A Loudspeaker-Based Projection Technique for Spatial Music Applications Using Virtual Microphone Control

This article describes a new sound-projection system for multichannel loudspeaker setups that has been developed by the authors. The system, called Virtual Microphone Control (ViMiC), is based on the simulation of microphone techniques and acoustic enclosures. In auditory virtual environments (AVEs), it is often required to position an anechoic point source in three-dimensional space. When sources in such applications are to be displayed using multichannel loudspeaker reproduction systems, the processing is typically based upon simple amplitudepanning laws. With an adequate loudspeaker setup, this approach allows relatively accurate positioning of spatial images in the horizontal plane, but it lacks the flexibility many composers of computer music would like to have. This article describes an alternative approach based on an array of virtual microphones. In the newly designed environment, the microphones, with adjustable directivity patterns and axis orientations, can be spatially placed as desired. Each virtual microphone signal is then fed to a separate (real) loudspeaker for sound projection. The system architecture was designed for a maximum flexibility in the creation of spatial imagery.

Despite its flexibility, the system is intuitive to use because it is based on the geometrical and physical principles of microphone techniques. It is also consistent with the expectations of audio engineers to create sound imagery similar to that associated with standard sound-recording practice, but it goes beyond the original concept by allowing strategic violations of physically possible parameters; namely, new supernatural microphone directivity patterns can be implemented into the ViMiC system.

This article begins with a review of various microphone techniques on which the ViMiC system relies and alternative sound-projection techniques. Next, the fundamental physical concepts on which the ViMiC system is based are described. In the following section, software implementation of the system is outlined with a focus on strategies to keep processor load and system latency low. The article concludes with a description of several projects that involved the ViMiC system.

Historical Background

In the 20th century, electroacoustic and electromechanical devices were invented to spatialize sounds dynamically, before computers became advanced enough to fulfill this task with real-time audio-processing algorithms. First, the introduction of sonic spatialization techniques based on microphone arrays should be mentioned. In 1931, Alan Blumlein filed a patent on a newly developed two-channel recording scheme based on two bidirectional microphones with coincident diaphragms angled at 90°. (Blumlein 1931). The family of stereo recording techniques with coincident microphone diaphragms were later referred to as XY techniques. In all XY techniques, the directionally dependent microphone sensitivities are used to encode the azimuth angle of the recorded sound source as level differences between both channels. Mr. Blumlein’s invention, which was made practical through the introduction of the ribbon microphone by Siemens (Weiss 1993), was the birth of stereo recording.

Almost at the same time, Steinberg and Snow (1934) introduced another microphone-based recording technique that was used in 1933 for a tele-presentation with Leopold Stokowski and the Philadelphia Orchestra. The basic idea was to capture a wavefront with several microphones, as shown in Figure 1. Their setup consisted of only two or three transducers, owing to technical limitations, but this number was shown in a psychoacoustic experiment to be sufficient for spatial coding purposes. In their recording scheme, the directions of the sound sources are coded almost solely through inter-channel time differences and to a lesser extent through inter-channel level differences. This type of method, which is based on spaced microphone placements with two receivers, is now well known as the AB technique (or spaced technique) in the USA and parts of Europe.

Figure 1.

Steinberg and Snow’s (1934) recording setup.

It took many years until a full “curtain of microphones” could be processed simultaneously to address a large loudspeaker array using the concept of wave field synthesis (WFS) (Berkhout 1988; Berkhout, Vogel, and de Vries 1993), which is based on Huygens’s Principle and the Kirchhoff– Helmholtz Integral. The present WFS technology shares two important features with the original work of...