MBROLA Hall of fame

The MBROLA team would like to celebrate the 10th anniversary of the MBROLA project. In order to emphasize the human network built around the project, we added this special page to our web site as well as the scrolling photo banner on the home page, in honour of the many people whose collaboration has made this project possible.

Not yet on this web page ?

Arthur Dirksen and Ludmila Menert

Arthur Dirksen and Ludmila Menert created the first Dutch voices for MBROLA (nl1 and nl2), then went on to found Fluency, which provides commercial text-to-speech software based on MBROLA. More info: www.fluency.nl.

Adina Svensson

I contributed with a Swedish female voice in 2002. I belong to the category "database provider". The database was created within the scope of my MA thesis in computational linguistics at Lund university.

Alain Ruelle

I provided my voice, aka fr7, to MBROLA. I also compiled MBROLA as a dll for MS Windows, and I developed the MBROLA tools for win32. I later contributed to MBROLA-based TTS systmes : EULER and now eLite

Alexas Stanislovas Girdenis and Pijus Kasparatis

Prof. Habil. Dr. Aleksas Stanislovas Girdenis, Corresponding Member of the Lithuanian Academy of Sciences, and Dr. Pijus Kasparaitis - providers of the Lithuanian diphone databases lt1 and lt2.

Antje Schweitzer

I'm a researcher at the Institute of Natural Language Processing at the University of Stuttgart, Germany. We provide a (non-commercial) German extension to the Festival Speech Synthesis System. My interests have been changing over past projects, including text analysis for prosody, prosody generation, speech database recording and management, and unit selection. Thus, I'm a TTS developer, MBROLA user and MBROLA database provider. :-)

Baris Bozkurt

Baris Bozkurt served as the Mbrola officer between August 2000 and September 2005, ie. he processed databases and provided support to Mbrola users. He also created two voices for Mbrola: Tr1 and Tr2(the two Turkish databases). He is with SVOX AG, Zurich since November 2005

Cesar Gonzales Ferreras

I am César González-Ferreras, from University of Valladolid (SPAIN) and I contributed to the MBROLA project as a database provider. The work was carried out with the help of some colleagues from my research group: Valentín Cardeñoso-Payo and David Escudero Mancebo. We designed a corpus with all required diphonemes, recorded it and segmented it. That corpus was used to build the voice es4.

Dimitrios Zachariadis

I have developed an MBROLA plugin for Wavesurfer, called speak.plug, which is listed in the "MBROLA Inside" page of your site. It adds speech functionality with natural voice intonation to Wavesurfer. It provides a connection to the Mbrola engine, to which it passes phoneme and voice intonation data entered or generated in Wavesurfer.

Fabrice Malfrere

I developed the MBROLIGN Project (a speech-to-text alignment tool) in 1995-1998. I designed a prosodic module used in Euler (and eLite) from 1996 to 1999. I also created the MBROLA French voice FR5, and various (commercial) voices for Babel Technologies (now Acapela Group). I work at Acapela Group since 1999.

Felix Burkhardt

Felix used Mbrola for his thesis about emotional speech synthesis and developed an emotional prosody transoformer called "Emofilt".

Francois Lanctot

With his company "Silex Creations Inc." (www.silexcreations.com), Francois Lanctot contributed to the creation of diphones databases in Canadian French (CA2, CA3 and CA5) as well as in Canadian English (CA4). They also created a platform for the MBROLA singing voice (some examples on www.klemradio.com) which combines a real-time registration module (MaxMBROLA with a MIDI musical keyboard) and a phonetic keyboard (selecting the diphones). Afterwards, an off-line module allows to refine the musical and interpratation parameters. This program provides also the functions from Felix Burkhardt (emotive speech) and can be used for normal voice.

Frederic Bechet

I've been developping Lia_Phon, a French Text-To-Speech system that is linked to MBROLA. Lia_Phon was first developped in 1995, so right at the beginning of the first release of MBROLA !!

Georges Sergiadis

Our team in the Telecommunications Laboratory of the Electrical Engineering Department at the Aristotle University, provided the Hellenic (Greek) database for the Mbrola project, in the effort to help blind people, and also our language and its Cultural Wealths to integrate in the new Information Society Era.

Joram Meron

I contributed the databases JP1, JP3, HB1

Karl Michael Schneider

I did not really contribute to the MBROLA project except using MBROLA. I use it mostly to announce the subjects of new emails I receive. I use a language recognizer to decide whether the subject of an email is in english or german. English subjects are announced using the Festival tts system. German subjects are announced using MBROLA together with txt2pho.

Kow Weng Onn

I'm the provider for the ma1 Malay female voice database. The ma1 database is currently being used with Fasih, our commercial Malay TTS that uses MBrola as its speech synthesizer. I'm also currently working with the MBrola team on adding 2 new voices (ma2 - male and ma3 - female). We have done the initial processing and is now in the testing phase to polish the rough edges before putting them up for download.

Manuel Rodriguez Hourcadette

I am one of a small number of professors, including Elsa Mora, of the Universidad de Los Andes in Merida, Venezuela which consists of the only group in Venezuela which works in Speech Technologies. Via a joint project, with the help of colleagues at the Laboratoire de Parole et Langage at the Université de Provence, in particular Christian Cavé, we developed the first Venezuelan Spanish diphone data base, vz1, in 2003. In 2004, I adapted a program written by Alistair Conkie to develop an automatic TTS synthesizer for our database. It has been very well received by local blind and mute people.

Marc Schroeder

I have been part of the team of DFKI and Saarland University that prepared the German voice databases de6 and de7 -- each of which actually consisting of three full diphone sets with different voice qualities related to vocal effort. MBROLA is still the best diphone synthesis algorithm that is publicly available! :-)

Markus Binsteiner

I created the bavarian voice (de8) as part of my thesis. The goal was to build a TTS-System that translates german text to a voice which speaks a certain (south-east) bavarian dialect.

Meelis Mihkla

Estonian text-to-speech synthesis was the common project of the Institute of Estonian Language, the Institute of Cybernetics and Filosoft Ltd. We joined the Mbrola-project in 1998. I took part in compiling Estonian diphone database and also in developing our TTS system. Now I'm dealing with statistical modelling of speech prosody for TTS synthesis.

Nicolas D'Alessandro

Together with Rapahel Sebbe, I created a MBROLA pluggin for MAX/MSP (realtime audio/video processing environment), which was first used in Art Zoyd's ARMAGEDDON opera for robots in 2004, and is now distributed freely and used by various artists. I am now the official MBROLA officer. I am also a PhD student at TCTS Lab, working on realtime audio synthesis.

Olivier Bianchi

Olivier Bianchi provided a diphone database for classical Latin (LA1), consisting of 1275 diphones, male voice. This database was created in the context of the author's doctoral project. LA1 provide diphone for classical Latin pronunciation only (i.e. 1st century BC).

Olivier Platteau

End of 1995, Thierry Dutoit proposed me to collaborate to the MBROLA project. I programmed a phoneme editor in Visual Basic, MbrEdit, based on the MBROLA synthetizer (MBREdit is the VB equivalent to MBROLI). I also later worked on the "W project" (a prototype abbvreviation-bsed fst keyboard fot speech synthesis) as a tester. I also took part in the development of a Euler interface with Visual Basic (eLite)

Piet Mertens

I mainly used MBROLA and diphones databases for french in order to realize a TTS synthetizer called Mingus (which uses also the phonetizer from Thierry and Vincent). Recently, I did a preliminary work for the realisation of a diphones databases for Vietnamese (with Chau Pham).

Raphael Sebbe

I contributed to the MBROLA porting for Mac OS X and to the realisation of a phoneme player (Mbrolix). I also created with Nicolas D'Alessandro an Maxbrola plugin Max/MSP for the Art Zoyd show, using keyboards for synthesis.

Thierry Dutoit

I imagined the MBROLA algorithm one evening in my morning shower (an excellent place for ideas!). It later became the center of my PhD in 2003. Together with Vincent Pagel, we later produced the MBROLA software as it stands now, and built up the web site.

Vincent Pagel

I initiated the project with Thierry Dutoit in December 1995, then joined TCTS lab in 1996, where (among other things!) I served as "MBROLA officer". I am also the author of the MBRDICO Project for multilingual corpus-based phonetization.

If you are also a MBROLA database provider, compiler, developper or user, and if you are not yet on this web page, you may send us your photo and a small piece of text with some details about your work with MBROLA to the adress matthieu.jottrand@tcts.fpms.ac.be

To the top