espeakEVENT_PHONEME, fix audio position of vowel when it follows a voiced consonant. Ensure enough space in the event list when espeak_Initialize() is called with a small buffer size. Tone languages: set the final stressed syllable to stress=7 so that it can be longer. lang=zh: adjust vowels git-svn-id: https://espeak.svn.sourceforge.net/svnroot/espeak/trunk@136 d46cf337-b52f-0410-862d-fd96e6ae7743master
a4o Au51 | a4o Au51 | ||||
a (DnK a | a (DnK a | ||||
a (DngK A | |||||
ch) a (DngK A | |||||
ch) a (4ngK a | |||||
ch) a (ng4K a | |||||
k) a (DngK A | |||||
_) a (DngK A | |||||
y) a iA | y) a iA | ||||
y) ao Au | y) ao Au | ||||
.group o | .group o | ||||
o o | o o | ||||
o (DngK u | |||||
ou ou | ou ou | ||||
o1u ou55 | o1u ou55 | ||||
o2u ou35 | o2u ou35 |
[N] zh | [N] zh | ||||
nn/nn_ [N] base | nn/nn_ [N] base | ||||
[N] en | [N] en | ||||
[N] zh | |||||
nn/nn@ [(@)] base | nn/nn@ [(@)] base | ||||
nn/nn#_ [N#] is | nn/nn#_ [N#] is | ||||
nn/nn2_ [N] zh | |||||
nn/nna [(a)] base | nn/nna [(a)] base | ||||
nn/nne [(e)] base | nn/nne [(e)] base | ||||
nn/nni [(i)] base | nn/nni [(i)] base | ||||
[ie] fi | [ie] fi | ||||
[ia] th | [ia] th | ||||
[i:a] th | [i:a] th | ||||
vdiph2/ie_2 [iE] zh | |||||
vdiph2/ii@ [i@] en | vdiph2/ii@ [i@] en | ||||
[i@3] en | [i@3] en | ||||
vdiph2/ii@_3 [i@] vi | vdiph2/ii@_3 [i@] vi | ||||
[i] hr | [i] hr | ||||
[i] id | [i] id | ||||
[i] tr | [i] tr | ||||
vowel/i#_6 [i.] zh | |||||
vowel/i#_7 [i[] zh | vowel/i#_7 [i[] zh | ||||
vowel/i_en [i:] en | vowel/i_en [i:] en | ||||
[i:] fr | [i:] fr | ||||
vowel/o-_4 [o] ro | vowel/o-_4 [o] ro | ||||
vowel/o_5 [O] nl | vowel/o_5 [O] nl | ||||
vowel/o_6 [O:] nl | vowel/o_6 [O:] nl | ||||
vowel/o_7 [u] zh | |||||
vowel/oe [W] en | vowel/oe [W] en | ||||
[W] af | [W] af | ||||
[W] de | [W] de | ||||
vowel/oe_4 [W] sv | vowel/oe_4 [W] sv | ||||
vowel/o_mid [O] fr | vowel/o_mid [O] fr | ||||
[o] hu | [o] hu | ||||
[o] zh | |||||
[o] tr | [o] tr | ||||
vowel/oo [O:] en_sc | vowel/oo [O:] en_sc | ||||
[O] en_wi | [O] en_wi | ||||
vwl_sv/r_sv3 [r] sv | vwl_sv/r_sv3 [r] sv | ||||
vwl_zh/aang [aan] zhy | vwl_zh/aang [aan] zhy | ||||
vwl_zh/aau [Au] zh | vwl_zh/aau [Au] zh | ||||
vwl_zh/a_n [a] zh | |||||
vwl_zh/ang [ang] zhy | vwl_zh/ang [ang] zhy | ||||
vwl_zh/eng [eng] zhy | vwl_zh/eng [eng] zhy | ||||
vwl_zh/iaa [iA] zh | vwl_zh/iaa [iA] zh | ||||
vwl_zh/iaau [iAu] zh | vwl_zh/iaau [iAu] zh | ||||
vwl_zh/ie [iE] zh | |||||
vwl_zh/ing [ing] zhy | vwl_zh/ing [ing] zhy | ||||
vwl_zh/iou [iou] zh | vwl_zh/iou [iou] zh | ||||
vwl_zh/i_rfx [i.] zh | |||||
vwl_zh/ng [ng] zhy | vwl_zh/ng [ng] zhy | ||||
vwl_zh/oeng [oen] zhy | vwl_zh/oeng [oen] zhy | ||||
vwl_zh/ong [ong] zhy | vwl_zh/ong [ong] zhy |
phoneme N //should be more different from n | phoneme N //should be more different from n | ||||
vcd vel nasal | vcd vel nasal | ||||
vowelout f1=3 f2=2300 300 400 f3=-200 80 len=40 rms=20 brk | |||||
vowelout f1=2 f2=1900 200 400 f3=400 80 len=50 rms=15 brk | |||||
formants NULL | formants NULL | ||||
before _ nn/nn_ | |||||
before _ nn/nn2_ | |||||
after _ nn/_nn | after _ nn/_nn | ||||
lengthmod 1 | |||||
lengthmod 7 | |||||
after (i) nn/inn | after (i) nn/inn | ||||
endphoneme | endphoneme | ||||
vowel starttype (a) endtype (a) | vowel starttype (a) endtype (a) | ||||
length 250 | length 250 | ||||
formants vowel/a | formants vowel/a | ||||
before N vwl_zh/a_n | |||||
endphoneme | endphoneme | ||||
phoneme A | phoneme A | ||||
phoneme i. //after ts. ts.h s. z. | phoneme i. //after ts. ts.h s. z. | ||||
vowel starttype (i) endtype (i) | vowel starttype (i) endtype (i) | ||||
length 250 | length 250 | ||||
formants vwl_zh/i_rfx | |||||
// formants vwl_zh/i_rfx | |||||
formants vowel/i#_6 | |||||
endphoneme | endphoneme | ||||
phoneme iA //try more before N | phoneme iA //try more before N | ||||
phoneme iE | phoneme iE | ||||
vowel starttype (i) endtype (e) | vowel starttype (i) endtype (e) | ||||
length 250 | length 250 | ||||
formants vdiph2/ie_2 | |||||
formants vwl_zh/ie | |||||
endphoneme | endphoneme | ||||
phoneme io | phoneme io | ||||
vowel starttype (o) endtype (o) | vowel starttype (o) endtype (o) | ||||
length 250 | length 250 | ||||
formants vdiph2/o_oo | formants vdiph2/o_oo | ||||
before N vowel/o_mid | |||||
endphoneme | endphoneme | ||||
phoneme o- | phoneme o- | ||||
vowel starttype (u) endtype (u) | vowel starttype (u) endtype (u) | ||||
length 250 | length 250 | ||||
formants vowel/u_bck | formants vowel/u_bck | ||||
before N vowel/o_7 | |||||
endphoneme | endphoneme | ||||
phoneme uA | phoneme uA |
case RULE_DIGIT: | case RULE_DIGIT: | ||||
if(IsDigit(letter_w)) | if(IsDigit(letter_w)) | ||||
{ | { | ||||
match.points += (21-distance_right); | |||||
match.points += (20-distance_right); | |||||
post_ptr += letter_xbytes; | post_ptr += letter_xbytes; | ||||
} | } | ||||
else | else | ||||
if(langopts.tone_numbers) | if(langopts.tone_numbers) | ||||
{ | { | ||||
// also match if there is no digit | // also match if there is no digit | ||||
match.points += (21-distance_right); | |||||
match.points += (20-distance_right); | |||||
post_ptr--; | post_ptr--; | ||||
} | } | ||||
else | else |
{ | { | ||||
if((p->type == phVOWEL) && (p->tone >= 4)) | if((p->type == phVOWEL) && (p->tone >= 4)) | ||||
{ | { | ||||
final_stressed = ix; | |||||
count_stressed++; | |||||
if(count_stressed == 0) | |||||
final_stressed = ix; | |||||
if(p->tone >= 4) | |||||
{ | |||||
final_stressed = ix; | |||||
count_stressed++; | |||||
} | |||||
} | } | ||||
} | } | ||||
{ | { | ||||
// LANG=vi | // LANG=vi | ||||
p = &phoneme_list[final_stressed]; | p = &phoneme_list[final_stressed]; | ||||
p->tone = 7; | |||||
if(p->tone_ph == 0) | if(p->tone_ph == 0) | ||||
p->tone_ph = LookupPh("7"); // change tone 1 to falling tone at end of clause | p->tone_ph = LookupPh("7"); // change tone 1 to falling tone at end of clause | ||||
} | } | ||||
p->pitch2 = pitch_adjust + phoneme_tab[tone_ph]->end_type; | p->pitch2 = pitch_adjust + phoneme_tab[tone_ph]->end_type; | ||||
} | } | ||||
} | } | ||||
phoneme_list[final_stressed].tone = 7; | |||||
} // end of Translator::CalcPitches_Tone | } // end of Translator::CalcPitches_Tone | ||||
if((out_start = outbuf) == NULL) | if((out_start = outbuf) == NULL) | ||||
return(EE_INTERNAL_ERROR); | return(EE_INTERNAL_ERROR); | ||||
// allocate space for event list. Allow 200 events per second | |||||
n_event_list = (buf_length*200)/1000; | |||||
// allocate space for event list. Allow 200 events per second. | |||||
// Add a constant to allow for very small buf_length | |||||
n_event_list = (buf_length*200)/1000 + 20; | |||||
if((event_list = (espeak_EVENT *)realloc(event_list,sizeof(espeak_EVENT) * n_event_list)) == NULL) | if((event_list = (espeak_EVENT *)realloc(event_list,sizeof(espeak_EVENT) * n_event_list)) == NULL) | ||||
return(EE_INTERNAL_ERROR); | return(EE_INTERNAL_ERROR); | ||||
#include "translate.h" | #include "translate.h" | ||||
#include "wave.h" | #include "wave.h" | ||||
const char *version_string = "1.31.05 04.Feb.08"; | |||||
const char *version_string = "1.31.06 05.Feb.08"; | |||||
const int version_phdata = 0x013105; | const int version_phdata = 0x013105; | ||||
int option_device_number = -1; | int option_device_number = -1; |
unsigned char *amp_env; | unsigned char *amp_env; | ||||
PHONEME_TAB *ph; | PHONEME_TAB *ph; | ||||
PHONEME_TAB *prev_ph; | PHONEME_TAB *prev_ph; | ||||
static int sourceix=0; | |||||
#ifdef TEST_MBROLA | #ifdef TEST_MBROLA | ||||
if(mbrola_name[0] != 0) | if(mbrola_name[0] != 0) | ||||
if(p->newword) | if(p->newword) | ||||
{ | { | ||||
last_frame = NULL; | last_frame = NULL; | ||||
sourceix = (p->sourceix & 0x7ff) + clause_start_char; | |||||
if(p->newword & 4) | if(p->newword & 4) | ||||
DoMarker(espeakEVENT_SENTENCE, (p->sourceix & 0x7ff) + clause_start_char, 0, count_sentences); // start of sentence | |||||
DoMarker(espeakEVENT_SENTENCE, sourceix, 0, count_sentences); // start of sentence | |||||
// if(p->newword & 2) | // if(p->newword & 2) | ||||
// DoMarker(espeakEVENT_END, count_characters, 0, count_sentences); // end of clause | // DoMarker(espeakEVENT_END, count_characters, 0, count_sentences); // end of clause | ||||
if(p->newword & 1) | if(p->newword & 1) | ||||
DoMarker(espeakEVENT_WORD, (p->sourceix & 0x7ff) + clause_start_char, p->sourceix >> 11, clause_start_word + word_count++); | |||||
DoMarker(espeakEVENT_WORD, sourceix, p->sourceix >> 11, clause_start_word + word_count++); | |||||
} | } | ||||
EndAmplitude(); | EndAmplitude(); | ||||
if(p->prepause > 0) | if(p->prepause > 0) | ||||
DoPause(p->prepause); | DoPause(p->prepause); | ||||
if(option_phoneme_events) | |||||
if(option_phoneme_events && (p->type != phVOWEL)) | |||||
{ | { | ||||
DoMarker(espeakEVENT_PHONEME, (p->sourceix & 0x7ff) + clause_start_char, 0, p->ph->mnemonic); | |||||
// Note, for vowels, do the phoneme event after the vowel-start | |||||
DoMarker(espeakEVENT_PHONEME, sourceix, 0, p->ph->mnemonic); | |||||
} | } | ||||
switch(p->type) | switch(p->type) | ||||
DoSpect(ph,prev->ph,next->ph,1,p,modulation); | DoSpect(ph,prev->ph,next->ph,1,p,modulation); | ||||
} | } | ||||
if(option_phoneme_events) | |||||
{ | |||||
DoMarker(espeakEVENT_PHONEME, sourceix, 0, p->ph->mnemonic); | |||||
} | |||||
DoSpect(p->ph,prev->ph,next->ph,2,p,modulation); | DoSpect(p->ph,prev->ph,next->ph,2,p,modulation); | ||||
memset(vowel_transition,0,sizeof(vowel_transition)); | memset(vowel_transition,0,sizeof(vowel_transition)); |
case L('z','h'): | case L('z','h'): | ||||
case L_zhy: | case L_zhy: | ||||
{ | { | ||||
static const short stress_lengths_zh[8] = {230,150, 230,230, 230,0, 230,250}; | |||||
static const short stress_lengths_zh[8] = {230,150, 230,230, 230,0, 230,300}; | |||||
static const unsigned char stress_amps_zh[] = {22,16, 22,22, 22,22, 22,22 }; | static const unsigned char stress_amps_zh[] = {22,16, 22,22, 22,22, 22,22 }; | ||||
tr = new Translator; | tr = new Translator; |