Imperial College London > Talks@ee.imperial > COMMSP Seminar > Perception-Based Methods for Spatial Audio

Perception-Based Methods for Spatial Audio

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Alastair Moore.

Imagine being at home while following a broadcast from the Royal Opera House, and having the impression of listening to it from the best seats of the opera house. Alternatively, imagine watching the 100m sprint race with your 3D glasses, and also hearing and feeling the trembling arena around you as the referee shoots the starter pistol. Achieving this when one has access to a very large number of accurately positioned loudspeakers is relatively easy. In most home surround systems, however, only 5-10 loudspeakers are available. In this case, one has to rely to some extent on perceptual and psychoacoustic phenomena. The first part of this talk will be focused on a framework for the analysis and design of multichannel audio systems based on concise modelling of underlying psychoacoustic phenomena. Formal listening experiments show that this framework improves over the state-of-the-art in terms of localisation accuracy and stability of the auditory perspective. While listening experiments provide the ground truth in this context, they are expensive and time-consuming to carry out. If one had access to reliable models of the auditory system, on the other hand, the analysis and design of multichannel audio systems could be carried out in a much larger parameter space. The second part of the talk will be concerned with the modelling of the binaural system. It is known that the most important perceptual cues used to localise sound sources are the interaural time and level differences. How these cues are combined together and across frequencies, however, is still unclear. During this part of the talk, a novel way of combining them will be presented, which will be shown to provide a remarkable agreement with experimental data. The model will then be used to provide new insights into the design of multichannel audio systems. Finally, if time allows, new perception-based room acoustic methods and microphone beamforming methods will be briefly presented as enabling technologies of the developed multichannel audio system.

This talk is part of the COMMSP Seminar series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

Changes to Talks@imperial | Privacy and Publicity