For those who want to experiment further, there were a couple of companies (I think Meitner was one of them) who worked on flexible panels using something similar to exciters. My memory is that at least one of them used mylar or polyethelene or something similar, damped with a layer of foam. Their explanation of how they worked was that rather than being a rigid piston, they allowed the wave to propagate outward, with the air being excited as it went, with delay as it went out acting as if the wave in the air had come from behind, from a spherical source. This explanation has some similarities to how Ohm explained their driver, and how Quad explained the concentric segments of the ESL 63, though obviously it was different. I think patent troubles boliixed up going into production, but Speaker Builder had an article on making a version yourself (the author was at pains to point out that if you tried to do it commercially, you'd probably get sued). I made a small demonstration one, and a bigger trial one. The small one was rather nice, but I never got the tension right on the big one. Interesting experience, though.