Compute Predefined Local Measurements

Audio signals have varying features over time. Therefore, local measurements are more insightful and applicable than global measurements. This example shows how to use AudioLocalMeasurements to analyze speech formants, which represent the distribution of frequencies filtered by the vocal tract.

Compute the max of the signal on each partition and display it on top of the waveform.

Compute the formants of the signal on each partition and display them on top of the spectrogram.

The recording environment can boost or attenuate certain frequencies, thus altering the formants. Use AudioReverb to simulate reverberation from a different environment.

Compute and visualize the new formants.