Another thought occurred to me for those geeking out on the internals. I take all the samples regardless of pressure rising or falling on the sensor. I think I may try pulling out the rising measurements and just smoothing the falling measurements and see what that gets me. Also, I think I can smooth the sampling better as well but need to think about how to best do that - with smoothing being a mathematical function as opposed to just adding delay or changing the number of samples taken.