English (UK) MFA dictionary v2.0.0#
@techreport{mfa_english_uk_mfa_dictionary_2022,
author={McAuliffe, Michael and Sonderegger, Morgan},
title={English (UK) MFA dictionary v2.0.0},
address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/English/English (UK) MFA dictionary v2_0_0.html}},
year={2022},
month={Mar},
}
G2P models |
Installation#
Install from the MFA command line:
mfa model download dictionary english_uk_mfa
Or download from the release page.
The dictionary available from the release page and command line installation has pronunciation and silence probabilities estimated as part acoustic model training (see Silence probability format and training pronunciation probabilities for more information. If you would like to use the version of this dictionary without probabilities, please see the plain dictionary.
Intended use#
This dictionary is intended for forced alignment of English transcripts.
This dictionary uses the MFA phone set for English, and was used in training the English MFA acoustic model. Pronunciations can be added on top of the dictionary, as long as no additional phones are introduced.
Performance Factors#
When trying to get better alignment accuracy, adding pronunciations is generally helpful, especially for different styles and dialects. The most impactful improvements will generally be seen when adding reduced variants that involve deleting segments/syllables common in spontaneous speech. Alignment must include all phones specified in the pronunciation of a word, and each phone has a minimum duration (by default 10ms). If a speaker pronounces a multisyllabic word with just a single syllable, it can be hard for MFA to fit all the segments in, so it will lead to alignment errors on adjacent words as well.
Ethical considerations#
Deploying any Speech-to-Text model into any production setting has ethical implications. You should consider these implications before use.
Demographic Bias#
You should assume every machine learning model has demographic bias unless proven otherwise. For pronunciation dictionaries, it is often the case that transcription accuracy and lexicon coverage for the prestige variety modeled in this dictionary compared to other variants. If you are using this dictionary in production, you should acknowledge this as a potential issue.
IPA Charts#
Consonants#
Obstruent symbols to the left of are unvoiced and those to the right are voiced.
Manner |
Labial |
Labiodental |
Dental |
Alveolar |
Alveopalatal |
Palatal |
Velar |
Glottal |
---|---|---|---|---|---|---|---|---|
Nasal |
Occurrences: 16,401 Examples: * zomb: [z ɒ m b] * mondo: [m ɒ n d əw] * boom: [b ʉː m] * reem: [ɹ iː m] Occurrences: 1,119 Examples: * mewl: [mʲ ʉː ɫ] * pmee: [pʰ iː mʲ iː] * mitra: [mʲ iː t ɹ ə] * jimi: [dʒ ɪ mʲ iː] Occurrences: 1,684 Examples: * totem: [tʰ əw t m̩] * kalam: [cʰ æ l m̩] * osm: [ɒː s m̩] * venom: [v ɛ n m̩] |
Occurrences: 947 Examples: * unfit: [ɐ ɱ f ɪ t] * nymph: [ɲ ɪ ɱ f] * infra: [ɪ ɱ f ɹ ə] * info: [ɪ ɱ f əw] |
Occurrences: 27,375 Examples: * mondo: [m ɒ n d əw] * calne: [kʰ ɑː ɫ n] * nests: [n ɛ s t s] * hone: [h əw n] Occurrences: 5,030 Examples: * cyan: [s aj n̩] * lumen: [l ʉː m n̩] * avant: [æ v n̩ t] * seiun: [s ej n̩] |
Occurrences: 6,303 Examples: * niall: [ɲ iː ɫ] * knees: [ɲ iː z] * upney: [ɐ p ɲ i] * renee: [ɹ ɪ ɲ ej] |
Occurrences: 6,054 Examples: * ring: [ɹ ɪ ŋ] * anglo: [æ ŋ ɡ l əw] * hangs: [h æ ŋ z] * tango: [tʰ æ ŋ ɡ əw] |
|||
Stop |
Occurrences: 9,434 Examples: * julep: [dʒ ʉː l ɛ p] * jape: [dʒ ej p] * pulp: [pʰ ɐ ɫ p] * spree: [s p ɹ iː] Occurrences: 11,922 Examples: * zomb: [z ɒ m b] * bipod: [b aj pʰ ɒ d] * wehrb: [v ɛː b] * boom: [b ʉː m] |
Occurrences: 26,137 Examples: * sleet: [s ʎ iː t] * pitt: [pʰ ɪ t] * nests: [n ɛ s t s] * aorta: [ej ɒː t ə] Occurrences: 20,290 Examples: * bipod: [b aj pʰ ɒ d] * mondo: [m ɒ n d əw] * aide: [ej d] * rode: [ɹ əw d] |
Occurrences: 4,476 Examples: * dicky: [d ɪ c i] * craps: [c ɹ æ p s] * crest: [c ɹ ɛ s t] * clear: [c ʎ ɪ ə] Occurrences: 2,556 Examples: * grift: [ɟ ɹ ɪ f t] * doggy: [d ɒ ɟ i] * voggy: [v ɒ ɟ i] * grim: [ɟ ɹ ɪ m] |
Occurrences: 14,363 Examples: * yolks: [j əw k s] * aches: [ej k s] * duck: [d ɐ k] * score: [s k ɒː] Occurrences: 5,156 Examples: * ghazi: [ɡ ɑː z i] * oghuz: [əw ɡ ʉː z] * anglo: [æ ŋ ɡ l əw] * saga: [s ɑː ɡ ə] |
Occurrences: 5,976 Examples: * sleet: [s ʎ iː ʔ] * pitt: [pʰ ɪ ʔ] * bated: [b ej ʔ ɪ d] * aorta: [ej ɒː ʔ ə] |
|||
Affricate |
Occurrences: 4,694 Examples: * ouch: [aw tʃ] * tew: [tʃ ʉː] * tubes: [tʃ ʉː b z] * cham: [tʃ æ m] Occurrences: 5,097 Examples: * julep: [dʒ ʉː l ɛ p] * jape: [dʒ ej p] * mangy: [m ej n dʒ i] * due: [dʒ ʉː] |
|||||||
Sibilant |
Occurrences: 34,516 Examples: * sleet: [s ʎ iː t] * yolks: [j əw k s] * dis: [d ɪ s] * nests: [n ɛ s t s] Occurrences: 13,750 Examples: * zomb: [z ɒ m b] * ghazi: [ɡ ɑː z i] * pours: [pʰ aw ə z] * oghuz: [əw ɡ ʉː z] |
Occurrences: 6,624 Examples: * sheol: [ʃ ej əw ɫ] * cush: [kʰ ʊ ʃ] * osh: [ɒ ʃ] * shew: [ʃ əw] Occurrences: 506 Examples: * lujvo: [l ʉ ʒ v əw] * asian: [ej ʒ ə n] * ezhes: [ɛ ʒ ɪ z] * usual: [j ʉː ʒ] |
||||||
Fricative |
Occurrences: 9,135 Examples: * foyer: [f ɔj ə] * grift: [ɟ ɹ ɪ f t] * efs: [ɛ f s] * guff: [ɡ ɐ f] Occurrences: 779 Examples: * fee: [fʲ iː] * goofy: [ɡ ʉː fʲ i] * feed: [fʲ iː d] * fused: [fʲ ʉː z d] Occurrences: 6,473 Examples: * wehrb: [v ɛː b] * vials: [v aj ɫ̩ z] * calve: [kʰ ɑː v] * veer: [v ɪ ə] Occurrences: 457 Examples: * vth: [vʲ iː tʰ iː ej tʃ] * vela: [vʲ iː l ə] * tovey: [tʰ əw vʲ i] * civvy: [s ɪ vʲ i] |
Occurrences: 2,418 Examples: * oath: [əw θ] * width: [w ɪ d θ] * roth: [ɹ ɒ θ] * vth: [f ɪ f θ] Occurrences: 530 Examples: * thass: [ð æ s] * hythe: [h aj ð] * dalet: [d ɑː ʎ ɪ ð] * thilk: [ð ɪ ɫ k] |
Occurrences: 950 Examples: * hick: [ç ɪ k] * hail: [ç ej ɫ] * hiply: [ç ɪ p ʎ i] * hits: [ç ɪ t s] |
Occurrences: 3,623 Examples: * hone: [h əw n] * hoary: [h ɒː ɹ i] * hares: [h ɛː z] * hangs: [h æ ŋ z] |
||||
Approximant |
Occurrences: 5,151 Examples: * wale: [w ej ɫ] * wood: [w ʊ d] * width: [w ɪ d θ] * what: [w ɒ t] |
Occurrences: 27,515 Examples: * ring: [ɹ ɪ ŋ] * rode: [ɹ əw d] * reem: [ɹ iː m] * rare: [ɹ ɛː] |
Occurrences: 1,419 Examples: * yolks: [j əw k s] * eau: [j ʉː] * usona: [j ʉː s əw n ə] * urus: [j ʊ ɹ ə s] |
|||||
Lateral |
Occurrences: 12,269 Examples: * julep: [dʒ ʉː l ɛ p] * lay: [l ej] * anglo: [æ ŋ ɡ l əw] * elena: [ɛ l ej n ə] Occurrences: 6,206 Examples: * wale: [w ej ɫ] * sheol: [ʃ ej əw ɫ] * calne: [kʰ ɑː ɫ n] * niall: [ɲ iː ɫ] Occurrences: 4,956 Examples: * niall: [n aj ɫ̩] * real: [ɹ iː ɫ̩] * aral: [ɛː ɹ ɫ̩] * vials: [v aj ɫ̩ z] |
Occurrences: 8,389 Examples: * sleet: [s ʎ iː t] * poley: [pʰ əw ʎ i] * litt: [ʎ ɪ t] * clear: [c ʎ ɪ ə] |
Vowels#
Vowel symbols to the left of are unrounded and those to the right are rounded.
Front |
Near-Front |
Central |
Near-Back |
Back |
|
---|---|---|---|---|---|
Close |
Occurrences: 12,825 Examples: * ghazi: [ɡ ɑː z i] * poley: [pʰ əw ʎ i] * hoary: [h ɒː ɹ i] * mangy: [m ej n dʒ i] Occurrences: 9,224 Examples: * sleet: [s ʎ iː t] * sheol: [ʃ iː əw ɫ] * dis: [dʲ iː aj ɛ s] * reem: [ɹ iː m] |
Occurrences: 903 Examples: * ueno: [j ʉ ej ɲ əw] * poo: [pʰ ʉ] * lukan: [l ʉ k n̩] * duets: [d ʉ ɛ t s] Occurrences: 6,380 Examples: * boom: [b ʉː m] * tew: [tʃ ʉː] * julep: [dʒ ʉː l ɛ p] * oghuz: [əw ɡ ʉː z] |
|||
Occurrences: 45,039 Examples: * ring: [ɹ ɪ ŋ] * dis: [d ɪ s] * pitt: [pʰ ɪ t] * bated: [b ej ʔ ɪ d] |
Occurrences: 3,041 Examples: * wood: [w ʊ d] * cush: [kʰ ʊ ʃ] * dour: [d ʊ ə] * pouf: [pʰ ʊ f] |
||||
Close-Mid |
Occurrences: 10,857 Examples: * wale: [w ej ɫ] * sheol: [ʃ ej əw ɫ] * aide: [ej d] * bated: [b ej ʔ ɪ d] |
||||
Occurrences: 45,598 Examples: * pours: [pʰ aw ə z] * foyer: [f ɔj ə] * aorta: [ej ɒː t ə] * orme: [ɒː m ə] |
|||||
Open-Mid |
Occurrences: 15,330 Examples: * dis: [dʲ iː aj ɛ s] * nests: [n ɛ s t s] * julep: [dʒ ʉː l ɛ p] * reem: [ɹ ɪ ɛ m] Occurrences: 1,060 Examples: * wehrb: [v ɛː b] * rare: [ɹ ɛː] * bear: [b ɛː] * hares: [h ɛː z] |
Occurrences: 197 Examples: * perv: [pʰ ɜ v] * derf: [d ɜ f] * ubert: [j ʉ b ɜ t] * furor: [fʲ ɜ ɹ ə] Occurrences: 3,341 Examples: * duh: [d ɜː] * berne: [b ɜː n] * glogg: [ɡ l ɜː ɡ] * burns: [b ɜː n z] |
|||
Occurrences: 14,824 Examples: * anglo: [æ ŋ ɡ l əw] * aloo: [æ l ʉː] * palp: [pʰ æ ɫ p] * cyan: [s aj æ n] |
Occurrences: 6,766 Examples: * bum: [b ɐ m] * pulp: [pʰ ɐ ɫ p] * uggo: [ɐ ɡ əw] * guff: [ɡ ɐ f] |
||||
Open |
Occurrences: 765 Examples: * karin: [kʰ ɑ ɹ ɪ n] * marfa: [m ɑ f ə] * agua: [ɑ ɡ w ə] * kaon: [cʰ ej ɑ n] Occurrences: 5,659 Examples: * ghazi: [ɡ ɑː z i] * calne: [kʰ ɑː ɫ n] * saga: [s ɑː ɡ ə] * real: [ɹ ej ɑː ɫ] Occurrences: 10,578 Examples: * zomb: [z ɒ m b] * bipod: [b aj pʰ ɒ d] * mondo: [m ɒ n d əw] * glogg: [ɡ l ɒ ɡ] Occurrences: 5,239 Examples: * aorta: [ej ɒː t ə] * orme: [ɒː m ə] * hoary: [h ɒː ɹ i] * score: [s k ɒː] |
Diphthongs#
aj
aw
ej
ɔj
əw