Final Implementation

Our spatial audio system used the HRTF data generated by the MIT Media Lab to achieve spatial directionality. To allieviate the "in the head" effect of headphones, we implemented a flexible reverb algorithm to simulate room effects and create a more natural sound.

mit_3D.m convolves HRTF with input sound
KEMAR data Complete set of HRTF data from MIT
HRTF mfiles Mfiles to be used with HRTF data from MIT
earlyEchos.m calculates locations of 16 virtual sources for the given room size, listener location and sound source location
echoFilter.m calculates delays and gains for all virtual sources computed in earlyEchoes.m
roomDemo.m demo of original speech signal and its response in 4 room types
roomDemo2.m demo of reverb algorithm where user can specifiy room size, listener location, and distance to sound source
roomResponse.m generates reverb algorithm by calling earlyEchoes and echoFilter
xyz_dist.m utility to compute the absolute distance between two objects in 3-space
makeSameLength.m Helper function used to make two sounds the same length
make_window.m a trapezoidal window generating function used to window hrtf outputs
processSound.m does directionality (HRTF) processing, then reverb processing on a sound

Back | Index | Next