Voice Report Generator — User Guide
Generates a standard Praat Voice Report for a selected sound object, detailing measures like jitter, shimmer, and other vocal features.
What this does
This script executes Praat's comprehensive Voice Report analysis on a selected Sound object. The Voice Report is a crucial analytical measurement in phonetics and speech science, providing several key vocal metrics including: mean fundamental frequency (F0), jitter (perturbation of F0), shimmer (perturbation of amplitude), degree of voice breaks, and measures of harmonicity.
What is the Voice Report? It's an automatic analysis feature in Praat that performs multiple steps—pitch extraction, cycle marking, and perturbation analysis—to quantify characteristics of vocal quality, often used to diagnose dysphonia or study prosody.
The script uses Praat's robust auto-correlation (CC) method for initial pitch tracking.
Quick start
- Load a Sound object (preferably a sustained vowel or clean speech) into Praat.
- Select the Sound object in the Praat Objects window.
- Run the script: Praat → Run script… →
voice report.praat. - The full report will be printed automatically to the Praat Info window.
Output: The report is a multi-line text output in the Info window containing all measured vocal parameters. Only the original Sound object remains in the Objects window after execution.
Important: This script requires a selected Sound object to run. If the sound is highly noisy, recorded at low quality, or contains significant unvoiced segments, the pitch tracking and resulting report metrics may be unreliable.
Implementation Details
Analysis Steps
The script performs the analysis in three sequential steps:
- Object Creation: The selected Sound object is first used to create a temporary Pitch object using the auto-correlation (CC) method.
- Cycle Marking: The Pitch object is then used to create a temporary PointProcess (CC) object, which accurately marks the beginning of each glottal cycle.
- Report Generation: The final Voice Report command is executed simultaneously on the original Sound, the temporary Pitch, and the temporary PointProcess objects.
- Cleanup: The script concludes by removing the temporary Pitch and PointProcess objects from the objects list.
Fixed Voice Report Parameters
The core of the analysis uses the Voice report command with the following fixed parameters:
| Parameter | Value | Description |
|---|---|---|
| Time range | 0 s to 0 s | Analyzes the full duration of the sound (0 to 0 means entire duration). |
| Minimum F0 | 75 Hz | Lowest acceptable fundamental frequency for pitch tracking. |
| Maximum F0 | 600 Hz | Highest acceptable fundamental frequency for pitch tracking. |
| Max period factor | 1.3 | Maximum factor by which adjacent periods can differ. |
| Max amplitude factor | 1.6 | Maximum factor by which adjacent amplitudes can differ. |
| Max period perturbation | 0.03 | Threshold for excluding periods from jitter analysis (3%). |
| Max amplitude perturbation | 0.45 | Threshold for excluding periods from shimmer analysis (45%). |