Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

Share this Page URL

1.3.5 Preparation > Dissecting the Impulse Response - Pg. 26

26 SecTion | 1 Test and Measurement Dissecting the Impulse Response The audio practitioner is often faced with the dilemma of determin- ing whether the reason for bad sound is the loudspeaker system, the room, or an interaction of the two. The impulse response can hold the answer to these and other perplexing questions. The impulse response in its amplitude versus time display is not particularly useful for other than determining the polarity of a system com- ponent, Fig. 1.15. A better representation comes from squaring impulse response (making all deflections positive) and displaying the square root of the result on a logarithmic vertical scale. This log-squared response allows the relative levels of energy arrivals to be compared, Fig. 1.16. The Envelope-Time Curve Another useful way of viewing the impulse response is in the form of the envelope-time curve, or ETC. The ETC is also a contribution of Richard Heyser. 2 It takes the real part of the impulse response and combines it with a 90 degrees phase shifted version of the same, Fig. 1.17. One way to get the shifted version is to use the Hilbert Transform. The complex combination of these two signals yields a time domain waveform that is often easier to interpret than the impulse response. The ETC can be loosely thought of as a smooth- ing function for the log-squared response, showing the envelope of the data. This can be more revealing as to the audibility of an FigURe 1.15 The impulse response, SIA-SMAART.