The accuracy of formant frequency measurements: a comparison of spectrographic analysis and linear prediction.

The accuracy of spectrographic techniques and of linear prediction analysis in measuring formant frequencies is compared. The first three formant frequencies of 90 synthetic speech tokens were measured by three experienced spectrographic readers and by linear prediction analysis. For fundamental frequencies between 100 and 300 Hz, both methods are accurate to within approximately +/- 60 Hz for both first and second formants. The third formant can be measured with the same degree of accuracy by linear prediction, but only to within +/- 110 Hz by spectrographic means. The accuracy of both methods decreases greatly when fundamental frequency is 350 Hz or greater. These limits of measurement appear to be within the range of the difference limens for formant frequencies.