Towards the beginning of a long audio recording, I can see the correspondence between loud sounds, like sneezes, with spikes in the visual display of the sound. However, over time those get further and further apart; the visual spike precedes the sound by about a second by the 3 minute mark. This makes it hard to segment the recording, since without the visualization it's hard to anticipate verbal pauses.
I'm running version 1.5.1, under Mac OSX 10.5.7, on a 2 GHz Intel macbook with 2G of RAM. Other than that, the software's great, so thanks for your hard work on this!