Its not impossible. If I was counting the frequency you are right the resolution would be depended on cycles per second. But, I am measuring the wave front and when it hits the two microphones. Therefore, this resolution is attainable. In fact microphones are used to measure wave propagation already but just for different applications. Check this thread: http://arduino.cc/forum/index.php/topic,7956.0.html. He uses a 2Mhz clock. Anybody else has any suggestions?