mjohnson/espeak-ng
1
0
Fork 0
mirror of https://github.com/rhasspy/espeak-ng.git synced 2026-10-31 05:35:35 +00:00
Code Issues Packages Projects Releases Wiki Activity

docs: Document the lexical set transcriptions used by the en language files.

Browse Source
This commit is contained in:
Reece H. Dunn
2017-02-25 10:32:03 +00:00
parent bd04b540e9
commit 5c213b7133
2 changed files with 76 additions and 65 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/languages/gmw/en.md
+50 -39
View File
@@ -15,60 +15,66 @@ Sets<sup>\[<a href="#ref1">1</a>\]</sup>. These were created by Wells in 1982
by comparing the Received Pronunciation British (RP) and General American
(GenAm) accents in use at that time.
The `en` transcriptions listed below are the phonemes used by eSpeak NG to
transcribe the different lexical sets.
## Short Vowels
| Lexical Set | RP | GenAm |
|-------------|-------|-------|
| KIT | ɪ | ɪ |
| DRESS | e | ɛ |
| TRAP | æ | æ |
| LOT | ɒ | ɑ |
| STRUT | ʌ | ʌ |
| FOOT | ʊ | ʊ |
| Lexical Set | en | RP | GenAm |
|-------------|-------|-------|-------|
| KIT | `I` | ɪ | ɪ |
| DRESS | `E` | e | ɛ |
| TRAP | `a` | æ | æ |
| LOT | `0` | ɒ | ɑ |
| STRUT | `V` | ʌ | ʌ |
| FOOT | `U` | ʊ | ʊ |
## Long Vowels
| Lexical Set | RP | GenAm |
|-------------|-------|-------|
| FLEECE | iː | i |
| PALM | ɑː | ɑ |
| THOUGHT | ɔː | ɔ |
| GOOSE | uː | u |
| Lexical Set | en | RP | GenAm |
|-------------|-------|-------|-------|
| FLEECE | `i:` | iː | i |
| PALM | `A:` | ɑː | ɑ |
| THOUGHT | `O:` | ɔː | ɔ |
| GOOSE | `u:` | uː | u |
## Rhotic Vowels
These are vowels that are followed by an `r` that is not part of the next syllable
when considering the root form of the word containing that vowel.
| Lexical Set | RP | GenAm |
|-------------|-------|-------|
| NURSE | ɜː | ɝ |
| START | ɑː | ɑɹ |
| NORTH | ɔː | ɔɹ |
| FORCE | ɔː | oɹ |
| CURE | ʊə̯ | ʊɹ |
| NEAR | ɪə̯ | ɪɹ |
| SQUARE | eə̯ | ɛɹ |
| Lexical Set | en | RP | GenAm |
|-------------|-------|-------|-------|
| NURSE | `3:` | ɜː | ɝ |
| START | `A@` | ɑː | ɑɹ |
| NORTH | `O@` | ɔː | ɔɹ |
| FORCE | `o@` | ɔː | oɹ |
| CURE | `U@` | ʊə̯ | ʊɹ |
| NEAR | `i@3` | ɪə̯ | ɪɹ |
| SQUARE | `e@` | eə̯ | ɛɹ |
__NOTE:__ `/i@3/` is used for the NEAR lexical set to differentiate it from
`/i@/` used in words like `million`.
## Reduced Vowels
These are unstressed vowels that differ from the vowels in the main lexical sets.
| Lexical Set | RP | GenAm |
|-------------|-------|-------|
| HAPPY | ɪ | i |
| COMMA | ə | ə |
| LETTER | ə | ɚ |
| Lexical Set | en | RP | GenAm |
|-------------|-------|-------|-------|
| HAPPY | `i` | ɪ | i |
| COMMA | `@` | ə | ə |
| LETTER | `3` | ə | ɚ |
## Diphthongs
| Lexical Set | RP | GenAm |
|-------------|-------|-------|
| FACE | eɪ̯ | eɪ̯ |
| PRICE | aɪ̯ | aɪ̯ |
| CHOICE | ɔɪ̯ | ɔɪ̯ |
| GOAT | əʊ̯ | oʊ̯ |
| MOUTH | aʊ̯ | aʊ̯ |
| Lexical Set | en | RP | GenAm |
|-------------|-------|-------|-------|
| FACE | `eI` | eɪ̯ | eɪ̯ |
| PRICE | `aI` | aɪ̯ | aɪ̯ |
| CHOICE | `OI` | ɔɪ̯ | ɔɪ̯ |
| GOAT | `oU` | əʊ̯ | oʊ̯ |
| MOUTH | `aU` | aʊ̯ | aʊ̯ |
## Split Vowels
@@ -76,10 +82,15 @@ These are lexical sets defined by John Wells that are merged with other lexical
sets in both RP and GenAm, so have split from one of those lexical sets and
merged with the other.
| Lexical Set | RP | GenAm |
|-------------|-------|-------|
| BATH | ɑː | æ |
| CLOTH | ɒ | ɔ |
| Lexical Set | en | RP | GenAm |
|-------------|-------|-------|-------|
| BATH | `aa` | ɑː | æ |
| CLOTH | `0` | ɒ | ɔ |
__NOTE:__ eSpeak NG does not have a distinct transcription for the CLOTH lexical
set. As such, it can only represent American English accents with the COT-CAUGHT
merger where COT is the American PALM-LOT merger and CAUGHT is the American
THOUGHT-CLOTH merger.
## References
phsource/ph_english
+26 -26
View File
@@ -97,7 +97,7 @@ endphoneme
// VOWELS
//*******************************************************************
phoneme @ // Schwa
phoneme @
vwl starttype #@ endtype #@
unstressed
length 140
@@ -109,7 +109,7 @@ phoneme @ // Schwa
endphoneme
phoneme 3 // Schwa, used for rhotic schwa in American, e.g. bett**er**
phoneme 3
CALL @
ipa ə
endphoneme
@@ -178,7 +178,7 @@ phoneme a# // **a**bout. This may be \[@\] or may be a more open schwa.
endphoneme
phoneme aa // 'b**a**th' etc. This is \[a\] in some accents, \[A:\] in others.
phoneme aa
vwl starttype #a endtype #a
ipa a
length 200
@@ -192,7 +192,7 @@ phoneme aa // 'b**a**th' etc. This is \[a\] in some accents, \[A:\] in other
endphoneme
phoneme E // dr**e**ss
phoneme E
vwl starttype #e endtype #e
length 140
IF nextPhW(n) THEN
@@ -223,7 +223,7 @@ endphoneme
phoneme I // k**i**t
phoneme I
vwl starttype #i endtype #i
length 130
IfNextVowelAppend(;)
@@ -263,7 +263,7 @@ phoneme I# // used for [I] which may be [@] in some accents
endphoneme
phoneme i // happ**y**. optional variant of [I] for end of words
phoneme i
vwl starttype #i endtype #i
unstressed
ipa i
@@ -278,7 +278,7 @@ phoneme i // happ**y**. optional variant of [I] for end of words
endphoneme
phoneme 0 // l**o**t
phoneme 0
vwl starttype #o endtype #o
length 140
ChangeIfDiminished(@)
@@ -304,7 +304,7 @@ phoneme 02 // becomes V in en-us
endphoneme
phoneme V // str**u**t
phoneme V
vwl starttype #a endtype #@
length 140
@@ -313,7 +313,7 @@ phoneme V // str**u**t
endphoneme
phoneme U // f**oo**t
phoneme U
vwl starttype #o endtype #o
length 150
@@ -327,7 +327,7 @@ endphoneme
phoneme A: // p**al**m
phoneme A:
vwl starttype #a endtype #a
length 230
IfNextVowelAppend(r-)
@@ -335,7 +335,7 @@ phoneme A: // p**al**m
endphoneme
phoneme A@ // st**ar**t. Used for [A:] when followed by 'r'.
phoneme A@
vwl starttype #a endtype #a
ipa ɑː
length 230
@@ -352,7 +352,7 @@ endphoneme
phoneme 3: // n**ur**se
phoneme 3:
vwl starttype #@ endtype #@
length 210
IfNextVowelAppend(r-)
@@ -361,7 +361,7 @@ phoneme 3: // n**ur**se
endphoneme
phoneme i: // fl**ee**ce
phoneme i:
vwl starttype #i endtype #i
length 175
IfNextVowelAppend(;)
@@ -372,7 +372,7 @@ phoneme i: // fl**ee**ce
endphoneme
phoneme O: // th**ou**ght
phoneme O:
vwl starttype #o endtype #o
length 230
ChangeIfDiminished(@)
@@ -388,7 +388,7 @@ phoneme O
endphoneme
phoneme O@ // n**or**th
phoneme O@
vwl starttype #o endtype #o
ipa ɔː
length 240
@@ -397,7 +397,7 @@ phoneme O@ // n**or**th
endphoneme
phoneme o@ // f**or**ce
phoneme o@
vwl starttype #o endtype #o
ipa ɔː
length 250
@@ -411,7 +411,7 @@ phoneme o@ // f**or**ce
endphoneme
phoneme u: // g**oo**se
phoneme u:
vwl starttype #u endtype #u
length 210
IF nextPh(l/2) THEN
@@ -424,14 +424,14 @@ phoneme u: // g**oo**se
endphoneme
phoneme aU // m**ou**th
phoneme aU
vwl starttype #a endtype #u
length 230
FMT(vdiph/au)
endphoneme
phoneme oU // g**oa**t
phoneme oU
vwl starttype #@ endtype #u
ipa əʊ
length 220
@@ -453,7 +453,7 @@ phoneme oU#
endphoneme
phoneme aI // pr**i**ce
phoneme aI
vwl starttype #a endtype #i
length 240
IF nextPh(#a) THEN
@@ -463,21 +463,21 @@ phoneme aI // pr**i**ce
endphoneme
phoneme eI // f**a**ce
phoneme eI
vwl starttype #e endtype #i
length 210
FMT(vdiph/eei)
endphoneme
phoneme OI // ch**oi**ce
phoneme OI
vwl starttype #o endtype #i
length 230
FMT(vdiph/ooi)
endphoneme
phoneme e@ // squ**are**
phoneme e@
vwl starttype #e endtype #@
length 230
IfNextVowelAppend(r-)
@@ -485,7 +485,7 @@ phoneme e@ // squ**are**
endphoneme
phoneme i@ // mill**io**n
phoneme i@
vwl starttype #i endtype #@
length 250
IfNextVowelAppend(r-)
@@ -493,7 +493,7 @@ phoneme i@ // mill**io**n
endphoneme
phoneme i@3 // n**ear**
phoneme i@3
vwl starttype #i endtype #@
length 250
@@ -506,7 +506,7 @@ phoneme i@3 // n**ear**
endphoneme
phoneme U@ // c**ure**
phoneme U@
vwl starttype #u endtype #@
length 200
IfNextVowelAppend(r-)