Ctrl+K
Ctrl+K

Swedish MFA dictionary v2.0.0a#

  • Maintainer: Montreal Forced Aligner

  • Language: Swedish

  • Dialect: N/A

  • Phone set: MFA

  • Number of words: 59,520

  • Phones: a a˥˧ a˥˩ a˧˩ a˩ b d̪ eː eː˥˧ eː˥˩ eː˧˩ eː˩ f h iː iː˥˧ iː˥˩ iː˧˩ j k kʰ l m n̪ oː oː˥˧ oː˥˩ oː˧˩ oː˩ p pʰ r s̪ t̪ t̪ʰ uː uː˥˧ uː˥˩ uː˧˩ yː yː˥˧ yː˥˩ yː˧˩ yː˩ øː øː˥˧ øː˥˩ øː˧˩ øː˩ ŋ œ œ˥˩ œ˧˩ ɑː ɑː˥˧ ɑː˥˩ ɑː˧˩ ɑː˩ ɔ ɔ˥˧ ɔ˥˩ ɔ˧˩ ɔ˩ ɕ ɖ ɛ ɛː ɛː˥˧ ɛː˥˩ ɛː˧˩ ɛ˥˧ ɛ˥˩ ɛ˧˩ ɛ˩ ɡ ɧ ɪ ɪ˥˧ ɪ˥˩ ɪ˧˩ ɪ˩ ɭ ɳ ɵ ɵ˥˧ ɵ˥˩ ɵ˧˩ ɵ˩ ʂ ʈ ʈʰ ʉː ʉː˥˧ ʉː˥˩ ʉː˧˩ ʊ ʊ˥˩ ʊ˧˩ ʊ˩ ʋ ʏ ʏ˥˧ ʏ˥˩ ʏ˧˩ ʏ˩

  • License: CC BY 4.0

  • Compatible MFA version: v2.0.0

  • Citation:

@techreport{mfa_swedish_mfa_dictionary_2022,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Swedish MFA dictionary v2.0.0a},
	address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Swedish/Swedish MFA dictionary v2_0_0a.html}},
	year={2022},
	month={May},
}

G2P models

Acoustic models
../../_images/full_logo_yellow.svg

Installation#

Install from the MFA command line:

mfa model download dictionary swedish_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 Swedish transcripts.

This dictionary uses the MFA phone set for Swedish, and was used in training the Swedish 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

Retroflex

Palatal

Velar

Glottal

Nasal
Occurrences:
14,317
Examples:
* måste:
[m o ː s ̪ t ̪ ɛ ˥ ˩]
* moské:
[m ʊ s ̪ k e ː]
* nämen:
[n ̪ ɛ m ɛ n ̪]
* omval:
[ɔ ˧ ˩ m ʋ ɑ ː l]
Occurrences:
39,682
Examples:
* fann:
[f a n ̪]
* sexan:
[s ̪ ɛ k s ̪ a n ̪]
* renar:
[r ɛ n ̪ ɑ ː r]
* ovant:
[o ː ʋ a n ̪ t ̪]
Occurrences:
2,292
Examples:
* erna:
[ɛ ɳ a]
* åarna:
[ɔ ˧ ˩ a ˥ ˩ ɳ a]
* torn:
[t ̪ ʰ u ː ɳ]
* garn:
[ɡ a ɳ]
Occurrences:
7,498
Examples:
* änka:
[ɛ ŋ k a]
* vagn:
[ʋ a ŋ n ̪]
* lunga:
[l ɵ ˧ ˩ ŋ a ˥ ˩]
* manga:
[m a ŋ ɡ a]

Stop
Occurrences:
7,258
Examples:
* yop:
[y ː ɔ p]
* pals:
[p ʰ a l s ̪]
* pusha:
[p ʰ ɵ ˧ ˩ ɧ a ˥ ˩]
* spis:
[s ̪ p i ː s ̪]
Occurrences:
9,730
Examples:
* busig:
[b ʉ ː s ̪ ɪ ɡ]
* bota:
[b u ː t ̪ a]
* sjöbo:
[ɧ ø ː b u ː]
* akbar:
[a k b ɑ ː ˥ ˩ r]
Occurrences:
31,064
Examples:
* tejp:
[t ̪ ʰ e ː e ː j i ː p ʰ e ː]
* ögats:
[ø ː ɡ ɑ ː t ̪ s ̪]
* tält:
[t ̪ ʰ ɛ l t ̪]
* vänts:
[ʋ ɛ ˧ ˩ n ̪ t ̪ s ̪]
Occurrences:
19,536
Examples:
* bodin:
[b u ː d ̪ ɪ n ̪]
* föds:
[f ø ː d ̪ s ̪]
* ålagd:
[o ː ˧ ˩ l a ˥ ˩ j d ̪]
* dana:
[d ̪ ɑ ː n ̪ a]
Occurrences:
1,325
Examples:
* lärt:
[l ɛ ʈ]
* kort:
[k ʰ ɔ ʈ]
* farit:
[f ɑ ː ʈ]
* mart:
[m a ˥ ˧ ʈ]
Occurrences:
1,760
Examples:
* gerda:
[ɡ ɛ ˥ ˩ ɖ a]
* hörd:
[h ø ː ɖ]
* hörde:
[h œ ɖ ɛ]
* körda:
[ɕ ø ː ɖ a]
Occurrences:
18,094
Examples:
* gick:
[j ɪ k]
* lukt:
[l ɵ k t ̪]
* kår:
[k ʰ o ː r]
* klent:
[k l e ː n ̪ t ̪]
Occurrences:
10,087
Examples:
* suget:
[s ̪ ʉ ː ɡ ɛ t ̪]
* gavs:
[ɡ ɑ ː ʋ s ̪]
* gasen:
[ɡ ɑ ː s ̪ ɛ n ̪]
* green:
[ɡ r i ː n ̪]

Sibilant
Occurrences:
38,194
Examples:
* bands:
[b a n ̪ d ̪ s ̪]
* sökas:
[s ̪ ø ː ˧ ˩ k a ˥ ˩ s ̪]
* sup:
[s ̪ ɵ p]
* tyst:
[t ̪ ʰ ʏ s ̪ t ̪]
Occurrences:
2,359
Examples:
* mars:
[m a ʂ]
* yrsel:
[y ː ʂ ɛ l]
* first:
[f ɪ ʂ t ̪]
* kors:
[k ʰ ɔ ˧ ˩ ʂ]
Occurrences:
1,376
Examples:
* kölen:
[ɕ ø ː ˧ ˩ l ɛ n ̪]
* köper:
[ɕ ø ː p ɛ r]
* kyss:
[ɕ ʏ s ̪]
* kemp:
[ɕ ɛ m p]

Fricative
Occurrences:
10,927
Examples:
* field:
[f ɪ ɛ l d ̪]
* furu:
[f ʉ ː r ɵ]
* fläck:
[f l ɛ k]
* färd:
[f ɛ ː ɖ]
Occurrences:
6,227
Examples:
* bath:
[b ɑ ː ˧ ˩ t ̪ h]
* hörda:
[h œ ɖ a]
* hålen:
[h o ː l ɛ n ̪]
* hells:
[h ɛ l s ̪]

Approximant
Occurrences:
11,757
Examples:
* kivik:
[ɕ i ː ʋ i ː k]
* virka:
[ʋ ɪ r k ʰ a]
* vlore:
[ʋ l ʊ r e ː]
* veva:
[ʋ e ː ˧ ˩ ʋ a ˥ ˩]
Occurrences:
5,785
Examples:
* jakob:
[j ɑ ː k ʊ b]
* jokk:
[j ɔ k]
* slöjd:
[s ̪ l œ j d ̪]
* bajs:
[b a j s ̪]

Trill
Occurrences:
43,405
Examples:
* borg:
[b ɔ r j]
* duer:
[d ̪ ʉ ː ɛ r]
* grava:
[ɡ r ɑ ː ʋ a]
* larm:
[l a r m]

Lateral
Occurrences:
25,226
Examples:
* blekt:
[b l ɛ k t ̪]
* kläs:
[k l ɛ ː s ̪]
* bildt:
[b ɪ l d ̪ t ̪]
* kalk:
[k ʰ a l k]

Lateral Tap
Occurrences:
407
Examples:
* jarl:
[j ɑ ː ɭ]
* sorl:
[s ̪ o ː ɭ]
* jarro:
[j ɑ ː ɭ s ̪ r ʊ]
* girls:
[j ɪ ɭ s ̪]

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:
0
Examples:
* pris:
[p r i ː s ̪]
* lita:
[l i ː t ̪ a]
* motiv:
[m u ː t ̪ i ː ʋ]
* dcjd:
[d ̪ e ː s ̪ e ː j i ː d ̪ e ː]
Occurrences:
7,176
Examples:
* viner:
[ʋ i ː ˥ ˧ n ̪ ɛ ˩ r]
* likud:
[l i ː ˥ ˧ k ʉ ː d ̪]
* vrida:
[ʋ r i ː d ̪ a]
* sidi:
[s ̪ i ː ˧ ˩ d ̪ i ː]
Occurrences:
0
Examples:
* hyrd:
[h y ː ˥ ˩ ɖ]
* yrsel:
[y ː ʂ ɛ l]
* easy:
[e ː a s ̪ y ː]
* mercy:
[m ɛ r s ̪ y ː]
Occurrences:
1,780
Examples:
* levy:
[l e ː ʋ y ː]
* hyrda:
[h y ː ˥ ˩ ɖ a]
* stay:
[s ̪ t ̪ a ˧ ˩ y ː]
* kenya:
[ɕ ɛ n ̪ y ː ˧ ˩ a ˥ ˩]
Occurrences:
0
Examples:
* bjuds:
[b j ʉ ː ˧ ˩ d ̪ s ̪]
* sura:
[s ̪ ʉ ː ˧ ˩ r a ˥ ˩]
* ljuv:
[j ʉ ː ʋ]
* puka:
[p ʰ ʉ ː ˧ ˩ k a ˥ ˩]
Occurrences:
4,347
Examples:
* buren:
[b ʉ ː ˥ ˧ r ɛ ˩ n ̪]
* judit:
[j ʉ ː ˥ ˩ d ̪ ɪ t ̪]
* judar:
[j ʉ ː ˥ ˩ d ̪ a r]
* slugt:
[s ̪ l ʉ ː ˥ ˧ k t ̪]
Occurrences:
0
Examples:
* groda:
[ɡ r u ː d ̪ a]
* krona:
[k r u ː n ̪ a]
* color:
[k ʰ u ː ˧ ˩ l ʊ r]
* doune:
[d ̪ u ː ɵ n ̪ ɛ]
Occurrences:
5,733
Examples:
* sotis:
[s ̪ u ː ˥ ˧ t ̪ ɪ ˩ s ̪]
* icon:
[ɪ s ̪ u ː n ̪]
* ove:
[u ː ˧ ˩ ʋ ɛ ˥ ˩]
* optis:
[u ː ˧ ˩ p t ̪ ɪ ˩ s ̪]
Occurrences:
22,982
Examples:
* innan:
[ɪ ˧ ˩ n ̪ a ˥ ˩ n ̪]
* lewis:
[l e ː ʋ ɪ ˩ s ̪]
* sotis:
[s ̪ u ː ˥ ˧ t ̪ ɪ ˩ s ̪]
* eniga:
[e ː n ̪ ɪ ɡ a]
Occurrences:
2,942
Examples:
* vegby:
[ʋ ɛ ˧ ˩ b ʏ ˥ ˩]
* conny:
[k ʰ ɔ n ̪ ʏ ˩]
* lilly:
[l ɪ l ʏ ˩]
* styva:
[s ̪ t ̪ ʏ ʋ a ˥ ˩]
Occurrences:
6,696
Examples:
* fotot:
[f u ː t ̪ ʊ t ̪]
* älgko:
[ɛ l j k ʊ ˩]
* dusko:
[d ̪ ɵ s ̪ k ʊ ˩]
* orsa:
[ʊ ˧ ˩ ʂ a ˥ ˩]

Close-Mid
Occurrences:
0
Examples:
* ange:
[a n ̪ j e ː]
* rfv:
[ɛ r ɛ f ʋ e ː]
* felen:
[f e ː l ɛ n ̪]
* lenin:
[l e ː n ̪ ɪ n ̪]
Occurrences:
11,384
Examples:
* bredd:
[b r e ː ˥ ˩ d ̪]
* nylén:
[n ̪ y ː ˧ ˩ l e ː n ̪]
* greta:
[ɡ r e ː ˧ ˩ t ̪ a ˥ ˩]
* adf:
[ɑ ː d ̪ e ː ɛ f]
Occurrences:
0
Examples:
* söul:
[s ̪ ø ː ʉ ː l]
* örfil:
[ø ː r f i ː l]
* bilkö:
[b ɪ l k ʰ ø ː]
* utför:
[ʉ ː ˧ ˩ t ̪ f ø ː ˥ ˩ r]
Occurrences:
5,328
Examples:
* köra:
[k ʰ ø ː ˧ ˩ r a ˥ ˩]
* spöar:
[s ̪ p ø ː ˩ ɑ ː r]
* bössa:
[b ø ː ˥ ˧ ɑ ː]
* sjönk:
[ɧ ø ː ŋ k]
Occurrences:
8,372
Examples:
* uppgå:
[ɵ ˧ ˩ p ɡ o ː]
* munk:
[m ɵ ŋ k]
* tunna:
[t ̪ ʰ ɵ ˧ ˩ n ̪ a ˥ ˩]
* kurva:
[k ʰ ɵ ˥ ˩ r ʋ a]
Occurrences:
0
Examples:
* kds:
[k ʰ o ː d ̪ e ː ɛ s ̪]
* hôtel:
[h o ː t ̪ ɛ l]
* henån:
[h ɛ n ̪ o ː n ̪]
* skåla:
[ɧ o ː l a]
Occurrences:
4,813
Examples:
* fåra:
[f o ː ˧ ˩ r a ˥ ˩]
* tablå:
[t ̪ ʰ a b l o ː ˥ ˩]
* åtala:
[o ː ˧ ˩ t ̪ a ˥ ˩ l a]
* svår:
[s ̪ ʋ o ː ˥ ˧ r]

Open-Mid
Occurrences:
47,840
Examples:
* väcks:
[ʋ ɛ ˧ ˩ k s ̪]
* svept:
[s ̪ ʋ ɛ ˥ ˧ p t ̪]
* svepa:
[s ̪ ʋ ɛ ˥ ˧ p a]
* hjärt:
[j ɛ ˧ ˩ ʈ]
Occurrences:
3,288
Examples:
* ihjäl:
[ɪ j ɛ ː l]
* gräl:
[ɡ r ɛ ː l]
* fläta:
[f l ɛ ː ˥ ˩ t ̪ a ˥ ˩]
* bärga:
[b ɛ ː ˥ ˩ r ɡ a]
Occurrences:
3,159
Examples:
* född:
[f œ d ̪]
* nöff:
[n ̪ œ f]
* östre:
[œ s ̪ t ̪ r ɛ]
* rört:
[r œ ˥ ˩ ʈ]
Occurrences:
10,312
Examples:
* tvåor:
[t ̪ ʋ ɔ ˧ ˩ ɔ ˥ ˩ r]
* åtta:
[ɔ ˧ ˩ t ̪ ʰ a ˥ ˩]
* lånet:
[l ɔ n ̪ ɛ t ̪]
* komet:
[k ʰ ɔ m e ː t ̪]

Open
Occurrences:
41,352
Examples:
* anad:
[a n ̪ a d ̪]
* aktat:
[a k t ̪ a t ̪]
* ruta:
[r ʉ ː ˧ ˩ t ̪ a ˥ ˩]
* satsa:
[s ̪ a ˧ ˩ t ̪ s ̪ a]
Occurrences:
0
Examples:
* tala:
[t ̪ ʰ ɑ ː l a]
* rara:
[r ɑ ː r a]
* vanan:
[ʋ ɑ ː n ̪ a n ̪]
* smal:
[s ̪ m ɑ ː l]
Occurrences:
12,465
Examples:
* haag:
[h a ɑ ː ˥ ˩ ɡ]
* roade:
[r ʊ ɑ ː ˥ ˩ d ̪ ɛ]
* hake:
[h ɑ ː k ɛ]
* ads:
[ɑ ː d ̪ e ː ɛ s ̪]

Tones#

  • ˥˧

  • ˥˩

  • ˧˩

  • ˩

On this page
Show Source