion that those sounds were played at the original location. We will construct said model, and present it at the conference along with a poster describing the research we did to come to it. Interested viewers will be able to listen to the different reverbs through a pair of headphones that will ”transport” them to different locations around the school. This technology has a lot of applications in sound design for film, video games, virtual reality, music production, audio forensics, among other related areas. The process of capturing the impulse response is essentially to measure the input and output signals generated by a location and deconvolve them. This will be done using a set of microphones and speakers, with the deconvolution done digitally in the Csound audio programming language. Our reverb model will convolute the impulse response with some

input signal (a users voice for example) and play back a transformed version of that signal to the user, repeating the process in reverse. Our model will be contained within a VST file suitable to be loaded into a multitude of Digital Audio Workstations (DAWs). By examining several acoustically interesting areas around TRU, we hope to successfully capture there impulse responses (soundprints) with minimal pre-ringing, abrupt pulsive noises, skewing, cancellation, and timesmearing. To accomplish this, we have conducted research on related protocols and propose to implement the preliminary tests, equipment alignments, signal capturing, and analysis methods presented by Angelo Farina and suggested by our co-supervisor Richard Taylor. These methods include Equipment Equalization, Exponential Sine Sweep (ESS) and Artifact Reduction. Analysis of the results will be conducted through the use of various software applications and hardware devices. To simulate the captured models we are developing a software application called a VSTi (Virtual Studio Technology Instrument) that will contain the captured models to be used for presentation. The VSTi will also contribute to the audio industry by being made available to the public for free.