Hello,
During the migration from 1.8.3 to 2.0 I had some troubles (solved) with voiceRSS, in the meantime I tried to add new languages to MaryTTS without success.
I tried to add the jar of the new languages to /var/lib/openhab2/cache/org.eclipse.osgi/226/0/.cp/lib/ where the other maryTTS jar are located but the new jar didn’t appear in the list of
paperUI (even after a restart).
Do you have some ideas?
Hallo,
hm, this is exactly what I am looking for. Have you been ablet to solve this? THX!
Stefan
Nop i stayed with voicerss
Hi
I see that M.A.R.Y. has a 5.2 version on their website, http://mary.dfki.de/download/index.html
Which seems to offer better quality speech, with inflections (to a degree). (when using their Web Demo)
Does anyone know if there are any plans on updating OH MARY from 5.1 to 5.2 and include the new voice options?
I assume it is NOT just a case of renaming the JAR files and dropping them in 
FYI..
The openHAB documentation suggests that a machine with reasonable resources is required to accommodate MARY, (there’s a joke in there somewhere…)
I’ve got 5.1 running on an Odroid C2 and Odroid UX4 (within OH) with good results.
I’ve just downloaded 5.2 and have it running as a webserver on my Odroid UX4 and it’s working REALLY well.
(Other than I can’t figure out how to enable more than the single language)
The only thing that I have NO idea how to do, is save the wave file from this URL into something that I can send to audio sinks..
http://{IP_ADDRESS}:59125/process?INPUT_TYPE=TEXT&OUTPUT_TYPE=AUDIO&INPUT_TEXT=Welcome%20to%20the%20world%20of%20speech%20synthesis!%0A&OUTPUT_TEXT=&effect_Volume_selected=&effect_Volume_parameters=amount%3A2.0%3B&effect_Volume_default=Default&effect_Volume_help=Help&effect_TractScaler_selected=&effect_TractScaler_parameters=amount%3A1.5%3B&effect_TractScaler_default=Default&effect_TractScaler_help=Help&effect_F0Scale_selected=&effect_F0Scale_parameters=f0Scale%3A2.0%3B&effect_F0Scale_default=Default&effect_F0Scale_help=Help&effect_F0Add_selected=&effect_F0Add_parameters=f0Add%3A50.0%3B&effect_F0Add_default=Default&effect_F0Add_help=Help&effect_Rate_selected=&effect_Rate_parameters=durScale%3A1.5%3B&effect_Rate_default=Default&effect_Rate_help=Help&effect_Robot_selected=&effect_Robot_parameters=amount%3A100.0%3B&effect_Robot_default=Default&effect_Robot_help=Help&effect_Whisper_selected=&effect_Whisper_parameters=amount%3A100.0%3B&effect_Whisper_default=Default&effect_Whisper_help=Help&effect_Stadium_selected=&effect_Stadium_parameters=amount%3A100.0&effect_Stadium_default=Default&effect_Stadium_help=Help&effect_Chorus_selected=&effect_Chorus_parameters=delay1%3A466%3Bamp1%3A0.54%3Bdelay2%3A600%3Bamp2%3A-0.10%3Bdelay3%3A250%3Bamp3%3A0.30&effect_Chorus_default=Default&effect_Chorus_help=Help&effect_FIRFilter_selected=&effect_FIRFilter_parameters=type%3A3%3Bfc1%3A500.0%3Bfc2%3A2000.0&effect_FIRFilter_default=Default&effect_FIRFilter_help=Help&effect_JetPilot_selected=&effect_JetPilot_parameters=&effect_JetPilot_default=Default&effect_JetPilot_help=Help&HELP_TEXT=&exampleTexts=&VOICE_SELECTIONS=cmu-slt-hsmm%20en_US%20female%20hmm&AUDIO_OUT=WAVE_FILE&LOCALE=en_US&VOICE=cmu-slt-hsmm&AUDIO=WAVE_FILE
Replacing the TTS Text in the URL is easy enough.. Now that I have had some OH rules training 
@Kai - Would you happen to know who maintains the MARY functionality?
Thanks in advance.
Last time I checked, adding more voices adds 100s of megabytes to the openHAB add-ons downloads so I don’t think it will ever happen this way. Maybe someone can create a bundle fragment that adds these voices. But Mary TTS is very OSGi unfriendly because there are no bundles and the same packages are split across several JARs. 
There’s also an issue for this:
Thanks for the advice