Korean IPA Transcription Tool

Transcribe Korean Hangul to IPA phonemic notation with assimilation rules

Convert Korean Hangul to IPA symbols in your browser, automatically applying Korean phonological rules such as nasalization, liquidization, aspiration, and palatalization across syllable boundaries. It runs free in your browser on Gera Tools, with nothing uploaded.

Last updated Source: Gera Tools

Does it handle nasalization?

Yes. When a stop coda such as /k t p/ meets a following nasal onset /n m/, the coda nasalizes to /ŋ n m/. The tool detects these boundaries and rewrites both consonants before producing IPA.

Korean IPA transcription converts Hangul into International Phonetic Alphabet symbols, applying the phonological sandhi rules that make written and spoken Korean diverge. Because Hangul is featural, each syllable block decomposes cleanly into an onset (초성), nucleus (중성), and optional coda (종성), which makes algorithmic transcription reliable.

How it works

Each syllable in the Unicode Hangul Syllables block (U+AC00–U+D7A3) is decomposed arithmetically. Given a code point c, the index is c - 0xAC00; the onset is index / 588, the nucleus is (index % 588) / 28, and the coda is index % 28. Each jamo index maps to a base IPA value, then assimilation runs across boundaries:

coda /k t p/ + onset /n m/  →  coda nasalizes to /ŋ n m/   (nasalization)
coda /n/ + onset /l/  or  /l/ + /n/  →  both become /l/    (liquidization)
coda /h/ + onset stop, or stop + /h/  →  aspirated /kʰ tʰ pʰ/ (aspiration)
coda /t/ (ㄷ/ㅌ) + onset /i,j/  →  palatalize to /tɕ tɕʰ/   (palatalization)

Obstruent codas are unreleased and marked with the no-audible-release diacritic. Before a vowel-initial syllable the coda resyllabifies into the next onset.

Why Korean requires sandhi rules

Written Hangul represents the underlying morphological form, not the surface pronunciation. This is a deliberate orthographic choice: the word 국물 (soup stock) is spelled with ㄱ as the coda of 국 and ㅁ as the onset of 물, but spoken Korean assimilates the stop to match the following nasal — so it is actually said [kuŋ.mul]. Without applying these rules, an IPA transcription of the spelling would give a pronunciation nobody actually uses.

The four main sandhi processes encoded in this tool are:

Nasalization. Obstruent codas (ㄱ, ㄷ, ㅂ) become nasals before nasal onsets (ㄴ, ㅁ). This is pervasive — it affects thousands of common words and word boundaries.

Liquidization. When /n/ and /l/ meet at a syllable boundary, both surface as [l]. The direction matters: ㄹ + ㄴ and ㄴ + ㄹ both yield [ll].

Aspiration. ㅎ combines with adjacent plain stops to produce their aspirated counterparts. ㄱ + ㅎ or ㅎ + ㄱ both surface as [kʰ].

Palatalization. Coronal stops ㄷ and ㅌ before /i/ or /j/ shift forward in articulation to [tɕ] and [tɕʰ], the sounds spelled with ㅈ and ㅊ.

Worked examples

HangulUnderlyingIPA (after rules)Rule applied
국물/kuk.mul/[kuŋ.mul]Nasalization (ㄱ → ŋ before ㅁ)
신라/sin.la/[ɕil.la]Liquidization (n+l → ll)
좋다/tɕoh.ta/[tɕoʰ.ta]Aspiration (ㅎ + ㄷ → tʰ)
해돋이/hɛ.tot.i/[hɛ.to.tɕi]Palatalization + resyllabification

This is a broad phonemic transcription intended for learners and linguists; allophonic vowel raising (e.g., 의 in unstressed position) and intonation contours are not marked.