Hi Slide, I think your text is directed to me (I’m Chris in Japan – and yes the photos are from my various rooms in Japan).
Yes, actually all the column widths are about 5″ in width (in all models).
I’m not sure I can explain the -6db point well or not…I’ll try to explain what I have understood. I believe it is understood that if a wave length is shorter than the width of the radiating plane, the sound wave will ‘beam’ directly off the membrane (and not have any radiating angle). And yes, a 2700 hz sound wave is basically 5″ as you wrote. If you take one cell for example, the mylar at the center of the cell will radiate an amount of energy and moving out to the sides of the cell the mylar will move less than at the center and therefore radiate less energy. Roger has this calculated to be a reduction of -6db if you were to compare the center to the side and this is for 20khz. As you can imagine, beaming and ‘not-beaming’ is progressive – it is not just an ‘on-or-off’ situation. If what I have described above is true for 20khz, frequencies less than 20khz will have more and more tendency to radiate with more angle. But let’s go back to 20khz and say there is a 6db drop in energy between the center and the sides of the cell. So at one side of a cell and the next adjoining cell there is a mergining of two points which are radiating at -6db which in essence are merging and providing the same energy as from the center of any one cell. This is what Roger explains is actually ‘over-kill’ to make sure that there is no ‘picket-fence’ effect if you were to move around in front of the speaker (it is built to sound as there is no drop in frequency energy between cells from 20khz and lower…)
I am interested to know if you can follow my explanation, and if anyone disagrees or has any comment, that is most welcome too of course.