The errors associated with two formant‐frequency estimators are compared theoretically for the simple case of a single steady‐state isolated formant. The estimators considered are (a) the zero‐crossing frequency, and (b) the frequency of the highest‐energy harmonic. The results of the analyses show that for a given harmonic the zero‐crossing counter gives the smallest magnitude error while the highest‐energy, harmonic tracker gives the smallest percentage error.

This content is only available via PDF.
You do not currently have access to this content.