Interpolation of binaural impulse responses for virtual auditory displays

Summary

The aim of this project is to minimise the number of recordings that need to be made by investigating methods for the interpolation of a limited number of recorded binaural impulse responses.

Supervisor(s)

Associate Professor Craig Jin

Research Location

Electrical and Information Engineering

Program Type

PHD

Synopsis

When we listen to sounds, we can often quite accurately localise the source of the sound in the space around us. However, when we listen to sounds over headphones, the sounds appear often to come from positions inside the head. The difference between the two situations is that during normal listening, our head and outer ears filter the incoming sound in a directionally dependent manner and this is absent during headphone presentation. We can record the directional transfer function of the ears (the binaural impulse responses) for a person by inserting miniature microphones in the ear canal and playing sounds from various positions in space around us. A sound can then be filtered with these impulse responses to create a virtual auditory display where, even though the sound is played over headphones, it appears to come from a specific position is space around us. At distances larger than 1m, the binaural impulse response is only a function of direction and not distance. But at distances less than 1m the impulse responses change with both distance and direction. To fully cover the 3D space around a listener many thousands or recordings would have to be made. Furthermore, different rooms will add different reverberation to the sound and in particular the early echoes will appear from different positions in space as the position of the source or the listener changes. This would require many additional recordings for varying source and listener positions. The aim of this project is to minimise the number of recordings that need to be made by investigating methods for the interpolation of a limited number of recorded binaural impulse responses. This involves the development of a technique to model the changes in the impulse response with direction and changes with distance for distances smaller than 1m. A second technique will be developed to model the changes in echoes in a reverberant room as the position of the listener or the source changes. The results of these techniques are verified with psychophysical listening tests.

Keywords

Audio Engineering, Signal Processing, Virtual Reality, Virtual Auditory Displays

Opportunity ID

The opportunity ID for this research opportunity is: 270

Other opportunities with Associate Professor Craig Jin

• Pattern analysis techniques for sound synthesis
• Sound field recording and recreation
• Beamforming with acoustic vector sensors - Multiple acoustic source localisation using acoustic vector sensor arrays
• Speech separation and localisation using particle filtering
• Mapping 2D Images to 3D Shape
• New technique for studying human brain activity
• Next Generation Audio Coding
• Spherical multi-modal scene analysis
• Statistical models of ear shape and ear acoustics
• Binaural signal processing algorithms for hearing aids
• Electrical Impedance Tomography for stroke, biophysical monitoring and medical device design
• Impedance tomography for cardiac imaging: high speed tomography
• Medical diagnostics for neonates in the developing world
• FPGA-based Low Latency Trading
• Floating Point FPGA Architectures
• Placement-aware Hardware Description Languages
• Scalable vision machines
• Modelling Parkinson's disease using control models
• Novel Electrodes for rapid electrophysiological recording
• FPGA-based low latency machine learning