Praat for Windows

Praat is Dutch for "talk." It's also the name of this lab-quality phonetics freeware that can record and analyze speech sounds in mono or stereo. While it's not difficult for lay users to pick up, it's designed for linguists and students of linguistics as well as other scientists and researchers who need to analyze human speech. It supports speech synthesis, including articulatory synthesis. It also creates high-quality images of data suitable for publication.

Praat's interface consists of two parts, Praat Objects and Praat Picture. We started by opening the Help file, which offers an excellent introduction to Praat and an overview of its functions before delving into the science. We began by recording speech via the system microphone and sound card. Praat lets you make both mono and stereo recordings. Mono recordings are preferable for analyzing speech, but mono recording equipment is uncommon outside the lab. However, most handheld note-takers record monophonic sound, and they are commonly used in fieldwork. Still, we chose stereo. The process for recording sound is a bit complicated, though the manual indicates it's easier in Linux and Mac versions of the program. We were able to create sound recordings not only from an external source such as a microphone but also from a WAV file and even directly inside Praat via formulas. This intriguing capability let us create tones like sine waves with noise added. Once we'd recorded a sound, we could view it as a waveform (two waveforms for stereo signals) in the SoundEditor window; Save, Cut, and Paste it; and open longer files with a LongSound feature that keeps most of longer sound files on your disk instead of being written to memory every time you play them.

From there, Praat delves into various analysis methods, and quickly left us speechless despite its clear and detailed explanations. However, it's doubtful that anyone who can really make use of Praat will be unfamiliar with its tools or its techniques.