diff options
author | Niko Tyni <ntyni@iki.fi> | 2005-10-18 14:32:20 +0200 |
---|---|---|
committer | Niko Tyni <ntyni@iki.fi> | 2005-10-18 14:32:20 +0200 |
commit | 8e03a58d62f4b8c00978634f116819ff246c09d3 (patch) | |
tree | 29284d785fe91b3473a19dcc257dfaa6066f4b12 /lib/voices/finnish/hy_fi_mv_diphone/festvox/suopuhe.dtd |
Import festvox-suopuhe-mv_20041119.orig.tar.gz
[dgit import orig festvox-suopuhe-mv_20041119.orig.tar.gz]
Diffstat (limited to 'lib/voices/finnish/hy_fi_mv_diphone/festvox/suopuhe.dtd')
-rw-r--r-- | lib/voices/finnish/hy_fi_mv_diphone/festvox/suopuhe.dtd | 153 |
1 files changed, 153 insertions, 0 deletions
diff --git a/lib/voices/finnish/hy_fi_mv_diphone/festvox/suopuhe.dtd b/lib/voices/finnish/hy_fi_mv_diphone/festvox/suopuhe.dtd new file mode 100644 index 0000000..8eb2d62 --- /dev/null +++ b/lib/voices/finnish/hy_fi_mv_diphone/festvox/suopuhe.dtd @@ -0,0 +1,153 @@ +<!-- + + SuoPuhe-project speech synthesizer input XML + + Copyright (c) 2000-2003 Dept. General Linguistics, University of Helsinki + All Rights Reserved. + + Author: Nicholas Volk (nvolk@ling.helsinki.fi) + + This program is distributed under Gnu Lesser General Public License (cf. the + file LICENSE in distribution). + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU Lesser General Public License as published by + the Free Software Foundation; either version 2 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Lesser General Public License for more details. + + TODO: different kinds of accents... + break-types need to be reconsidered + audio could read stuff from www + + v.0.6.1 -audio tag + + v.0.6.0 -changed "sentence" to "utterance" + -initial doubling is done with sayas, not with + "triggers_doubling" + -token can contain breaks and phrase boundaries + -added language + + v0.5.1 (nv) minor bug fix + modification on breaks + v0.5.0 replaced sayas with original + the motivation was that I needed include tags within the tokens: + + <token original="1351"> yksi kolme <phrase/> viisi yksi </token> + + + v0.4.5 (Nicholas Volk 18.6.2002) + -changed accent from accent ( yes | no ) "no" + to accent ( yes | neutral | no ) "neutral" > + + + v. 0.4.1 added list-argument to <TOKEN> + -sanaluokkatiedot kesken ja muuttuvat varmasti + -puhuvalle päälle voi välittää ilme-tiedon ennen sanaa ja sanan jälkeen, + ei sanan aikana + -mitenköhän kielen generointi + - +--> + +<!ELEMENT suopuhe ((speaker|audio)+)> + + + +<!ELEMENT speaker ((utterance, pause?)+)> +<!ATTLIST speaker + name CDATA #IMPLIED + gender (male|female) #IMPLIED + age (child|youth|adult|aged) #IMPLIED > + +<!-- speaker --> +<!-- if the name refers to an existing voice, try to use that one, --> +<!-- instead of the default voice. Otherwise the voice selection --> +<!-- is primarily based on the sex and secondarily on the age. --> +<!-- Without only 2 Finnish, you don't get much support for these. --> +<!-- Jos nimi viittaa johonkin käytettävissä järjestelmässä olevaan --> +<!-- ääneen, käytetään sitä, muuten oletusääntä. --> +<!-- Muuten ääni valitaan ensisijaisesti sukupuolen ja toissijaisesti --> +<!-- iän perusteella. Sitten kun toinen ääni on valmis, --> +<!-- niin tällä tagilla on oikeasti väliä. --> + +<!ELEMENT pause EMPTY> +<!ATTLIST pause + duration CDATA #IMPLIED > + +<!-- pause --> +<!-- A long pause between utterances, not a phrase break --> +<!-- Aito pitkän hiljainen hetki, ei hapentankkauskatko keskellä --> +<!-- lausetta --> + +<!ELEMENT audio EMPTY> +<!ATTLIST audio + file CDATA #IMPLIED> + +<!ELEMENT utterance ((facial?, token, (break|phrase)?)*, token, facial?)> +<!ATTLIST utterance + language CDATA "FI" + type (declaration|question|exclamation) "declaration" + manner (normal|whisper|creaky|fast) "normal" + tone (neutral|sarcastic) "neutral"> + +<!-- utterance --> +<!-- "type" sets the baseline/intonation type --> +<!-- "manner" and "tone" are hard to support in diphone-based synthesis --> +<!-- so they are not really supported yet. --> +<!-- "type" määrää lauseen baselinen (fujisakin fraasikomponentin jne.), --> +<!-- "manner" ja "tone" ovat --> +<!-- vaikeasti tuettavissa difonipohjaisessa synteesissä, mutta ehkä --> +<!-- ne joskus tekevät jotain... --> + + +<!ELEMENT facial EMPTY> +<!ATTLIST facial + expression CDATA "plain"> + + +<!-- facial --> +<!-- simplistic talking head support --> +<!-- expression tells what expression to use at the end of that word --> +<!-- expression kertoo millaiseen yleisilmeiseen pitäisi päätyä sanan --> +<!-- loputtua. Alkeellinen Puhuva pää -tuki. --> + +<!ELEMENT token (#PCDATA|break|phrase)*> +<!ATTLIST token + pos ( abbr | adjective| adverb | char | conj | foreign | + hybrid | multiword | name | noun | num | punc | + unknown | verb ) "unknown" + original CDATA #IMPLIED + list ( yes | no ) "no" + accent ( yes | neutral | no ) "neutral" > + + +<!-- token --> +<!-- "token" means any string. Punctuation marks have been separated --> +<!-- from the words they were attached to and now appear as individual --> +<!-- tokens --> +<!-- Lisäksi ilmoitetaan onko sana painotettu vai ei lausepainon --> +<!-- kannalta. --> + + + <!ELEMENT phrase EMPTY> + +<!-- marks a phrase break thus supporting fujisaki model --> +<!-- uuden fraasin alku, signaloi fujisakin mallille --> +<!-- phrase-komentoa, voitaneen käyttää myös potentiaalisen --> +<!-- tauon ilmaisemiseen... --> + + + <!ELEMENT break EMPTY> + <!ATTLIST break level ( -1 | 0 | 1 | 2 | 3 | 4 ) "1" > + + +<!-- -1 no break --> +<!-- 0 no break, phrase boundary; we have a separate phrase though --> +<!-- 1+ breaks with phrase boundaries --> + + + + |