Speech annotation and analysis made simple

Install SPPAS, open the Dashboard, manage your files, annotate automatically, analyze results, and edit data in a complete graphical interface.

Why choose SPPAS?

Easy installation with a Setup
A clear Dashboard to get started
A workspace to organize files
Automatic annotation in a few clicks
Analysis tools for annotated data
Editors to inspect and correct files

SPPAS offers open source cross-platform, customizable automatic annotation and analysis solutions for audio and video media.

Brigitte Bigi, a computer scientist at Laboratoire Parole et Langage (LPL), in Aix-en-Provence, France, is the author of SPPAS.

Easy installation

Install SPPAS with a simple Setup. No command line required.

SPPAS installation setup window — Install SPPAS in a few minutes.

SPPAS dashboard interface — The Dashboard gives direct access to the main tools.

Clear starting point

The Dashboard helps users understand where to start and gives access to the main parts of the application.

Manage a workspace

Keep input files, generated annotations, and related resources together in the same working environment.

SPPAS workspace and file management screen — Manage files inside a workspace.

SPPAS automatic annotation screen — Launch automatic annotation in a few clicks.

Annotate automatically

Run automatic annotations from the graphical interface, without relying on command-line tools.

Analyze annotated data

Explore the results of annotations with dedicated tools designed for speech data analysis.

SPPAS analysis screen — Analyze annotated files directly in SPPAS.

SPPAS annotation editor — Edit and inspect annotation files.

Edit your files

Inspect, correct, and refine annotation files with the editors included in SPPAS.

SPPAS Web Apps

SPPAS 5.0 introduces a new Dashboard that progressively replaces parts of the previous wxPython interface with a modern web-based UI. Some features are now provided as standalone applications that can run either inside SPPAS or directly online.

The first SPPAS Web App is available below. Additional applications will be progressively added.

TextCueS

Generates the sequence of cues to code from a written text.

Open

Why should you trust SPPAS?

Reliable software

The application behaves as documented and tolerates unexpected user actions.

Local processing

SPPAS runs on your computer. No data are uploaded or collected.

Actively maintained

Continuous development, bug fixes, and support for new languages.

Evidence and Impact

Documentations:
- The SPPAS book: 160 pages
- The resources documentation: 60 pages
- The XRA file format: 15 pages
- The transcription convention: 5 pages

Cited 585+, including:
- cited by 214 references: SPPAS - Multi-lingual Approaches to the automatic Annotation of Speech
- cited by 140 references: SPPAS: a tool for the phonetic segmentations of Speech
- cited by 129 references: SPeech Phonetization Alignment and Syllabification (SPPAS): a tool for the automatic analysis...

How to cite?

Brigitte Bigi.
SPPAS - MULTI-LINGUAL APPROACHES TO THE AUTOMATIC ANNOTATION OF SPEECH.
The Phonetician. Journal of the International Society of Phonetic Sciences, 2015,
Journal of ISPhS/International Society of Phonetic Sciences, 111-112 (ISSN:0741-6164), pp.54-69.

Brigitte Bigi, Béatrice Priego-Valverde.
Search for Inter-Pausal Units: application to Cheese! corpus.
9th Language & Technology Conference: Human Language Technologies
as a Challenge for Computer Science and Linguistics, May 2019,
Poznań, Poland. pp.289-293. ⟨hal-02428485⟩

References: 30
- 19 about the annotations
- 4 about the software tool itself
- 3 about the linguistic resources construction
- 2 about the analyses
- 2 about the data representation

SPPAS is Open Research Software

SPPAS was awarded by the French Ministry of Higher Education, Research and Innovation at the 2022 Open Source Research Software Competition.

French Open Science Award
Honourable Mention for the 2022 Special Jury Prize

Award ceremony with Brigitte Bigi and Sylvie Retailleau, French Minister of Higher Education and Research — Award ceremony: Brigitte Bigi and Sylvie Retailleau, French Minister of Higher Education and Research.

SPPAS vs other tools for Speech Segmentation (forced-alignment)

Handles disfluencies and non-speech events
Peer-reviewed evaluation of all processing steps
GUI + CLI + Python API
Multilingual and extensible
Fully open source (code + linguistic resources)

SPPAS differs from other speech segmentation tools in several aspects. The comparison below summarizes the main differences. For a detailed comparison (in French) follows this link.

Detailed feature comparison with other tools

Comparative Table – Speech Segmentation Feature Support
Feature	SPPAS	Others (e.g. WebMAUS, MFA, torchaudio, ...)
Fully open source for both source code and all linguistic resources	✔	✘
Dedicated label for filled pauses (e.g., “fp” for “euh” in French)	✔	✘ (mapped to nearest vowel)
Alignment of non-speech events like laughter	✔	✘
Manual control of each processing step	✔	△
Peer-reviewed evaluation of all processing steps	✔	✘
Handles disfluencies (fragments, repetitions, unknown words)	✔	△
Alignment of long segments beyond Inter-Pausal Units (IPUs)	✘	✔
Both multilingual (FR, EN, IT, ZH...) and extensible by users	✔	△
Editable HMM-based acoustic models	✔	△
Import/export with multiple formats (TextGrid, ELAN, HTK...)	✔	△
Uses international phoneme standard (X-SAMPA)	✔	△ (tool-specific encodings)
Graphical user interface (GUI)	✔	✘
Command-line interface (CLI)	✔	△
Python API/interface	✔	△
Web-based interface	✘	△

Legend:

✔ = Fully supported
△ = Partially supported / varies by tool
✘ = Not supported or not explicitly implemented

SPPAS with Praat and ELAN

SPPAS is the perfect companion to Praat and ELAN. It automates time-consuming annotation tasks, while Praat and ELAN remain ideal for inspection, manual correction, and fine-grained manual analysis.

Use SPPAS to generate annotations automatically
Open results in Praat or ELAN
Inspect, correct, and refine your annotated tiers
Then use SPPAS to generate new tiers from your queries (not just tables) and compute statistical distributions

Current Supports

LPL — Speech and Language Laboratory

Since 2011

website

Firah — Applied Disability Research

2023 - 2026

website

A*Midex University Foundation

2025 - 2028

website

Past Supports

ORTOLANG - Investissement d'Avenir

2012-2019

ORTOLANG receives state aid under the « Investissements d’avenir » program (ANR–11–EQPX–0032)

Visit the website

CoFee - Conversational Feedback

2013-2015

Multidimensional analyses and modeling (ANR-12-JCJC-JSH2-006-01)

Visit the website

Variamu - Variations in Action

2014-2015

a Multilingual approach (Campus France - Procore PHC)

Visit the website

Adding Cantonese into SPPAS

2015-2016

In collaboration with PolyU (Campus France - Procore PHC)

PolyU website

NaijaSynCor - Common Nigerian Pidgin

2017-2021

a corpus-based study of the nature and functions of Naija in Nigeria (ANR-16-CE27-0007)

Visit the website

VAPVISIO

2018-2022

the training of language trainers in online environments using videoconferencing (ANR-18-CE28-0011)

The ANR website