Introduction to Ambisonics and B-Format

As it is my intention to capture and reproduce audio in such a way that the listener is able to perceive sound as if being present at the time of recording, it is necessary to provide a brief introduction to the methods that I shall be employing.

A comprehensive description of the technology associated with this approach has been included in my Dissertation literature review and in the supporting documentation for the Honours Project Module. I will therefore give only a brief outline of how this technology works.

Ambisonics, devised by Gerzon, Fellget, Cooper et al, aims to reproduce spatial audio including height or 'periphony'. Traditional forms of spatial reproduction had no means of capturing periphonic information. Ambisonics however, expanded on the work of Alan Blumlein by adding an omnidirectional microphone and a third figure of eight microphone to Blumlein's pair of figure of eight microphones used for capturing stereo images.

Blumlein's pair



With the addition of an omnidirectional and third figure of eight capsule we now have the ability to capture periphonic information.



The WXYZ relationship is described further by this diagram where W acts as a reference for XY&Z:



As both techniques rely on phase relationships between the relative capsules, both therefore rely on the 'coincidence' of the microphone capsules to each other. If there are any phase discrepancies introduced by the spacing of the microphone capsules then the mathematical assumptions upon which these techniques are based cannot be realised. As it is impossible for any two microphone capsules to occupy the same space then there are bound to be phase inaccuracies, which in turn will skew the resultant stereo image. The manufacturers of the Soundfield microphone have got round this problem by calculating the impact that audio has on a sphere. As the relationship between the surface of a sphere and its centre can be clearly defined the calculations necessary to 'virtually' put the sound sources at the centre of a sphere can be carried out.

The resultant outputs from the four microphone capsules produce what is known as B-format audio. This four channel format relies on various combinations of sum and difference of microphone signal phase to replicate spatial information. These signals can then be directed to any number of loudspeakers, where soundfields or wavefronts, corresponding to those created at the recording stage, are reproduced.

The advantages of Ambisonics and B-format over traditional spatial techniques are; the ability to capture and reproduce periphonic information, not needing to rely on any single playback system and the flexibility of what can be achieved by 4 channels of audio in terms of output configurations.