Moving on to the measured frequency response, this was taken on an axis level with the midpoint of the electrostatic panel, 42" from the floor. Visually smoothing out all the little ripples, the SL3's response appears to shelve down above 1kHz. Remember, however, that there will be an interaction between the speaker and the measured response due to the relatively close microphone distance. There will be a "proximity effect" that raises the speaker's apparent midrange output the closer the microphone is to the speaker.

This is graphically shown in fig.3, which shows the on-axis response measured at my usual 50" mike distance (bottom trace) and at 108" (top), the farthest away I could place the microphone in the Stereophile listening room where I perform my speaker measurements. You can see that not only does the midrange dramatically shelve up at the closer mike position, but that the speaker acquires more presence-region energy at the farther position. What is true for the microphone will also be true for the listener. In small rooms, the balance of the SL3 will depend very much on how far away the listener sits.

Fig.3 Martin-Logan SL3, anechoic response on mid-panel axis at 50" (top) and at 108" (bottom) (5dB/vertical div.)

Vertically, the Martin-Logan seemed quite uncritical of exact listening height. Sitting with your ears anywhere between 32" and 48" from the floor will give similar balances. Laterally, however, fig.4 shows that while the curvature of the panel gives reasonably good dispersion below 1kHz, the speaker is still very beamy in the treble (other than in the 10kHz region, where the speaker actually puts out slightly more energy off-axis). As a result, the Martin-Logan's balance might sound rather uninvolving in rooms that are heavily furnished. Dispersive room surfaces will be better than absorptive. Much of the ragged-looking behavior at extreme off-axis angles is due to the presence of reflections of the panel's output from the tapered "wings." This should be subjectively inconsequential, however.

Fig.4 Martin-Logan SL3, horizontal response family at 50", normalized to response on mid-panel axis, from back to front: differences in response 90 degrees-5 degrees off-axis; reference response; differences in response 5 degrees-90 degrees off-axis.