North Carolina Sociolinguistic Archive and Analysis Project (NCSLAAP)
A digital audio archive of North Carolina speech with analytical tools designed for linguists.
Overview
Professor Walt Wolfram of the NC State University Linguistics program in the English department, director of the North Carolina Language and Life Project, has been interviewing and recording North Carolinians since the 1960s for the purpose of studying North Carolina speech. By making this collection of more than 1100 recordings available in a digital archive, we are able to preserve the recordings, make them more widely available online (when permissible), and greatly facilitate linguistic analysis. Linguists will be able to search and browse recordings and transcripts by subject demographics and other criteria; to locate, extract, and save relevant sound clips; and to perform advanced analyses of pitch, variance, and other variables. The project began in the spring of 2005 and is due to be completed in the summer of 2006.
To explore the project, visit http://ncslaap.lib.ncsu.edu and log in as user "guest" with no password.
Team and Partners
Tyler Kendall, built the MySQL database and PHP interface and provided Linguistics subject expertise. Email: tsk3@duke.edu
Walt Wolfram, William C. Friday Distinguished Professor, directs the North Carolina Language and Life Project.
The North Carolina Language and Life Project records and transcribes the audio to be archived.
The Digital Media Lab provides equipment and staff for digitizing the audio.
How We Did It
Approximately 1100 field recordings on audio cassette, reel-to-reel tape, CD, video, and digital video are being digitally archived in a MySQL database with a PHP interface. Analog tapes played on Marantz PMD222 cassette decks are being digitized with a single (mono) channel at a 44.1kHz sample rate and a 16-bit bit depth using Peak 4.0 on Macintosh G5 workstations; files are stored in .wav format, but users can automatically extract segments as .mp3 files. The database has been integrated with the open-source phonetic analysis program Praat. All components of the project currently reside on a Macintosh G5 but will be moved to an unknown production server in the NC State University Libraries. There are tentative plans to adopt the Library of Congress's Metadata Encoding and Transmission Standard (METS).