Loudspeakers

Measuring a horn speaker such as the La Scala AL5 using a quasi-anechoic, time-windowed technique is complicated by the time delays associated with the path lengths through the midrange and bass horns. If the measured impulse response is windowed to remove the first (floor) reflection as usual, you end up measuring the tweeter horn alone as the midrange and bass sections have barely responded. So for the forward frequency responses, measured at 1m on the axis of the tweeter horn, I've extended the time window to include output from the midrange and bass horns and used third-octave (rather than tenth-octave) smoothing to quell the response ripples resulting from the included reflection.

The resulting response traces [Graph 1] should be viewed with some caution but are reliable for determining trends. The ripples in the lower midrange reflect the highly uneven impedance vs. frequency trace in this region (not shown) and the CSD waterfall [Graph 2], generated from the nearfield measurement of the bass horn, clearly shows that the La Scala has multiple resonances here. Generating the CSD conventionally, to view tweeter performance only, shows resonances there too. Bearing in mind the foregoing caveats, we couldn't match Klipsch's claimed sensitivity of 105dB SPL for 2.83V/1m but our pink noise figure of 101.6dB is still impressive, particularly as the AL5 is not tough to drive. Although Klipsch's '8 ohms compatible' doesn't square with our measured impedance minimum of 3.9ohm, the minimum EPDR of 2.0ohm at 60Hz is fairly amp-friendly, although a second dip to 2.0ohm at 377Hz complicates the picture very slightly. KH