jouml / nchlt-praat-plugin

Praat plug-in for exploration and acoustic analysis of the NCHLT speech corpora of South African languages

Praat NCHLT plug-in

Description

This plugin enables Praat to search in the orthographic transcriptions of the NCHLT Speech Corpus and open the audio files of corresponding search results. The NCHLT Speech Corpus contains orthographically transcribed broadband speech corpora for all of South Africa’s eleven official languages. Each language corpus consists of a large training suite (trn) and a smaller test suite (tst). The NCHLT Speech Corpus must be installed on your machine (at least one language) before using this plugin. The corpus may be obtained from SADiLaR.

After launching from the Praat Open menu, you first select one of your installed languages, pick one of the suites, and specify a search pattern. The XML transcription is loaded into Praat and searched for the pattern. You can further refine the search using speaker attributes (gender, age, location). When you are done with searching you can view the results in a table (including orthographic transcription, speaker ID, age, gender, and location) and open corresponding audio files for acoustic analysis.

Manual

A detailed manual is included: manual.pdf

Recquirements

• Praat 5.4.x or newer
• NCHLT Speech Corpus (at least one language)

Installation

• quit Praat
• Click Clone or download button and download ZIP archive (or click here: Download ZIP)
• unzip, change into the newly created directory NCHLT-Praat-Plugin-master, and copy the included sub-directory plugin_NCHLT to the Praat preferences directory:
  • Linux: /home/your-username/.praat-dir/
  • macOS: /Users/your-username/Library/Preferences/Praat Prefs/
  • Win: C:\Users\your-username\Praat\ or C:\Documents and Settings\your-username\Praat\
• launch Praat
• select the new item in Praat's Open menu: NCHLT: Search corpus...

Usage

When launched the first time, the plugin asks for the base directory of the NCHLT Speech Corpus. This is the directory/folder that contains the language-specific sub-directories called nchlt_<ISO 639-3> (old format) or nchlt.speech.corpus.<ISO 639-3> (current format) (<ISO 639-3> is a placeholder for the 3-letter language code):

base directory
│
├──nchlt_<ISO 639-3>
│	├── audio
│	│   ├── <spk_id>
│	│   │   ├── nchlt_<ISO 639-3>_<spk_id><gender>_<file_number>.wav
│	│   ...
│	└── transcriptions
│	   ├── nchlt_<ISO 639-3>.trn.xml
│	   └── nchlt_<ISO 639-3>.tst.xml
│
├──nchlt.speech.corpus.<ISO 639-3>
│  ├──...
│  ...
...

After selection, the base directory path is stored in the file CorpusDir in the plugin directory. That means you have to do this only once and the plugin is aware of your corpus directory in the future. If you move your corpus to an other location or if there are problems with base dir detection, just delete CorpusDir in the plugin directory. You'll be asked for the base dir the next time you launch the plugin.

The plugin scans the corpus base directory for installed languages and presents a menu to choose a language from (if you add a new language to your corpus, it will be recognized automatically the next time you launch the plugin). Then you have to select a suite (training or test or both) and specify a search pattern. Search patterns have two modes: simple and regex. If you choose simple search pattern all lower-case characters in a pattern are regarded as literal characters. Additionally, you may use the upper-case character "V" which isn't interpreted literally but represents a set of characters, namely vowels: [aeiou].

Some examples:

pattern	match examples
VpV	api, kupo, jepas ...
khV	khi, akho, teskhu ...
tVpV	tapi, stupo, tepas ...

Prohibited are upper-case characters (except for "V"), special symbols (like &%$), punctuation marks (like ?!,.), and spaces.

If you choose regex search pattern you can assemble a regular expression as complex as you wish. Regex syntax is described in the Praat regular expressions tutorial.

After you start the search, please be patient. Some XML files are pretty large and it may take some seconds (depends on your machine) to perform the search. The search results (if any) are available as a Praat table object. A dialog window will open (and stay open) to let you refine your search and/or view the table. You can browse through all results and perform acoustic analysis either one by one or you pick specific items identified in the table.

Changelog

v1.0

• support for new SADiLaR folder naming pattern
• available on Github

v0.4

• detailed manual added

v0.3

• Python no longer required (slower but more robust)
• regular expression search
• filtering of search results

v0.2

• usability improvements and bug fixes

v0.1

• initial release
• XML parsing with Python ElementTree

License

This is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

This software is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this software. If not, see http://www.gnu.org/licenses/.