You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
That's the algorithm to determine where the next time point should come when we convert a pitch tier to a series of time points. A naive algorithm would be: if the latest point is at time t, and the current pitch is f, then the next point should be at time t' = t+1/f. This is incorrect, because the pitch may have changed between these two times. The correct algorithm should be such that the area under the pitch curve between t and t' is exactly 1. This means that if the pitch changes linearly from f to f' (between t and t'), t' and its accompanying f' should be chosen in such a way that (t' - t) times 0.5*(f+f') equals 1, where the last expression means the average pitch between t and t'. The area under the linear curve is a quadratic function of t' (the definite integral of the pitch curve from t to t'). hence those "slope" (the slope of the pitch curve) and "discriminant" (the discriminant of the solution of a quadratic equation).
Does that help, or do you want to see the derivation of the precise solution as well?
Hi,
I am trying to understand the maths behind Praat. Would you be able to explain where the algorithm here comes from?
Many thanks!
Tim
The text was updated successfully, but these errors were encountered: