An impulses response measurement is an indicator of a loudspeaker's performance in the time domain. It is a fact of nature that the impulse response and the frequency response both contain exactly the same information but exhibited in two different ways. Because of this one is tempted to view one or the other as the more necessary requirement for good sound reproduction. The problem is that some things that are highly audible are "masked" or not readily evident in one domain versus the other (frequency versus time).

For example a slightly delayed signal as would occur with cabinet or waveguide diffraction or reflection may have only a very small effect in the frequency domain - perhaps barely even noticeable, while in the time domain it can be readily seen as a delayed impulse of energy. In the time domain it can be difficult to see peaks and dips in the frequency response which would be clearly audible. Yes, the two domains are equivalent in theory, but in practice they both have important uses.

The main feature of the impulse response to consider is how compact it is in the time domain. An accurate impulse response, like that of Ai speakers, shows a very tight energy signature in time composed of a double impulse which decays rapidly. There should not be any tails or delayed impulses from diffraction etc. The ideal loudspeaker response is a doublet with only a very small (but long) ringing resulting from the low frequency cut-off. The doublet response (a dual impulse both positive and negative) is required since no speaker can produce DC - there must be no net impulse when averaged over time. The time average of a loudspeakers impulse response must go to zero, which a single impulse would not.

The ESP series of loudspeakers all have extremely short impulse responses usually not found in systems with horns or systems with high efficiency.