Festival 2.4 seems to break compatibility with previous festival 2.1 HTS voices.
I will update the Catalan voices as soon as possible (ideally this week) to be compatible with festival 2.4 and I will check if there are other HTS voices in Debian. Thanks for the report. El 20/05/2015, a les 12:32, Ximo Baldó i Soriano <[email protected]> va escriure: > Package: festvox-ca-ona-hts > Version: 1.2-1 > Severity: important > > --- Please enter the report below this line. --- > > Since last festival package upgrade on May 11th, festvox-ca-ona-hts catalan > voice for festival no longer works. > > I'm using speech-dispatcher with festival as output module wit jovie > (formerly > ktts. kde text speech system). When I try to add a "talker", during > configuration, all languages appear correctly recognized but catalan. Whith > catalan voice, Catalan name does no appear on the list, it shows as "none", > making it incorrectly configurable. > > > Here you can see, when starting Jovie, talker "Kanija" shows lang="es" > correctly, and talker "Kaya" shows lang="none" instead lang="ca" (which I > think is the correct form): > > ------- > jovie(32650)/kttsd (daemon) SpeakerPrivate::readTalkerData: > TalkerListWidget::loadTalkerCodes: talkerID = "Kanija" > jovie(32650)/kttsd (daemon) SpeakerPrivate::readTalkerData: > TalkerCodeWidget::loadTalkerCodes: talkerCode = "<voice name="Kanija" > lang="es" outputModule="festival" voiceName="" voiceType="6" ><prosody > volume="20" rate="10" pitch="5" punctuation="1"/></voice>" > jovie(32650)/kttsd (daemon) SpeakerPrivate::readTalkerData: > TalkerListWidget::loadTalkerCodes: talkerID = "Kaya" > jovie(32650)/kttsd (daemon) SpeakerPrivate::readTalkerData: > TalkerCodeWidget::loadTalkerCodes: talkerCode = "<voice name="Kaya" > lang="none" outputModule="festival" voiceName="" voiceType="6" ><prosody > volume="20" rate="10" pitch="5" punctuation="1"/></voice>" > ------- > > I don't know what info do you need, but when trying to selecto voice in > festival and test it I get an aplay error: > > festival> (voice.list) > (upc_ca_ona_hts el_diphone kal_diphone) > festival> (voice_el_diphone) > el_diphone > festival> (SayText "testing festival voices") > #<Utterance 0x7f3dc116f8b0> > festival> (voice_kal_diphone) > kal_diphone > festival> (SayText "testing festival voices") > #<Utterance 0x7f3dc1213270> > festival> (voice_upc_ca_ona_hts) > upc_ca_ona_hts > festival> (SayText "testing festival voices") > > Warning: HTS_fopen: Cannot open hts/htsvoice. > aplay: main:593: bad speed value 0 > #<Utterance 0x7f3dc1984430> > festival> > > First two voices works correctly (spanish and english), but catalan voice > gives an error. > > > As far I know, my config files are corrects, configured with festival and > catalan > as default module and voice: > > > ~/.speech-dispatcher/speechd.conf: > # Global configuration for Speech Dispatcher > # ========================================== > > # -----SYSTEM OPTIONS----- > > # CommunicationMethod specifies the method to be used by Speech Dispatcher to > communicate with > # its clients. Two basic methods are "unix_socket" and "inet_socket". > # > # unix_socket -- communication over Unix sockets represented by a file in the > # filesystem (see SocketPath below). This method works only locally, but is > # prefered for standard session setup, where every user runs his own instance > of Speech > # Dispatcher to get voice feedback on his own computer. > # > # inet_socket -- alternatively, you can start Speech Dispatcher on > # a TCP port and connect to it via hostname/port. This allows for a more > # flexible setup, where you can use Speech Dispatcher over network > # from different machines. See also the Port and LocalhostAccessOnly > # configuration variables. > # > # CommunicationMethod "unix_socket" > > # SocketPath is either "default" or a full path to the filesystem > # where the driving Unix socket file should be created in case the > # CommunicationMethod is set to "unix_socket". The default is > # $XDG_RUNTIME_DIR/speech-dispatcher/speechd.sock where $XDG_RUNTIME_DIR > # is the directory specified by the XDG Base Directory Specification. > # Do not change this unless you have a reason and know what you are doing. > > # SocketPath "default" > > # The Port on which Speech Dispatcher should be available to clients if the > "inet_socket" > # communication method is used. > > # Port 6560 > > # By default, if "inet_socket" communication method is used, the specified > port is opened only > # for connections comming from localhost. If LocalhostAccessOnly is set to 0 > it disables this > # access controll. It means that the port will be accessible from all > computers on the > # network. If you turn off this option, please make sure you set up some > system rules on what > # computers are and are not allowed to access the Speech Dispatcher port. > > # LocalhostAccessOnly 1 > > # -----LOGGING CONFIGURATION----- > > # The LogLevel is a number between 0 and 5 that specifies > # how much of the logging information should be printed > # out on the screen or in the logfile (see LogFile) > # 0 means nothing, 5 means everything (not recommended). > > LogLevel 3 > > # The LogDir specifies where Speech Dispatcher writes its logging messages > # (status information, error messages, etc.). Specify "stdout" for > # standard console output or a directory path. 'default' means that > # the logs are written to the default destination (e.g. a preconfigured > # system directory or the home directory if .speech-dispatcher is present) > # DO NOT COMMENT OUT THIS OPTION, SET IT TO "default" if you do not > # want to influence it. > > LogDir "default" > #LogDir "/var/log/speech-dispatcher/" > #LogDir "stdout" > > # The CustomLogFile allows logging all messages of the given kind, > # regardless their priority, to the given destination. > > #CustomLogFile "protocol" "/var/log/speech-dispatcher/speech-dispatcher- > protocol.log" > > # ----- VOICE PARAMETERS ----- > > # The DefaultRate controls how fast the synthesizer is going to speak. > # The value must be between -100 (slowest) and +100 (fastest), default > # is 0. > > DefaultRate 10 > > # The DefaultPitch controls the pitch of the synthesized voice. The > # value must be between -100 (lowest) and +100 (highest), default is > # 0. > > DefaultPitch 5 > > # The DefaultVolume constrols the default volume of the voice. It is > # a value between -100 (softly) and +100 (loudly). Currently, +100 > # maps to the default volume of the synthesizer. > > DefaultVolume 100 > > # The DefaultVoiceType controls which voice type should be used by > # default. Voice types are symbolic names which map to particular > # voices provided by the synthesizer according to the output module > # configuratuion. Please see the synthesizer-specific configuration > # in etc/speech-dispatcher/modules/ to see which voices are assigned to > # different symbolic names. The following symbolic names are > # currently supported: MALE1, MALE2, MALE3, FEMALE1, FEMALE2, FEMALE3, > # CHILD_MALE, CHILD_FEMALE > > DefaultVoiceType "FEMALE3" > > # The Default language with which to speak > > DefaultLanguage ca > > > # ----- MESSAGE DISPATCHING CONTROLL ----- > > # The DefaultClientName specifies the name of a client who didn't > # introduce himself at the beginning of an SSIP session. > > # DefaultClientName "unknown:unknown:unknown" > > # The Default Priority. Use with caution, normally this shouldn't be > # changed globally (at this place) > > # DefaultPriority "text" > > # The DefaultPauseContext specifies by how many index marks a speech > # cursor should return when resuming after a pause. This is roughly > # equivalent to the number of sentences before the place of the > # execution of pause that will be repeated. > > # DefaultPauseContext 0 > > # -----SPELLING/PUNCTUATION/CAPITAL LETTERS CONFIGURATION----- > > # The DefaultPunctuationMode sets the way dots, comas, exclamation > # marks, question marks etc. are interpreted. none: they are ignored > # some: some of them are sent to synthesis (see > # DefaultPunctuationSome) all: all punctuation marks are sent to > # synthesis > > # DefaultPunctuationMode "none" > > # The DefaultCapLetRecognition: if set to "spell", capital letters > # should be spelled (e.g. "capital b"), if set to "icon", > # capital letters are indicated by inserting a special sound > # before them but they should be read normally, it set to "none" > # capital letters are not recognized (by default) > > # DefaultCapLetRecognition "none" > > # The DefaultSpelling: if set to On, all messages will be spelled > # unless set otherwise (this is usually not something you want to do.) > > # DefaultSpelling Off > > # ----- AUDIO CONFIGURATION ----------- > > # -- AUDIO OUTPUT -- > > # Chooses between the possible sound output systems: > # "pulse" - PulseAudio > # "alsa" - Advanced Linux Sound System > # "oss" - Open Sound System > # "nas" - Network Audio System > # "libao" - A cross platform audio library > # Pulse audio is the default and recommended sound server. OSS and ALSA > # are only provided for compatibility with architectures that do not > # include Pulse Audio. NAS provides network transparency, but is not > # very well tested. libao is a cross platform library with plugins for > # different sound systems and provides alternative output for Pulse Audio > # and ALSA as well as for other backends. > > AudioOutputMethod pulse > > # -- Pulse Audio parameters -- > > # Pulse audio server name or "default" for the default pulse server > > #AudioPulseServer "default" > > #AudioPulseMinLength 1764 > > # -- ALSA parameters -- > > # Audio device for ALSA output > > #AudioALSADevice "default" > > # -- OSS parameters -- > > # Audio device for OSS output > > #AudioOSSDevice "/dev/dsp" > > # -- NAS parameters -- > > # Route to the Network Audio System server when NAS > # was chosen for the audio output. Note that NAS > # server doesn't need to run on your machine, > # you can use it also over network (for instance > # when working on remote machines). > > #AudioNASServer "tcp/localhost:5450" > > > > # -----OUTPUT MODULES CONFIGURATION----- > > # Each AddModule line loads an output module. > # Syntax: AddModule "name" "binary" "configuration" "logfile" > # - name is the name under which you can acces this module > # - binary is the path to the binary executable of this module, > # either relative (to lib/speech-dispatcher-modules/) or absolute > # - configuration is the path to the config file of this module, > # either relative (to etc/speech-dispatcher/modules/) or absolute > > #AddModule "espeak" "sd_espeak" "espeak.conf" > AddModule "festival" "sd_festival" "festival.conf" > #AddModule "flite" "sd_flite" "flite.conf" > #AddModule "ivona" "sd_ivona" "ivona.conf" > #AddModule "pico" "sd_pico" "pico.conf" > #AddModule "espeak-generic" "sd_generic" "espeak-generic.conf" > #AddModule "espeak-mbrola-generic" "sd_generic" "espeak-mbrola-generic.conf" > #AddModule "swift-generic" "sd_generic" "swift-generic.conf" > #AddModule "epos-generic" "sd_generic" "epos-generic.conf" > #AddModule "dtk-generic" "sd_generic" "dtk-generic.conf" > #AddModule "ibmtts" "sd_ibmtts" "ibmtts.conf" > #AddModule "cicero" "sd_cicero" "cicero.conf" > #AddModule "pico-generic" "sd_generic" "pico-generic.conf" > > # DO NOT REMOVE the following line unless you have > # a specific reason -- this is the fallback output module > # that is only used when no other modules are in use > #AddModule "dummy" "sd_dummy" "" > > # The output module testing doesn't actually connect to anything. It > # outputs the requested commands to standard output and reads > # responses from stdandard input. This way, Speech Dispatcher's > # communication with output modules can be tested easily. > > # AddModule "testing" > > # The DefaultModule selects which output module is the default. You > # must use one of the names of the modules loaded with AddModule. > > DefaultModule festival > > # The LanguageDefaultModule selects which output modules are prefered > # for specified languages. > > #LanguageDefaultModule "en" "espeak" > #LanguageDefaultModule "cs" "festival" > #LanguageDefaultModule "es" "festival" > LanguageDefaultModule "ca" "festival" > > # -----CLIENT SPECIFIC CONFIGURATION----- > > # Here you can include the files with client-specific configuration > # for different types of clients. They must contain one or more sections with > # this structure: > # BeginClient "emacs:*" > # DefaultPunctuationMode "some" > # ...and/or some other settings > # EndClient > # The parameter of BeginClient tells Speech Dispatcher to which clients > # it should apply this settings (it does glob-style matching, you can use > # * to match any number of characters and ? to match one character) > > # There are some sample client settings > > Include "clients/*.conf" > > # The DisableAutoSpawn option will disable the autospawn mechanism. > # Thus the server will not start automatically on requests from the clients > # DisableAutoSpawn > > > /etc/speech-dispatcher/modules/festival.conf and ~/.speech- > dispatcher/modules/festival.conf are identical, and have all lines on they > commented but Debug parameter: > > ~$ cat /etc/speech-dispatcher/modules/festival.conf > > # -- FESTIVAL SERVER SETTINGS -- > > # Address where the Festival server runs (you have to > # have a Festival server running, please see documentation). > > #FestivalServerHost "localhost" > #FestivalServerPort 1314 > > > # -- CACHING -- > > # Festival output module supports a caching mechanism for > # repeated events like characters, keys and sound icons > # that saves the wavefiles retrieved from the synthesizer > # and uses them next time an identical request is issued. > # This way, the synthesis is much more responsive when > # you type in keys, you spell some word by going over it > # with cursor keys, sound icons are faster and so on. > # If you wish to enable this caching mechanism, turn > # FestivalCacheOn to 1, or switch it off by 0. > > # FestivalCacheOn 1 > > # How large should the memmory assigned to output module for > # cache should be. Festival will never overcome this limit. > # If there are more messages to save, the ones that are least > # accessed will be removed from the cache. So if you set this > # parameter too low, the module will often have to synthesize > # your request (which takes time) instead of just retrieving > # it from the cache, but you will save system resources. If you > # set it higher, it might make the module more responsive. > # The value is given in kilobytes, default is 10 megabytes. > > # FestivalCacheMaxKBytes 10240 > > # Normally, the cache pool is common for all voices, rates and > # pitches. This means that when for example the key 'r' gets cached, > # you change the voice and request the key 'r' again, it will be > # retrieved from the cache and you will hear the old 'r' instead of a > # new one. By default, it only switches between languages. In most > # cases, this isn't a problem for characters, keys and sound icons and > # it saves system resources. But if you want to keep separate caches > # for different voices (and even rates and pitches), please set the > # corresponding parameters to 1. (0 means don't distinguish). > > # FestivalCacheDistinguishVoices 0 > # FestivalCacheDistinguishRate 0 > # FestivalCacheDistinguishPitch 0 > > # -- FESTIVAL PERFORMANCE -- > > # Switching FestivalReopenSocket to 1 will make the module close the > # connection to Festival each time the currently synthesized wavefile > # is no longer needed to finish and new text is waiting. This may improve > # Festival responsivity on slower machines, but might cause a more network > # trafic. Currently, the option is set to 0 by default, because there > # is a network problem in Festival socket communication layer that introduces > # unnecessary delays that affect the performance heavily in this mode. > # Unless your Festival is patched against this bug, switching this on > # is not recommended. > > # FestivalReopenSocket 0 > > > # -- DEBUGING -- > > # Debug turns debugging on or off > # See speechd.conf for information where debugging information is stored > Debug 0 > > # If FestivalDebugSaveOutput is set to 1, it writes the produced sound tracks > # to /tmp/debug-festival-*.snd before it says them. You can later browse them > # for debugging purposes. They are numbered in order starting from 0. > > > > If you need any information I can collect or any test I can do, tell me and > I'll do it without any problem. > > > --- System information. --- > Architecture: amd64 > Kernel: Linux 4.0.0-1-amd64 > > Debian Release: stretch/sid > 950 unstable ftp.deb-multimedia.org > 900 unstable ftp.debian.org > 800 experimental ftp.debian.org > 500 stable people.debian.org > > --- Package information. --- > Depends (Version) | Installed > ==========================-+-=========== > festival-ca (>= 2.1.5) | 3.0.6-1 > > > Package's Recommends field is empty. > > Package's Suggests field is empty. -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
