Phase Information in Spatial Audio Techniques

Some spatial audio techniques use the intensity and the phase when decoding the signals, whereas some other techniques only use the intensity information and disregard the phase. From the following methods, which ones rely on the phase information? Transaural reproduction, VBAP, WFS, and Ambisonics.

Analysis of Specific Methods

(a) Transaural reproduction: Uses phase information. It relies on the precise cancellation of signals between the left and right loudspeakers, so that it depends on constructive and destructive interference and therefore uses phase information.

(b) VBAP (Vector Based Amplitude Panning): Almost does not use any phase information. With VBAP, the final level is adjusted only taking into account the level of each speaker, assuming an incoherent sum of the signals. It does not rely on interference phenomena.

(c) WFS (Wave Field Synthesis): Uses phase information, which is vital for the reconstruction of the wavefronts by using the principle of interference.

(d) Ambisonics: Depending on the decoding principle, it uses phase information or not. The basic decoding relies on the coherent addition of signals between the different loudspeakers, but the max-rE and in-phase depend on the incoherent summation and therefore do not take phase into account (except that in-phase requires no speakers in anti-phase).

Wave Field Synthesis Challenges and Solutions

Truncation and Diffraction Effects

a) The wave field turns around beyond the array of loudspeakers and reflects back inside the array. This is a phenomenon called truncation, which causes diffraction and is produced because we have a limited amount of loudspeakers.

Possible solutions:

• Tapering window: Similar to an analysis window used before applying FFT in speech signal processing.
• Wave-Equation modification: Add a new term to the Wave-Equation to compensate for the effect (modification of amplitude on the loudspeaker on the edge).

Loudspeaker Density and Wavefront Reconstruction

b) Figure a) (Top) only uses 3 loudspeakers, which is not sufficient to create a proper wavefront. It can be observed that 3 independent wavefronts, generated by each individual loudspeaker, are created, but they do not add to each other.

As the number of loudspeakers is increased, the wavefront can be better created. The more loudspeakers there are, the less distance there is between them, and the more frequencies can be synthesized because of the Nyquist theorem.