Plotting Pitch and Intensity Contours with Annotations in Praat

Added:

Programmatic Acoustic Analysis (Python/R) vs. Legacy Praat Workflows

While Praat is the traditional gold standard in phonetics, relying solely on its built-in GUI and legacy scripting language for plotting and annotation has significant limitations. Modern phonetic science increasingly favors programmatic workflows using programming languages like Python (via libraries like Parselmouth, Librosa, and Matplotlib) or R. These modern tools address several of Praat's drawbacks by offering fully automated, high-throughput batch processing, vastly superior and highly customizable visualization options, and seamless integration with statistical analysis pipelines. Furthermore, Praat’s default pitch-tracking algorithms are prone to tracking errors like octave jumps; alternative frameworks allow researchers to easily swap in more robust, state-of-the-art pitch estimators (such as YIN or deep learning-based trackers like CREPE). Transitioning from Praat to code-based environments fosters greater reproducibility, scalability, and alignment with modern data science practices.

Basic familiarity with the Praat user interface, including loading audio files and opening the Objects and Editor windows.
Fundamental concepts of acoustic phonetics, specifically what fundamental frequency (F0/pitch in Hz) and intensity (amplitude/loudness in dB) represent physically.
An understanding of Praat TextGrids and how interval and point tiers are used for phonetic segmentation and annotation.
Basic knowledge of digital audio properties, such as sampling rates and how background noise can affect pitch tracking.
Praat scripting to automate the extraction of pitch, intensity, and duration measurements over large acoustic datasets.
Exporting raw acoustic data coordinates to R or Python for advanced plotting (e.g., using ggplot2 or matplotlib) and statistical modeling.
Applying pitch and intensity contours to prosodic analysis frameworks, such as ToBI (Tones and Break Indices) for intonational phonology.
Advanced acoustic visualization techniques, such as tracking and plotting formant trajectories (F1, F2) to analyze vowel quality over time.
37.4K views512likes10:22@ethanweedOriginal Release: 2016-04-08

This tutorial demonstrates how to extract pitch and intensity contours from audio recordings in Praat software and plot them together with word-level annotations to analyze suprasegmental features of speech, including the steps to import sound files, extract acoustic measurements, create text grids for annotations, and customize multi-axis plots for visual comparison.