11.4.1 Basic signal processing

Figure 11.12: An overview of a linear filter and its relationship to Fourier analysis. The top row of blocks corresponds to the time domain, whereas the bottom row is the frequency (or spectral) domain.

The importance of frequency components in sound waves should be clear by now. This remains true for the engineering problem of synthesizing sounds for VR, which falls under the area of signal processing. A brief overview is given here; see [11,191] for further reading. As the core of this subject is the characterization or design of filters that transform or distort signals. In our case the signals are sound waves that could be fully synthesized, captured using microphones, or some combination. (Recall that both synthetic and captured models exist for the visual case as well.)

Figure 11.12 shows the overall scheme, which will be presented over this section. The original signal appears in the upper left. First, follow the path from left to right. The signal enters a black box labeled linear filter and becomes distorted, as shown in the right. What is a linear filter? Some background concepts are needed before returning to that question.