Using a concert grand piano, and using Matlab-based frequency analysis and the best-known inter-bin interpolation, the pitch of the note C2 drops 1.8 cents in 1.5 sec after the attack, and drops a total of 3.2 cents 3 sec after the attack. I was challenged by the assertion that the pitch does not drop after 0.5 sec. This is proof to the contrary. It all has to do with the nonlinearity of the piano string. Frequency decay along with amplitude decay is a known phenomenon in 'spring theory.'