Polarity

One piece of information that the impulse response can yield quite accurately is the polarity of the transducer. To keep things simple. The loudspeaker used was a simple piston radiator, a 4” cone in a sealed box. When such a loudspeaker is fed a signal from a properly polarized sound system, it’s first movement from rest will be forward toward the listener in response to a positive going applied voltage. This creates a positive pressure wavefront that will be the first part of the sound wave to arrive at the microphone, causing an inward deflection of the microphone diaphragm.

This in turn, produces an initial spike that is pointing upward. If the wires connected to our loud speaker were reversed, the initial spike would be down ward, indicating that the wavefront had a negative-going pressure component that first arrived at the microphone. This polarity information is obscured by most of the post-processes that we will perform on the impulse response. So, the first step in interpreting the impulse response might be to “zoom in” on the leading edge and determine the polarity of the transducer (Figure 2).

Multi-way loudspeakers can yield a response that is more difficult to interpret. First, since the transducers typically emit from different points in space (due to construction and physical location), their arrival times will be displaced in time on the impulse response display. If some of the transducers have correct polarity, but others are reversed, the result will be a curve that fluctuates wildly above and below the ambient. It may be necessary to disconnect the devices and measure them independently to determine the polarity of each (Figure 3).

Step Response

We will now introduce a post process that can make the time displacement of the transducer(s) easier to determine. The step response of a two-way loudspeaker system is shown in Figure 4. This is the impulse response after integration. The time offset of the woofer and tweeter is clearly visible, indicating a loudspeaker system that is not “signal aligned” to synchronize the arrival of the woofer and tweeter. Such a display might be used to adjust the physical position of the transducers, or to calibrate a delay device to correct the offset.

An impulse response usually starts at time zero, with the first sound arrivals delayed by the propagation distance through the air and any electronic delay in the signal processing chain. Following the first energy arrivals, the reflections from the room will be visible. The longer that the room “stores” acoustic energy, the longer the impulse response must be to record it. Once the complete decay of sound energy in the room is recorded, it can be post processed to yield informative information about the loud speaker and/or room.

Let us now square the impulse response, take the square root, and display the result on a logarithmic vertical scale. This yields the absolute value of the data displayed on a log scale. The polarity information is lost in this process, but we gain the ability to examine the sound arrivals at the microphone on a relative or absolute dB scale. The audible effect of a reflection can be evaluated much more easily when displayed in this manner (See Figure 5 below).