Whispers of a working chip, page-167

Read through the patent, here's my take. Per the diagrams (as well as diagrams in their other docs), there are a significant number of holes/gaps in the top/bottom surfaces for air to flow through, not to mention over the edges of the sprung membrane as it moves. I can't see how there's much compression occurring because of this. Interestingly, the fact that the air moves over the edges of the membrane is actually a major difference to conventional speaker drivers which need a tight seal with an enclosure to ensure the front and back of the speaker operate in separate air spaces. I suspect AKP pixels don't need to be sealed as they always move the same distance, i.e. produce the exact same output.

I could of course be mistaken.