The automatic annotation and analysis of speech

Download SPPAS

Download Plugins

Phonemes Classifier

Create articulatory annotations from the time-aligned phonemes: class, place of articulation, manner of articulation, phonation, etc.

Audio Segmenter

AudioSegmenter is a tool to segment audio files into several tracks. Bounds of the tracks are indicated in an annotated file of any format supported by SPPAS (xra, TextGrid, eaf, ...).

Download Plugins -- DEPRECATED

ALL THESE PLUGINS NEED TO BE UPDATED. THEY WON'T WORK WITH ANY RECENT VERSION OF SPPAS.

IPUs Cleaner

This plugin cleans the IPUs after the orthographic transcription was performed: it removes the non-transcribed IPUs, re-indexes, it merges consecutive silences, etc.

Download the IPUs Cleaner plugin

Sampa to IPA

This plugin converts transcriptions from the SAMPA (Speech Assessment Methods Phonetic Alphabet) annotation to the International Phonetic Alphabet.

Download the SAMPA to IPA plugin

Stats Groups

This plugin estimates statistics (occ, mean, stdev, intercep and slope) on sequences of numerical intervals.

Download the Stats Groups plugin

Audio concatenate

Word Concatener is a tool to concatenate into a new audio file, the audio chunks corresponding to the occurrences of a word or a sequence in an audio file. Bounds of such tracks are indicated in an annotated file of any format supported by SPPAS (xra, TextGrid, eaf, ...).

Download the Audio Concatenate plugin

MarsaTag

MarsaTag is a Part-Of-Speech tagging for French developed at Laboratoire Parole et Langage, by S. Rauzy and G. De Moncheuil. MarsaTag is available for download at: https://www.ortolang.fr/market/tools/sldr000841
This plugin allows running MarsaTag on time-aligned tokens generated by the automatic annotation 'Alignment' of SPPAS.

Download the MarsaTag plugin

To propose a new plugin...

Please contact the author

ESPPAS (deprecated)

ESPPAS (in 2012) was supported by ANR OTIM project Ref Nr ANR-08-BLAN-0239. This was an add-on to SPPAS to deal with Enriched Orthographic Transcriptions. It was using LIA_Phon to perform grapheme-to-phoneme conversion.

Send an e-mail to the author to get it (but it is not compatible with any current version of SPPAS).