[mary-dev] Differences between svn and standalone-installation

Marc Schroeder marc.schroeder at dfki.de
Fri Oct 15 16:46:18 CEST 2010


Hi Fabio,

the DummyAllophones2AcoustParams steps in when we don't have 
voice-specific acoustic models. Could you verify if there is any 
difference between the config files in MARYBASE/conf, when you compare 
those of the released version with those of the svn version?

Specifically, I can reproduce the problem by commenting out the 
following lines from en_GB-dfki-obadiah-hsmm.config:

#voice.dfki-obadiah-hsmm.acousticModels = duration F0


Best,
Marc

On 15.10.10 14:20, Fabio Tesser wrote:
> Hi Ingmar,
>
> thanks for your quick reply! In attachment the log file.
> I actualy see from the logs that DummyAllophones2AcoustParams is used...
> is this the default behaviour in svn?
>
> Best,
> Fabio.
>
>
>
> On 10/15/2010 01:20 PM, Ingmar Steiner wrote:
>> Hi Fabio,
>>
>> Thanks for the debug logs! Both of these errors seem to be caused by
>> the deprecated DummyAllophones2AcoustParams module being called (which
>> doesn't predict the required duration and F0 info). This should not
>> happen. Could you please provide the server.log of your maryserver (at
>> startup, before any synthesis requests)?
>>
>> Best wishes,
>>
>> /**
>> * Ingmar Steiner
>> * Researcher, Language Technology
>> * German Research Center for Artificial Intelligence
>> *
>> * Campus D3 1 +1.18
>> * D-66123 Saarbrücken
>> * Germany
>> * Phone: ++49-681-857-75-5263 (NEW!)
>> * Email: ingmar.steiner at dfki.de
>> *
>> * Deutsches Forschungszentrum für Künstliche Intelligenz GmbH
>> * Trippstadter Straße 122, D-67663 Kaiserslautern, Germany
>> * Geschäftsführung:
>> * Prof. Dr. Dr. h.c. mult. Wolfgang Wahlster (Vorsitzender)
>> * Dr. Walter Olthoff
>> * Vorsitzender des Aufsichtsrats:
>> * Prof. Dr. h.c. Hans A. Aukes
>> * Amtsgericht Kaiserslautern, HRB 2313
>> */
>>
>> On 15 Oct 2010, at 13:09, Fabio Tesser wrote:
>>
>>> Hello everyone,
>>> I just started the work of adding the Italian language in OpenMary.
>>>
>>> Marc has already created a branch for that, but as first step, in
>>> orrder to better understand the sources, I have tried to check out
>>> the latest tag from svn<https://mary.opendfki.de/repos/tags/4.1.1>.
>>>
>>> I have installed two english voices (mary-dfki-obadiah and
>>> mary-dfki-obadiah-hsmm) using the mary-component-installer.
>>> After starting the maryserver and making a synth request with a
>>> web-client, the audio is not produced and I receive the errors
>>> reported in the attached logs.
>>>
>>> In the UnitSelection case seem that some data is missed:
>>> "Caused by: java.lang.IllegalStateException: No duration information
>>> in MaryXML -- check log file for messages warning about unloadable
>>> acoustic models"
>>> but I checked and seem that in lib/voices/dfki-obadiah there are the
>>> relative data files.
>>>
>>> The HMM voice error seem more general:
>>> "Caused by: marytts.exceptions.SynthesisException: HMM Synthesiser
>>> could not synthesise:"
>>>
>>>
>>> My system is composed by:
>>> * Linux Ubuntu 10.04 x86_64
>>> * java -version
>>> java version "1.6.0_18"
>>> OpenJDK Runtime Environment (IcedTea6 1.8.1) (6b18-1.8.1-0ubuntu1)
>>> OpenJDK 64-Bit Server VM (build 16.0-b13, mixed mode)
>>>
>>> I tried to install the OpenMary 4.1.1 using also the standalone
>>> procedure (java -jar openmary-standalone-install-4.1.1.jar) and after
>>> installing the same voices, this maryserver synthesizes the sentences
>>> without problems in the same computer.
>>>
>>> Does anyone have a suggestion of why I get the errors using the
>>> maryserver built from the svn tag?
>>>
>>> Thanks,
>>> Fabio.
>>>
>>>
>>> <mary-dfki-obadiah-hsmm_request.log><mary-dfki-obadiah_request.log>_______________________________________________
>>>
>>> Mary-dev mailing list
>>> Mary-dev at dfki.de
>>> http://www.dfki.de/mailman/cgi-bin/listinfo/mary-dev
>>
>
>
> _______________________________________________
> Mary-dev mailing list
> Mary-dev at dfki.de
> http://www.dfki.de/mailman/cgi-bin/listinfo/mary-dev

-- 
Dr. Marc Schröder, Senior Researcher at DFKI GmbH
Coordinator EU FP7 Project SEMAINE http://www.semaine-project.eu
Project leader for DFKI in SSPNet http://sspnet.eu
Project leader PAVOQUE http://mary.dfki.de/pavoque
Associate Editor IEEE Trans. Affective Computing http://computer.org/tac
Editor W3C EmotionML Working Draft http://www.w3.org/TR/emotionml/
Portal Editor http://emotion-research.net
Team Leader DFKI TTS Group http://mary.dfki.de

Homepage: http://www.dfki.de/~schroed
Email: marc.schroeder at dfki.de
Phone: +49-681-85775-5303
Postal address: DFKI GmbH, Campus D3_2, Stuhlsatzenhausweg 3, D-66123 
Saarbrücken, Germany
--
Official DFKI coordinates:
Deutsches Forschungszentrum fuer Kuenstliche Intelligenz GmbH
Trippstadter Strasse 122, D-67663 Kaiserslautern, Germany
Geschaeftsfuehrung:
Prof. Dr. Dr. h.c. mult. Wolfgang Wahlster (Vorsitzender)
Dr. Walter Olthoff
Vorsitzender des Aufsichtsrats: Prof. Dr. h.c. Hans A. Aukes
Amtsgericht Kaiserslautern, HRB 2313


More information about the Mary-dev mailing list