次の方法で共有

  • [アーティクル]

Note

Please see Azure Cognitive Services for Speech documentation for the latest supported speech solutions.

Consonants (Microsoft.Speech)

This is a listing and cross reference of the consonant phones that are used by the three phonetic alphabets that Microsoft.Speech supports. See Phonetic Alphabet Reference (Microsoft.Speech).

UPS Consonant Labeling Conventions

Consonant symbols are one, two, or three characters in length. Single character labels are used for highly frequent, language-universal consonants. In general the first character signifies place plus voicing, and the second signifies manner. Labeling conventions carry across phonetic classes in the following ways:

Conventions for place of articulation and voicing labeling:

  • Dentals and alveloars use D for voiced and T for voiceless

  • Alveolars use N for nasal and R for approximant

  • Labiodentals use V for voiced and F for voiceless

  • Velars use G for voiced and K for voiceless, NG for nasal

  • Bilabials use B for voiced and P for voiceless stops, M for nasal

  • Laterals use L

  • Uvulars use Q

  • Pharyngeals use H

  • Palatals use X for voiceless and Y for voiced

Conventions for manner of articulation labeling:

  • Single-character labels have manner encapsulated in choice of symbol: for example, M is nasal

  • Trills double up the first symbol

  • Fricatives use H

  • Approximants use X

  • Retroflex uses R

Consonant Tables

There are two tables, each listing a different category of consonant phones:

  • Consonants

  • Affricates

The table columns contain the following information:

  • UPS. The phone label specified in Microsoft’s Universal Phone Set (UPS).

  • IPA. The phone label specified in the International Phonetic Alphabet (IPA) phone set.

  • Unicode. The Unicode value from the IPA phone set.

  • SAPI ID. The Speech API (SAPI) phone ID, which in most cases is the Unicode value.

  • IPA Description. This column describes the features of the phone, such as place of articulation, manner of articulation, phonation type, and airstream mechanism. See the Glossary of Phonetic Terms (Microsoft.Speech) for more information about the features of phones.

  • ipaASCII. The equivalent phone in ipaASCII phone set for reference.

  • X-SAMPA. The equivalent phone in X-SAMPA for reference.

  • Example. A word that uses the phone and illustrates its sound.

  • Language. The language of the example word.

Consonant Phones

Consonants are speech sounds that are articulated with complete (voiceless) or partial closure of the vocal tract.

UPS

IPA

Unicode

SAPI ID

IPA Description

ipaASCII

X-SAMPA

Example

Language

P

p

U+0070

0070

{blb,stp,vls}

P

P

put

English

B

b

U+0062

0062

{blb,stp,vcd}

B

B

big

English

M

m

U+006D

006D

{blb,nas}

M

M

mat

English

BB

ʙ

U+0299

0299

{blb,trl,vcd}

B<trl>

B\

PH

ɸ

U+0278

0278

{blb,frc,vls}

P

p\

BH

β

U+03B2

03B2

{blb,frc,vcd}

B

B

kabra

Spanish

MF

ɱ

U+0271

0271

{lbd,nas}

M

F

F

f

U+0066

0066

{lbd,frc,vls}

F

F

fork

English

V

v

U+0076

0076

{lbd,frc,vcd}

V

V

vat

English

VA

ʋ

U+028B

028B

{lbd,apr,vcd}

R<lbd>

V\

TH

θ

U+03B8

03B8

{dnt,frc,vls}

T

T

thin

English

DH

ð

U+00F0

00F0

{dnt,frc,vcd}

D

D

then

English

T

t

U+0074

0074

{alv,stp,vls}

T

T

talk

English

D

d

U+0064

0064

{alv,stp,vcd}

D

D

dig

English

N

n

U+006E

006E

{alv,nas}

N

N

no

English

RR

r

U+0072

0072

{alv,trl,vcd}

R<trl>

torre, rojo

Spanish

DX

ɾ

U+027E

027E

{alv,flp,vcd}

*

4

butter

US English

S

s

U+0073

0073

{alv,frc,vls}

S

S

sit

US English

Z

z

U+007A

007A

{alv,frc,vcd}

z

Z

zap

US English

LSH

ɬ

U+026C

026C

{alv,lat,frc,vls}

K

LH

ɮ

U+026E

026E

{alv,lat,frc,vcd}

z<lat>

K\

caballo

Spanish/Zulu

RA

ɹ

U+0279

0279

{alv,apr}

r

r\

puro

Spanish

L

l

U+006C

006C

{alv,lat,apr,vcd}

l

L

lid

US English

L vel

ɫ

U+026B

006C 02E0

{alv,lat,apr,vcd} vel

SH

ʃ

U+0283

0283

{pla,frc,vls}

S

S

she

US English

SH pal

ʆ

U+0286

0283 02B2

{pla,frc,vls} pal

ZH

ʒ

U+0292

0292

{pla,frc,vcd}

Z

Z

pleasure

US English

ZH pal

ʓ

U+0293

0292 02B2

{pla,frc,vcd} pal

TR

ʈ

U+0288

0288

{rfx,stp,vls}

t.

t`

DR

ɖ

U+0256

0256

{rfx,stp,vcd}

d.

D`

NR

ɳ

U+0273

0273

{rfx,nas,vcd}

n.

N`

DXR

ɽ

U+027D

027D

{rfx,flp,vcd}

*.

r`

SR

ʂ

U+0282

0282

{rfx,frc,vls}

s.

S`

ZR

ʐ

U+0290

0290

{rfx,frc,vcd}

z.

Z`

R

ɻ

U+027B

027B

{rfx,apr,vcd}

r.

R

red

US English

LR

ɭ

U+026D

026D

{rfx,lat,vcd}

l.

l`

RR rho

ɼ

U+027C

0072 02DE

{rfx,trl}

CT

c

U+0063

0063

{pal,stp,vls}

c

C

JD

ɟ

U+025F

025F

{pal,stp,vcd}

J

J\

NJ

ɲ

U+0272

0272

{pal,nas,vcd}

J

oignon

French

C

ç

U+00E7

00E7

{pal,frc,vls}

C

C

sicher

German

CJ

ʝ

U+029D

029D

{pal,frc,vcd}

C<vcd>

j\

J

j

U+006A

006A

{pal,apr,vcd}

j

J

yard

US English

LJ

ʎ

U+028E

028E

{pal,lat,apr,vcd}

l^

L

gli

Italian

W

w

U+0077

0077

{lbv,apr,vcd}

w

W

with

US English

K

k

U+006B

006B

{vel,stp,vls}

k

K

cut

US English

G

g

U+0067

0067

{vel,stp,vcd}

g

G

gut

US English

NG

ŋ

U+014B

014B

{vel,nas}

N

N

sing

US English

X

x

U+0078

0078

{vel,frc,vls}

x

X

mujer

Spanish

GH

ɣ

U+0263

0263

{vel,frc,vcd}

Q

7

luego

Spanish

GA

ɰ

U+0270

0270

{vel,apr,vcd}

j<vel>

M\

GL

ʟ

U+029F

029F

{vel,lat,vcd}

L

L\

QT

q

U+0071

0071

{uvl,stp,vls}

q

Q

QD

ɢ

U+0262

0262

{uvl,stp,vcd}

G

G\

QN

ɴ

U+0274

0274

{uvl,nas,vcd}

n"

N\

QQ

ʀ

U+0280

0280

{uvl,trl,vcd}

-

R\

QH

χ

U+03C7

03C7

{uvl,frc,vls}

X

X

RH

ʁ

U+0281

0281

{uvl,frc,vcd}

g"

R

rond

French

HH

ħ

U+0127

0127

{phr,frc,vls}

H

X\

HG

ʕ

U+0295

0295

{phr,frc,vcd}

H<vcd>

?\

GT

ʔ

U+0294

0294

{glt,stp,vls}

?

?

H

h

U+0068

0068

{glt,frc,vls}

h

H

help

US English

WJ

ɥ

U+0265

0265

{lbp,apr,vcd}

j<rnd>

H

huit;juin

French

Affricates

Affricates are consonants that begin as stops but release as a fricative.

The SAPI UPS includes explicit phones for the most frequent affricates in the highest priority languages. The IPA Unicode specification includes some of these phones, though not all of them. The SAPI phone IDs for affricates do not use the IPA Unicode values but instead concatenate the phone IDs of the component phones. Each SAPI phone ID is four characters long. For example JH 006A0361006A is equivalent to D + ZH 006A 0361 006A.

This approach maintains the semantic equivalence of compounds and their constituents. It has been used to generate the phone IDs of all compound phones in the UPS. This allows a speech recognition engine to easily decompose the compound phone into its constituent phones, which may be required if the engine does not model one of these compounds. The table shows the constituent phones (UPS Compound) that were combined to create the resulting explicit phone (UPS).

UPS

UPS Compound

IPA

Unicode

SAPI ID

IPA Description

Example

Language

PF

P + F

p.f

007003610066

{lbd,aff,vls}

Pfahl

German

TS

T + S

t.s / ʦ

U+02A6

007403610073

{den,aff,vls}

Zahl

German

CH

T + SH

t.ʃ / ʧ

U+02A7

007403610283

{pla,aff,vls}

chin

English

JH

D + ZH

d.ʒ / ʤ

U+02A4

006403610292

{pla,aff,vcd}

joy

English

JJ

J + J

j.j

006A0361006A

{pal,aff,vcd}

hielo

Spanish

DZ

D + Z

d.z / ʣ

U+02A3

00640361007A

{den,aff,vcd}

zona

Italian

CC

T + SC

t.ɕ / ʨ

U+02A8

007403610255

{pla,aff,vls}

Mandarin

TSR

T + SR

t.ʂ

007403610282

{rfx,aff,vls}

Mandarin

JC

D + ZC

d.ʑ / ʥ

U+02A5

006403610291

{pla,aff,vcd}

Many other affricates or compound consonants exist and may be required by other languages. For example, there are about 20 Italian geminate consonants (doubled consonants). UPS does not include dedicated phones for these, but allows them to be defined using the + symbol, or the length diacritic ‘lng’ in the case of geminates. For more details, see Parsing Guidelines for SAPI Speech Recognition Phone Converters (Microsoft.Speech).