MSBTF hierarchical bandlimitation

The MSBTF hierarchical encoding of surround sound was introduced by Gerzon, developed as a platform for psychoacoustically near-optimum transmission and conversion of surround sound ( Gerzon,1992). It is expected that this type of hierarchical representation of multichannel audio signals will provide a framework for psychoacoustically optimum bandwidth limitation. To verify the assumption, the degradations of audio quality caused by limiting the total bandwidth with different algorithms were studied with formal listening tests. A selected result is shown in the figure 1.

Figure 1: Basic Audio Quality for different processes

The results showed that if a multichannel audio recording is transmitted under highly restricted bandwidth conditions and if some form of bandlimitation is unavoidable, dynamic allocation of bandwidth to the standard audio channels (loudspeaker signals) based on Quality Advisor (the QA) could result in smaller degradation of audio quality compared to the situation when all standard channels are bandlimited equally (the Anchor). However, the best results in terms of audio quality could be obtained when the bandwidth of multichannel audio recordings is limited using the MSBTF-based hierarchical bandlimitation method described in the paper (the MSBTF1, MSBTF2 and MSBTF3).

This work is reported in detail in [P1].