What to expect from Hawaiian
Hawaiian is one of the least-resourced languages in the speech model. The draft will contain many errors and, on difficult audio, may be closer to a sound-alike guess than a transcript. It works best on slow, clear speech from one speaker: a kumu reading a text, a prepared talk, a careful interview.
With that in mind, a timed draft can still speed up work on long recordings, because you correct lines while the player follows along instead of typing everything from the beginning.
Upload the recording as it is: MP3, M4A, WAV, OGG, OPUS, FLAC, a voice message or a phone recording. You can also record from your microphone in the upload box. Nothing needs converting first.
ʻOkina and kahakō
Modern Hawaiian spelling uses the ʻokina (glottal stop) and the kahakō (macron on long vowels, as in ā, ē, ī, ō, ū). Both change meaning, and both are the most likely things to be missing or misplaced in the draft. Plan to add them during review.
Exports are UTF-8, so the characters you add are kept in Word, subtitles and CSV. When you type corrections, use the proper ʻokina character (U+02BB) rather than an apostrophe if your publication requires it.
The transcript opens next to a player that highlights the current line. Click a line to hear it again, correct the word in place, and your changes save automatically.
Hawaiian, Pidgin and English
Many recordings mix Hawaiian with English, and with Hawaiʻi Creole English (Pidgin), which is a different language from Hawaiian. Choose the language by what dominates:
- Mostly Hawaiian: keep Hawaiian selected and fix English words as needed.
- Mostly English or Pidgin with Hawaiian words and place names: choose English. Pidgin is not a separate option.
- Mele and oli (song and chant) are very hard for any speech model; expect to rewrite most of the text.
Who records in Hawaiian
Typical files are interviews with kūpuna for family and community history, archival radio and oral-history recordings, immersion-school and university language materials, and talks at cultural events. A rough draft gives teachers and archivists something to annotate.
Export TXT, DOCX, PDF, CSV, SRT or VTT, with timestamps and speaker names switched on or off per export. Speaker labels are available on paid plans or with a Minute Pack.
For video files, see the Hawaiian video to text page.