B-Format Can Provide Acoustical Predictions

By Pat Brown

In this article, Pat Brown will show how a B-Format can provide acoustical predictions.

In addition to recording and room measurements, B-Format is also applicable to acoustical predictions. The holy grail of sound design has always been to calculate the system response during the design phase of an auditorium. The WXYZ approach is well-suited for this, as it allows the 3-dimensional characteristics of the sound field to be considered. Figure 1 shows a settings dialog for the auralization module of CATT-Acoustic™ www. catt.se that allows the selection of a receiver model. Note that B-Format is one of the choices. The use of B-Format allows a number of microphone and listener configurations to be synthesized as a post-process.

B Format allows 3 dimensional auralization

Predicting the RIR

Figure 2 shows a predicted room impulse response (RIR) for a listener position in a virtual auditorium. It was generated using a Reflection Cone Tracing algorithm that combines image source and ray tracing methods. The ISO3382 acoustic metrics can be derived from this RIR (also shown in the figure).

(RIR) for a listener position in a virtual auditorium


Figure 3 shows the WXYZ echograms that are generated when B-Format is selected as the receiver model. As with all RIRs, these are specific to each listener seat. These RIRs preserve the direction of sound travel as well as arrival time and level. A software decoder can process the WXYZ signals for playback on the desired loudspeaker configuration, including stereo, 5.1 surround and others. It is necessary to generate RIRs for a sufficient number of seats to assure that the system performance is adequate.

An advancement of this concept is implemented in CATT-Walker™. A matrix of listeners is placed on the audience plane and the B-Format echogram is generated for each. The user can then select a WAV file for auralization with these responses. The B-Format RIRs are used for real-time auralization as the listener moves around the auditorium (using the arrows keys on the keyboard). Seating positions that fall in between the actual listener locations are generated using a triangulation method. The designer is literally free to “walk the coverage” of the proposed design, complete with head turning via mouse movement.


There is often a disconnect between the listening process and the design process. At first exposure they seem like two unrelated exercises. Many system designers proceed by intuition and listening alone, because these are the tools that they are most comfortable with. Walk-through convolution as implemented in CATT-Walker brings the design process full circle by allowing the designer to walk the coverage and listen to their design before it is installed, while assessing speech intelligibility and other factors along the way. This allows the familiar tools of listening and intuition to be used just like they are in live spaces. All that is needed are a decent PC, some software tools and some expertise in using the tools that comes from self-study, practice and experience. pb