The first installment of this series discussed the methods used to obtain the impulse response of a system. The “system” can be an electronic device, transducer or even a room. It can also be the combination of several systems. We will now turn our attention to interpreting the impulse response and applying the information to sound reinforcement systems.

The Impulse Response

Let us use as an example the impulse response of a sound system / room combination. We will proceed upon the assumption that the response has been gathered in a correct manner, utilizing a method that provided sufficient signal-to-noise ratio and repeatable measured data. Such an impulse response might look something like Figure 1.

First, remember how we obtained the data. A well-de fined stimulus was input into the system, and the output of that system was sampled and compared to the input, yielding the transfer function of the system. The “squiggles” of Fig. 1 are the electrical fluctuations above and below zero volts measured at the output of the test microphone. Zero volts corresponds to ambient atmospheric pressure. If we assume that the microphone has a linear response, then our display could just as well be labeled in Pascals (sound pressure) rather than volts.

With that assumption, what we see in the figure is the deviation of the air pressure above and below the ambient atmospheric pressure. A positive deviation means a compression of air particles and a negative deviation means a rarefaction of air particles, plotted as a function of time. As Dick Heyser put it:

The sound that we are interested in characterizing is a restoration to equilibrium conditions of the air around us following the disturbance of that equilibrium fry an event. Sound is what happens when air gets pushed.

What can we tell from this display? Unfortunately, without further processing the information yielded is minimal. The manner in which the human ear/brain system process this information requires that we process it into a more meaningful form. It must be emphasized that the following manipulations are ‘post processes” on the impulse response. Since some operations require discarding part of the data, the “raw” impulse response should be kept on file for future reference. Even so, the manipulation of the impulse response is essential and justified if we are to draw conclusions concerning the sound that we are examining.