Nice, but does it takes into account that some droplets are directly issued from the hose itself ? Not to mention the double stream problem ^u^ Cool paper nonetheless.