An application for generation of spatial sound using the collection of head model

Matej Kladnik (2013) An application for generation of spatial sound using the collection of head model. EngD thesis.

Preview

PDF Download (2184Kb)

Abstract

The purpose of the thesis was to develop an application for the generation of 3D sound. 3D sound is a type of spatial sound that adds spatial impression by generating two, slightly different versions of the sound for both ears. Term Surround sound in contrary means “spatial sound” produced by transmission of sound waves from several spatially placed sources in the room. In the thesis we developed application based on the online public library MIT-KEMAR. The library contains head related impulse responses (HRIRs) measured on artificial head model that are necessary for the generation of 3D spatial sound. The first part of the thesis focuses on the measurement technique and data organization in the library. Use of data from the library is demonstrated in a simple program function, developed in Scilab – environment for numerical computation and signal processing. A function for the graphical demonstration of the impulse responses for left and right ear from the library was written. In addition, responses from different locations of the sound sources were compared. Afterwards, using a new function, 3D sound files were created. The second part of the thesis shows creation of a virtual development system using software tool Oracle VirtualBox. Such system is highly portable and can be used on various operating systems. Main application was developed in Python programming language that is already included in Linux distribution installed on virtual system. Python modules for reading of sound files, matrices and multidimensional arrays operations, graph drawing and other graphical user interface elements were added. Later on, main application was implemented which in addition to generation of elementary 3D sounds also combines them into more complex sound scenes. Using Py2exe module, we produced final version of application that can run without installing Python environment on Windows operating systems.

Actions (login required)