[mary-dev] Support for building vocalization on marytts 5

[Fabio Tesser]

Hello,

I would like to know the current state of the vocalization support in 
marytts5. I know that  there is support for running old voices that had 
been made with the vocalization (i.e. dfki-poppy-hsmm), and I see that 
there are modules able to run voiceimport modules that produces the 
vocalizations files (i.e. 
https://github.com/marytts/marytts/blob/master/marytts-builder/src/main/java/marytts/tools/voiceimport/vocalizations/VocalizationUnitfileWriter.java).

However I have not found anything about including vocalization on 
VoiceCompiler or HMMVoiceCompiler. So I think that the building of a 
voice with vocalizations it is not supported  with mary 5 in this 
moment. Am I right?

On the other way I see that the old VoicePackager 
https://mary.opendfki.de/repos/branches/4.x-branch/java/marytts/tools/voiceimport/VoicePackager.java 
was able to build a voice including vocalizations.

There are any plans of migrating this functionality on mary 5?

If not, I could help on that.
In this case could you tell me, if apart from:

S. Pammi, M. Schröder, M. Charfuelan, O. Türk, and I. Steiner, 
"Synthesis of listener vocalisations with imposed intonation contours," 
in ssw7, 2010

there are  any other manual/readme/wiki that explains what are the voice 
import modules to run in order to built a voice that includes the 
vocalizations?


Thanks & Best,
Fabio.