Would it be accurate to assume that the M2 waveguide "pinches" the waveform in the horizontal and vertical plane? It looks like two sets of fingers is pinching the mouth.
I think those "fingers" gently introduce diffraction, which actually widens high frequency coverage in those directions. But arguably most of the sound from the compression driver goes in between the fingers, and it is not clear to me what the radiation pattern looks like in those directions (especially along the diagonals). Listening tests have confirmed that the M2 is superb, so whatever happens at other angles is obviously not a problem. It also looks to me like that waveguide's geometry avoids detrimental internal reflections. And if there are any, I don't think they would be in the direct sound - I think they'd careen off in harmless directions. Kudos to Charles Sprinkle!
I remember reading somewhere that Charles Sprinkle no longer believes in constant directivity as the best solution and that he now thinks smooth but steadily rising directivity is preferable. I asked directly and he said it had a lot to do with needing a high-shelf filter from around 2khz with the JBL M2 in normal/small rooms to sound natural that led him onto that path.
VERY INTERESTING that Sprinkle finds (presumably gently) rising directivity to be preferable to a gently downward-shelved direct sound (with constant directivity) in a small room. Thank you for passing that along.
The approach I use is constant directivity but with a gently downward-tilting direct sound, instead of a "flat" direct sound. In my opinion this gives a natural-sounding tonal balance while minimizing the spectral discrepancy between the direct sound and the reflections, which in turn is arguably beneficial to timbre. My designed-for listening axis is about 20 degrees off the horn's centerline (after Geddes), so there is some tilt-down of the highs in the direct sound already happening. The horn's inevitably slightly "hotter" on-axis energy is actually making its first appearance at the central "sweet spot" as reflected sound, thereby tipping the spectral balance of the reverberant field up a bit, which reduces the net discrepancy between it and the direct sound.
Of course I realize this goes against Harman's findings of what listeners prefer (flat on-axis, tilt-down off-axis), and unfortunately for my credibility, it also goes against what Charles Sprinkle now believes to work best. Hmmm. Don't tell my marketing department, they'll demand a bigger budget.
I have JBL M2's now and had Kii Three previously, both more or less constant directivity designs. Can't say I feel they suffer from it. Is this perhaps one of the more academic discussions where the issue is solved by somewhat different tuning?
I tried working with the tweeter Kii uses years ago, back before the Kii Three came out. Measurements were superb, but I heard a slight edginess in the sound which imo became distracting at high levels, and nothing in my measurements explained it. [speculation] The diffraction introduced by the fingers in the Sprinkle waveguide is incoherent enough to be perceptually benign, while the diffraction introduced by the sharp edges between the concentric flat rings in the Kii tweeter's waveguide is coherent enough to not be. [/speculation]
Last edited: